<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:cc="http://cyber.law.harvard.edu/rss/creativeCommonsRssModule.html">
    <channel>
        <title><![CDATA[Stories by Poli Dey Bhavsar on Medium]]></title>
        <description><![CDATA[Stories by Poli Dey Bhavsar on Medium]]></description>
        <link>https://medium.com/@bhavsar-poli?source=rss-28d4f7b81521------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*pC76J_zMRQpgI_dtzZkS3w.jpeg</url>
            <title>Stories by Poli Dey Bhavsar on Medium</title>
            <link>https://medium.com/@bhavsar-poli?source=rss-28d4f7b81521------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Fri, 20 Nov 2026 01:05:40 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@bhavsar-poli/feed" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[Cracking Bayes’ Theorem: When Understanding a Concept Isn’t the Same as Solving It]]></title>
            <link>https://bhavsar-poli.medium.com/cracking-bayes-theorem-when-understanding-a-concept-isn-t-the-same-as-solving-it-38a8eefc0a19?source=rss-28d4f7b81521------2</link>
            <guid isPermaLink="false">https://medium.com/p/38a8eefc0a19</guid>
            <category><![CDATA[bayes-theorem]]></category>
            <category><![CDATA[bioinformatics]]></category>
            <category><![CDATA[career-change]]></category>
            <dc:creator><![CDATA[Poli Dey Bhavsar]]></dc:creator>
            <pubDate>Thu, 19 Nov 2026 12:43:53 GMT</pubDate>
            <atom:updated>2026-11-19T14:26:32.735Z</atom:updated>
            <content:encoded><![CDATA[<p><em>I understood the idea. The numericals were another story.</em></p><p>If you read my last post on why I moved from Marketing to Bioinformatics, you know I’ve spent the last while relearning how to learn, quant subjects especially. This post is about one specific concept that humbled me more than I expected: <strong>Bayes’ Theorem.</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*XdsbXwKIafp92I8YpX51_A.png" /><figcaption><strong>Bayes’ Theorem: A simple formula, a powerful way of thinking — and a surprisingly challenging one to master.</strong></figcaption></figure><h3>The Challenge</h3><p>Bayes’ Theorem is one of those ideas that feels deceptively simple the first time you meet it. Prior, likelihood, evidence, posterior. Say it out loud and it sounds like common sense: update your belief when new evidence comes in. Easy, right?</p><p>Then I sat down to actually solve a numerical. This is the kind of problem that stopped me cold:</p><blockquote>At a specific genomic locus in a diploid sample, a candidate variant is equally likely to be a true heterozygous somatic mutation or a true homozygous-reference genotype (prior probability = 0.5 for each). A low-coverage targeted sequencing run is performed at this site. Because of allelic dropout and read sampling, the variant caller detects the alternate allele in 3/5 of true heterozygous cases, missing it in 2/5 of cases. Homozygous reference sites never produce a positive variant call under this strict filter.</blockquote><blockquote>Given that the variant caller returns a negative (reference-only) call, what is the probability that the sample actually harbours the true heterozygous mutation?</blockquote><p>I could name every piece in that problem. What I kept fumbling was the setup: which number is the prior, which is the likelihood, and (the trap in this one) that the question asks for the probability of a mutation given a negative call, not the other way around.</p><p>That gap between “I get the concept” and “I can compute the answer” is more common than people admit, and almost nobody warns you about it. Bayes’ Theorem in particular has a way of making you feel smart in a sentence and lost in a formula.</p><h3>Going Back to the Fundamentals</h3><p>Instead of grinding through more practice problems, I went back a few steps. NPTEL’s statistics and machine learning courses gave me the scaffolding I was missing: conditional probability, total probability, and how they fit together before Bayes even enters the picture.</p><p>If you’re stuck the way I was, the fix may not be more repetition on the same kind of problem. It may be rebuilding the layer underneath it.</p><h3>The Blog That Made Me Actually Enjoy It</h3><p>Somewhere along the way, I stumbled upon Eliezer Yudkowsky’s <a href="https://www.yudkowsky.net/rational/bayes">An Intuitive Explanation of Bayes’ Theorem</a>, and it did something I didn’t expect: it made me laugh while explaining Bayes. It opens by teasing the reader about friends who treat the theorem like the secret of the universe, and even imagines someone whose crush will only date Bayesians. Then comes the line that hooked me:</p><blockquote>Soon you will know. Soon you will be one of us.</blockquote><p>There’s a reason humour works so well for dense concepts like this: it lowers your guard. You stop treating the material like something to be endured and start engaging with it the way you’d engage with a good story. I found myself re-reading sections not because I had to, but because I wanted to see how it would explain the next part.</p><h3>The Video That Sealed It</h3><p>If the blog opened the door, this video is what walked me all the way through it.</p><blockquote><a href="https://www.youtube.com/watch?v=Ab4viREnP74&amp;t=677s">Bayes’ Theorem | TRICK that NEVER fails | Solved Examples</a> by Six Sigma Pro SMART</blockquote><p>Most explanations of Bayes’ Theorem get bogged down in notation before you ever see how the pieces move. What makes this walkthrough click is its structure and pacing: it pairs a single, reliable problem-solving framework (“the trick”) with fully worked-out examples. Instead of guessing which number is the prior and which is the likelihood, you learn a mechanical way to set up the problem every single time, turning a concept that usually trips people up into something you can solve confidently.</p><p>After watching it, I went back to the numericals that had stumped me earlier and solved them without hesitation. That’s when I knew the concept had actually settled in, not just been memorized.</p><h3>Putting It to Work: When a Negative Variant Call Doesn’t Rule Out a Mutation</h3><p>Here’s the problem from earlier, solved the way the video taught me.</p><p>In low-coverage sequencing, a reference-only call is easy to misread as “nothing there.” When allelic dropout and uneven read sampling enter the picture, a true heterozygous somatic mutation can hide in plain sight, and Bayes’ Theorem gives us an exact way to quantify how much a negative call should change our belief.</p><p><em>A caveat for fellow bioinformaticians: this is a deliberately simplified toy example. Real somatic variants are often subclonal, and real callers are rarely perfect in both directions.</em></p><h4>1. Translating the biology into clear events</h4><p>Before touching any formula, define the two competing hypotheses and the observed outcome:</p><p>• <strong>M (true heterozygous somatic mutation): </strong>the alternate allele is genuinely present at the locus in one chromosome copy.</p><p>• <strong>R (homozygous reference): </strong>both chromosome copies match the reference genome; there is no true somatic variant at this locus.</p><p>• <strong>N (negative / reference-only call): </strong>the variant caller does not detect the alternate allele.</p><h4>2. Listing what we know</h4><p>First, the priors: what we believed before looking at this sequencing call.</p><p><strong>P(M) = 0.5</strong></p><p><strong>P(R) = 0.5</strong></p><p>Next, the likelihoods: how probable a negative call is under each scenario.</p><p>If the mutation is truly present, assume the variant caller detects it 3/5 of the time and misses it 2/5 of the time. Thus, for this example, the probability of a negative call is 0.4:</p><p><strong>P(N | M) = 2/5 = 0.4</strong></p><p>If the site is truly homozygous reference, assume the strict filter never produces a positive variant call. Therefore, a negative call occurs every time under this hypothesis:</p><p><strong>P(N | R) = 1.0</strong></p><h4>3. Plugging the numbers into Bayes’ Theorem</h4><p>We want the posterior: the chance of a true heterozygous mutation given that the caller returned a negative result, written P(M | N).</p><p><strong>P(M | N) = [P(N | M) × P(M)] / [P(N | M) × P(M) + P(N | R) × P(R)]</strong></p><p>Now substitute the values carefully.</p><p><strong>Numerator</strong> (true mutation and a negative call):</p><p><strong>0.4 × 0.5 = 0.2</strong></p><p><strong>Denominator</strong> (all the ways a negative call can happen):</p><p><strong>(0.4 × 0.5) + (1.0 × 0.5) = 0.2 + 0.5 = 0.7</strong></p><p>Put them together:</p><p><strong>P(M | N) = 0.2 / 0.7 = 2/7 ≈ 0.2857, or about 28.6%</strong></p><h4>4. Understanding what this result actually means</h4><p>Under the assumptions of this simplified example, even after the variant caller returns a negative result, the posterior probability of a true heterozygous somatic mutation remains approximately 28.6%. In other words, the negative result reduces our initial belief from 50% to 28.6%, but does not eliminate the possibility of a mutation.</p><p>Why isn’t it close to zero? Because a negative call can happen in two ways:</p><p>1. There truly is no mutation: 0.5 prior probability × 1.0 likelihood of a negative call = 0.5.</p><p>2. There is a mutation, but the low-coverage sequencing run misses it: 0.5 prior probability × 0.4 likelihood of a negative call = 0.2.</p><p>Once you see those two pieces side by side, 0.2 / (0.2 + 0.5) stops feeling like a magic formula and starts feeling like straightforward accounting of possibilities.</p><h3>Closing the Loop</h3><p>Looking back, this whole exercise mirrors the bigger shift I wrote about in my last post. Moving from Marketing to Bioinformatics was never just about switching industries. It’s been about relearning how to learn technical material from scratch, without shortcuts, and without pretending to understand something before I actually did.</p><p>Bayes’ Theorem was a small test case for that bigger process. Struggle honestly, find the right resources (even unconventional ones like a funny blog post), and give the concept time to actually land.</p><p>If you’re stuck on Bayes’ Theorem right now, or on anything that feels conceptually obvious but practically slippery, I hope this is proof that the fog does lift. It just takes the right combination of resources and a little patience with yourself.</p><p><em>Have a resource that made a stubborn concept click for you? I’d love to hear about it in the comments.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=38a8eefc0a19" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[From Marketing to Bioinformatics — Why I’m Making the Switch]]></title>
            <link>https://bhavsar-poli.medium.com/from-marketing-to-bioinformatics-why-im-making-the-switch-cc15fec5f0d2?source=rss-28d4f7b81521------2</link>
            <guid isPermaLink="false">https://medium.com/p/cc15fec5f0d2</guid>
            <category><![CDATA[career-transitions]]></category>
            <category><![CDATA[life-sciences]]></category>
            <category><![CDATA[computational-biology]]></category>
            <category><![CDATA[bioinformatics]]></category>
            <category><![CDATA[never-stop-learning]]></category>
            <dc:creator><![CDATA[Poli Dey Bhavsar]]></dc:creator>
            <pubDate>Sun, 19 Jul 2026 14:43:26 GMT</pubDate>
            <atom:updated>2026-11-15T06:32:01.788Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="The image shows a double helical structure of a DNA molecule" src="https://cdn-images-1.medium.com/max/1024/1*AfDgvHTjN3wTMuyJo7V7Vw.jpeg" /><figcaption>Two strands, running parallel, always connected — science and life have more in common than we think. The duality never really left. It was just waiting for the right conditions to express itself. (<em>Photo by </em><a href="https://unsplash.com/@3dparadise?utm_source=unsplash&amp;utm_medium=referral&amp;utm_content=creditCopyText"><em>Braňo</em></a><em> on </em><a href="https://unsplash.com/photos/dna-double-helix-render-on-blue-Mm1VIPqd0OA?utm_source=unsplash&amp;utm_medium=referral&amp;utm_content=creditCopyText"><em>Unsplash</em></a><em>)</em></figcaption></figure><p>There’s a particular kind of restlessness that doesn’t announce itself loudly. It doesn’t show up as a dramatic moment of clarity or a single conversation that changes everything. It just sits quietly somewhere in the background, patient, like a tab you never quite close.</p><p>That’s the best way I can describe the last two decades of my life.</p><h4><strong>It started with a double life</strong></h4><p>In the early 2000s, I was finishing my Master’s in Zoology at Bangalore University when I enrolled in a one-year Diploma in Bioinformatics at STG International — a private institute that had a partnership with Red Hat. At the same time, I had just joined a publishing company doing typesetting for scientific journals from Cambridge University Press.</p><p>So there I was: commuting to a full-time job during the day, going to class in the evenings, and somewhere in the middle of all that, trying to keep up with both. It sounds exhausting now. Honestly, it was. But there was something about Bioinformatics specifically — the idea that you could sit at a computer and <em>make sense of life itself</em>, sequence by sequence — that kept pulling me forward.</p><p>I finished the diploma. I was proud of it.</p><p>And then, quietly, I had to set it aside.</p><h4><strong>The practical decision</strong></h4><p>The honest truth was that Bioinformatics jobs in India at that time were scarce. The field existed, but the ecosystem hadn’t caught up yet. I had a job, a degree, real responsibilities — and a career path that was actually open in front of me. So I walked down it.</p><p>That path eventually took me further than I had expected — including to Auckland, New Zealand, where I completed a full-time MBA in International Business. It was one of those experiences that quietly reshapes how you see everything. Working and studying across cultures, learning to communicate with people whose professional contexts and assumptions were entirely different from my own, understanding that good work travels across borders when you understand the people on the other side — all of that became part of how I operate. Back in India, my marketing career reflected it: I worked for companies serving international clients, and that cross-cultural fluency turned out to matter more than any single technical skill.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*FxRrfcy_xYLMuaz3dvRTCQ.jpeg" /><figcaption><strong>Google Webmaster Conference 2019, Mumbai — 19 July 2019.</strong> A memorable day of learning, sharing ideas, and connecting with content creators from across India.</figcaption></figure><p>During this phase of my career, I was also selected to attend the <strong>Google Webmaster Conference 2019 in Mumbai</strong>. It was a wonderful opportunity to meet and interact with fellow content creators from across India, exchange experiences, and gain fresh perspectives on the evolving world of search and digital content. It remains one of the memorable experiences of my marketing journey.</p><p>A career took shape. On paper, it looked like a clean narrative — someone who had found their lane and stayed in it.</p><p>But it never quite felt that way from the inside.</p><p>I can’t point to a single moment during those years when Bioinformatics came rushing back to me. No documentary. No headline. No conversation with a colleague that made me think <em>that should have been me.</em> It was more like a feeling that sat in the back of my chest, steady and low, surfacing occasionally and then receding. Unfinished business, not urgent — but never gone.</p><h4><strong>The years in between — and why they weren’t wasted</strong></h4><p>Here’s something I’ve come to believe: a detour is only wasted if you sleepwalk through it.</p><p>I didn’t. Throughout my marketing career, upskilling was a constant — not because someone told me to, but because I genuinely couldn’t help it. I pursued courses through IxDF and Udemy, along with a mini-degree through CXL, exploring areas such as marketing, conversion optimisation, and user experience. Not necessarily to change careers, but because I was curious about what makes people think, feel, and respond to the things they interact with.</p><p>Understanding the user experience made me a sharper, more empathetic marketer. It changed how I structured content, how I thought about client needs, and how I asked questions before jumping to solutions.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Ky7AU9ZKRbyXABmLtqt2fA.png" /><figcaption><strong>A few milestones, many lessons, and countless moments of curiosity that made the journey worthwhile.</strong></figcaption></figure><p>Life also had its own plans during this period — years that asked more of me personally than professionally, and slowed things down in ways I hadn’t anticipated. But those years, difficult as some of them were, only made the pull toward Bioinformatics feel more necessary, not less.</p><p>I came through that stretch with a different kind of clarity. The kind you don’t get from reading about it.</p><h4><strong>The slow return</strong></h4><p>A few years ago, while still working full-time as a marketer and content creator, I made a quiet decision: I enrolled in a Master of Computer Applications through an Open University — IGNOU.</p><p>It wasn’t glamorous. Open university courses while holding down a full-time job rarely are. But I wasn’t chasing glamour — I was trying to rebuild something. Specifically, the computational foundation I knew I would need if I ever wanted to re-enter the life sciences space seriously. I couldn’t go back to a wet lab. That window had closed practically, even if not emotionally. But computational biology — that was a door I could open from where I was standing.</p><p>I completed the MCA recently. And alongside it, I had been quietly working through NPTEL courses in Computational Biology, picking up what I could, piecing things back together.</p><p>It was a gradual slow burn. No epiphany. Just a steady accumulation of choices that all pointed in the same direction.</p><h4><strong>Why Bioinformatics, and why now</strong></h4><p>Bioinformatics sits at an unusual intersection. It demands statistics, mathematics, programming, and biology — all at once. For most people entering the field fresh, that sounds like a mountain. For me, returning to it with a background in zoology, a diploma in the field, a fresh MCA, and years of structured thinking from a global corporate career, it feels less like starting over and more like assembling pieces that were always meant to fit together.</p><p>The cross-cultural communication skills, the UX-informed empathy, the ability to work with international teams and clients — none of that disappears when I open a genome browser or write a Python script. If anything, I think it makes me a different kind of contributor in this space. Science doesn’t happen in isolation, and the ability to think clearly about how people understand and use information is not a soft skill. It’s increasingly essential.</p><p>I’m also entering this at a genuinely exciting moment. Genomics, proteomics, single-cell RNA sequencing, drug discovery pipelines powered by machine learning — the field has moved enormously since those early days. The ecosystem that didn’t exist when I first tried to enter it has now exploded. The timing, which once felt like my obstacle, now feels like an unexpected advantage: I’m returning to a field that has finally grown into its own potential.</p><h4><strong>What comes next — and why I’m writing about it</strong></h4><p>I’ll be honest: I don’t know exactly how this transition will unfold. I don’t have a job offer in hand or a research position waiting. What I have is a background that spans more disciplines than most people would have deliberately collected, a curiosity that never actually went away, and a willingness to document the process openly.</p><p>That’s what this blog series is going to be. Not a polished after-the-fact success story, but a real-time account — the learning, the confusion, the small wins, and the moments where I have to sit with something difficult before it clicks. If you’re someone who’s building in an interdisciplinary space, or thinking about a return to something you once had to set aside, I hope some of this resonates.</p><p>It’s never too late to return to the work that always felt like yours.</p><p>I’m living proof that the feeling doesn’t go away. And I’ve decided to stop waiting for a better moment to act on it.</p><p><em>This is the first post in a series documenting my transition into Bioinformatics and Computational Biology. If you’ve made a similar journey — across fields, across years — I’d love to hear from you in the comments.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=cc15fec5f0d2" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[4 Different Optimization Approaches to Make Your Customers Convert — Review]]></title>
            <link>https://bhavsar-poli.medium.com/4-different-optimization-approaches-to-make-your-customers-convert-review-f238ae22709?source=rss-28d4f7b81521------2</link>
            <guid isPermaLink="false">https://medium.com/p/f238ae22709</guid>
            <category><![CDATA[b2b]]></category>
            <category><![CDATA[conversion-optimization]]></category>
            <category><![CDATA[customer-value]]></category>
            <category><![CDATA[neuromarketing]]></category>
            <category><![CDATA[digital-psychology]]></category>
            <dc:creator><![CDATA[Poli Dey Bhavsar]]></dc:creator>
            <pubDate>Sun, 09 May 2021 18:27:28 GMT</pubDate>
            <atom:updated>2021-05-10T09:30:42.336Z</atom:updated>
            <content:encoded><![CDATA[<h3><strong>4 Different Optimization Approaches to Make Your Customers Convert — Review</strong></h3><p>Not all optimization efforts are always profitable! Are you in the same boat? Then you’re on the right page. Today I am going to share with you four different optimization approaches that are proven and tested by industry veterans. You can also use them to make your customers take action and boost your revenue.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*qiedfHXSHOKZIQbgk7F-GA.jpeg" /><figcaption><a href="https://unsplash.com/@xpsteven?utm_source=unsplash&amp;utm_medium=referral&amp;utm_content=creditCopyText">Image</a>: 4 Different Optimization Approaches for four different scenarios.</figcaption></figure><h3><strong>#1 An optimizing strategy for B2B scenarios</strong></h3><p>According to Bill Leake, a marketing expert, McKinsey &amp; Co. alumni and Dell Computer veteran, B2B sales normally occur on a much longer sales cycle. They are often conducted in person and it’s crucial to take into account the persona of the person visiting a B2B website.</p><p>Here are 5 key points you must consider while optimizing for a B2B website:</p><ol><li><strong>Define your goals</strong> — The first thing you need to do is to set your goals. It is crucial and you must also decide how you’re going to measure them. Video engagement, time spent on-site and the pathing how they landed are some of the engagement metrics worth tracking.</li></ol><p>Also, Bill suggested three optimization goal categories to implement from the beginning:</p><ul><li><strong>Leads</strong></li><li><strong>Account-based marketing</strong></li><li><strong>Tracking for attribution</strong></li></ul><p>2. <strong>Determine &amp; prioritize lead type</strong> — In B2B wrong lead generation strategy might result in poor quality leads which in turn might aversely affect sales team’s attitude. Therefore, it’s important to understand your sales engine first and keep a balance between these three elements — cost, quality and quantity.</p><p>3. <strong>Identify your personas </strong>— Begin a persona analysis to identify your persona, i.e., do their psychological &amp; personality profiling and professional role assessment. Bill recommends targeting your personas based on the customers you want rather than customers you have. Once you have identified your desired customer, start optimizing from bottom funnel content up.</p><p>4. <strong>Optimize CTA and content</strong> — You can optimize your CTA based on what you’re selling, your product price, how would you want your customers to buy it while avoiding over optimizing for sales.</p><p>The form and style of your content would be shaped by its purpose — whether you want to educate or generate leads. Mobile-friendly content, video, third-party content, showing earned media wins are forms of content recommended by Bill for B2B sites.</p><p>5. <strong>Common mistakes to avoid</strong> — Lack of credibility in your content, poor language choices, auto-play videos, treating audience like B2C, not focusing on the needs of customers and talking about the business itself are some of the common mistakes digital marketers should avoid.</p><p>To get an in-depth knowledge of optimization strategies, enrol for the <a href="https://cxl.com/institute/programs/conversion-optimization/">Conversion Optimization minidegree at the CXL Institute</a>.</p><p>Now, let’s move on to the next approach.</p><h3><strong>#2 Customer Value Optimization Model — a 5-step framework</strong></h3><p>Now, I am going to share with you a proven and tested framework by Justin Rondeau, CEO at Digital Marketer, to convert your prospects across a range of industries. Here are the five steps:</p><p><strong>Step 1: Determining market fit</strong></p><p>This is the longest and most important step in this process. You have to dig into your customers ideal after-state, i.e., their improved life after using your product or service. And then you can construct your statement of value and connect that to your offering. However, still they might not want to buy from you due to multiple reasons including they don’t know you.</p><p><strong>Step 2: Generating leads</strong></p><p>In this step, Justin suggests acquiring quality prospects using a lead magnet. A good lead magnet is highly actionable, high value (both perceived and actual), easy to consume and a very specific chunk of information that is used in exchange for an opt-in.</p><p>Examples of lead magnets are checklists, case studies, free downloads, free software trials, quizzes, assessments, tests, etc.</p><p><strong>Step 3: Turning lead into a buyer</strong></p><p>Now, that you have attracted some quality prospects, it’s time to develop a strategy to convert them into customers. To do that you need to build trip wires aka “a tempting offer that is intended to convert prospects into buyers”. It’s aimed at alleviating acquisition costs. Some of the examples of trip wires are E-book, coupon for service, physical books, physical product sample, etc.</p><p><strong>Step 4: Turning that buyer into a repeat customer</strong></p><p>This is the monetization stage wherein you can maximize your profit by cross-selling and using a profit maximizer after your customer purchases your core offer.</p><p><strong>Step 5: Re-engaging lost customers</strong></p><p>Now that you have already recruited a solid customer base, it’s time to follow-up to the previous four steps. This step is known as the return path and you can build one by using three proven approaches:</p><ul><li>Automated follow-up</li><li>Exit offers</li><li>Remarketing via display networks and custom audiences.</li></ul><h3><strong>#3 Applying Psychology to technology for optimization</strong></h3><p>Now, I’ll tell you how we can apply persuasive psychological techniques to the various elements and components of our website to optimize it for better conversion.</p><p>Here’s where we can apply Psychology to technology:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/590/1*u7DSh7u9Z3k66Q9OYVT4Pw.png" /><figcaption><a href="http://cxl.com/">Image</a>: These are stages wherein you can apply Psychology to Technology.</figcaption></figure><p>We can inform our design using psychological and neuroscience concepts as shown in the image below:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*wVPTq-qNpMDLdZYTQjzcmg.png" /><figcaption><a href="http://cxl.com/">Image</a>: Applying concepts of Psychology and Neurology in Web Design</figcaption></figure><p>Actually, this course Digital Psychology and Behavioural Design Training by Dr. Brian Cugelman is very interesting. In this course, he teaches simulation of the real world created in the digital space using metaphors. For example, information architecture originally comes from Library Science.</p><p>The way we move around in physical spaces, metaphors are used to give us the same feeling of moving from one room to another in our digital experience.</p><p><strong>Directional cues</strong></p><p>Also, Dr. Cugelman explains the importance of directional cue, that is how human looking at form can increase conversions than human looking away from form.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*DLrY7gu8Ei6tDYkjPw7cYA.png" /><figcaption><a href="http://cxl.com/">Image</a>: Behavioral Psychology by Stage.</figcaption></figure><p>In the above image, Dr. Cugelman explains the state of mind of your customers according to Behavioral Psychology. You can now apply them as shown in the image and persuade your prospects to take action.</p><p>He also teaches how to do Psychology-inspired design as shown below:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*s7eh1FAfFVabMP-6CyOMtQ.png" /><figcaption><a href="http://cxl.com/">Image</a>: Psychology-inspired web design.</figcaption></figure><p>So, you have a very high-level idea of how you can apply psychology in your design and help your customers take actions.</p><p>Let’s get into the inner faculties of the human brain to understand what goes on in your customers’ brain and how you can use that knowledge to make more money.</p><h3><strong>#4 Understanding inner mechanisms of human brain to increase sales</strong></h3><p>Before we deep dive into the inner mechanisms of the human brain, let’s get ourselves on the same page about this one fact that to optimize conversion rates, you need to change human behaviour. How do we do that?</p><p>Please stay tuned as I give a glimpse of the course <a href="https://cxl.com/institute/online-course/applied-neuromarketing/">Applied Neuromarketing by André Morys at the CXL Institute</a>. I would recommend you to enrol for this course because without this knowledge, we’re actually shooting in the dark and wasting our marketing efforts.</p><p>Let’s begin with the notion that your customer will only stay on your site if they find <strong>relevant </strong>information, regardless of how cool or fancy your site aesthetics are. For that you would need to be more effective in communicating your message to your target audience. Here’s how!</p><p><strong>Messaging that gets into the mind of your audience</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/454/1*4OyRn_28VQHVRfFOqRaYEg.png" /><figcaption><a href="https://www.interaction-design.org/">Image</a>: The 3 Brain model to understand the workings of human brain.</figcaption></figure><p>This is known as the <strong>3-brains model </strong>and it would help you to <a href="https://bhavsar-poli.medium.com/lessons-from-neuromarketing-to-effectively-attract-persuade-sell-to-your-audience-review-db49d08eeb7d"><strong>understand how the human brain works</strong></a> and how you apply this knowledge to engage your audience and persuade them to take actions. I would suggest you read <a href="https://bhavsar-poli.medium.com/lessons-from-neuromarketing-to-effectively-attract-persuade-sell-to-your-audience-review-db49d08eeb7d">this blog</a> first before proceeding further.</p><p>Now that you know how our brain works you would be able to align your messaging according to that and also apply <strong>the power of implicit code</strong>. It is conveying your users what your product can do for them and what they’re with your product in an implicit manner mostly via images.</p><p>André Morys calls it the <strong>principle of emotional resonance</strong>. This implicit code is immediately perceived by your old brain or limbic brain and then passed on to your new brain for rationalization. He says that,</p><blockquote>“100% of decisions are made unconscious, but some decisions will be ‘post-rationalized’ more than others”</blockquote><p><strong>Control how your audience perceives your messaging</strong></p><p>So, how do you control attention? André Morys teaches <strong>seven easy to apply principles</strong> to control your customers’ attention.</p><ol><li>Contrast</li><li>Space</li><li>People/faces</li><li>Movement</li><li>Breaking rules</li><li>Directional cues</li><li>Person’s name/their own face</li></ol><p>And here he makes an important point:</p><blockquote>“90% of experiments fail because the change that was made on a website never reached a customer’s mind.”</blockquote><p><strong>Increasing your customers’ motivation to buy</strong></p><p>To do that you have to get rid of common personas that are not precise enough. Instead you have to come up with emotionally resonated personas using these four steps:</p><p><strong>Step 1: Eliminate conflicting values and focus on one value</strong></p><p><strong>Step 2: Set a clear value</strong></p><p><strong>Step 3: Remove unneeded information</strong></p><p><strong>Step 4: Deduce specific information about your product/service</strong></p><p>So, you have created personas that are emotionally focused by removing clutter. Now get them straight to the point to create emotional resonance. This will be helpful and valuable for your optimization efforts.</p><p>Use these personas permanently to challenge your optimization ideas and hypothesis.</p><p>Hope these four approaches would help you design customer-centric optimizations.</p><p>This is my last blog in the series of 12 blogs wherein I shared with you my learnings from the CRO program at the CXL Institute.</p><p>Next time I’ll see you with a blog with my takeaways from reading a book. <a href="https://www.linkedin.com/in/peeplaja/">Peep Laja </a>has recommended many books, and I’ll start reading one by one once I’m done with my CRO program.</p><p>Till then bye; stay safe!</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=f238ae22709" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Mastering A/B Testing: What to do Before, During & After Tests to Get More Wins — Review]]></title>
            <link>https://bhavsar-poli.medium.com/mastering-a-b-testing-what-to-do-before-during-after-tests-to-get-more-wins-review-118e7c253903?source=rss-28d4f7b81521------2</link>
            <guid isPermaLink="false">https://medium.com/p/118e7c253903</guid>
            <category><![CDATA[conversion-optimization]]></category>
            <category><![CDATA[statistical-significance]]></category>
            <category><![CDATA[split-testing]]></category>
            <category><![CDATA[a-b-testing]]></category>
            <category><![CDATA[null-hypotheses]]></category>
            <dc:creator><![CDATA[Poli Dey Bhavsar]]></dc:creator>
            <pubDate>Sun, 02 May 2021 18:19:10 GMT</pubDate>
            <atom:updated>2021-05-02T21:13:31.179Z</atom:updated>
            <content:encoded><![CDATA[<h3><strong>Mastering A/B Testing: What to do Before, During &amp; After Tests to Get More Wins — Review</strong></h3><p>As a CRO expert, wouldn’t you like to get more value out of your experimentation efforts? Well, A/B testing can be complex and you can encounter pitfalls!</p><p>With A/B testing tools evolving over time, we have come a long way when it comes to convenience, however, the success of your tests largely depends on you. How you use your most powerful tool — your brain — to design optimal testing strategies and determine the best outcome of your tests.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*2PHT3fHwG8pfvPVGSnTxZw.jpeg" /><figcaption><a href="https://unsplash.com/@mattseymour?utm_source=unsplash&amp;utm_medium=referral&amp;utm_content=creditCopyText">Image</a>: A/B Testing: Find your winner between Control vs. Challenger.</figcaption></figure><p>To get more wins out of your optimization efforts, please stay tuned. I’m going to share my learnings from the course <a href="https://cxl.com/institute/online-course/ab-testing-mastery/">A/B testing mastery</a> by <a href="https://www.linkedin.com/in/tonwesseling/">Ton Wesseling</a> at the CXL Institute. I’ll cover before the test, how you should plan, during the test, how you should execute and after the test, how you should make sense of the result.</p><h3><strong>Origin of A/B testing</strong></h3><p>A/B Testing, as the name suggests, is all about testing or comparing two versions of something, A (control) and B (Challenger), to find out which performs better. Do you know that it’s origin dates back to the early twentieth century? Yes, in the 1920’s statistician and biologist Ronald Fisher discovered the principles of A/B testing then known as controlled experiment.</p><p>This online testing method, however, has risen to popularity in the last couples of decades as companies began to evaluate almost everything from headlines to product descriptions to website design to CTA copy or button.</p><p>Let’s begin!</p><h3><strong>How to plan an A/B test?</strong></h3><p>While planning for your test, the most important question that arises is that: Do you have enough data to conduct A/B tests?</p><p>Ton Wesseling teaches in the course, how to use the ROAR model created by him to find an answer to this question.</p><p>ROAR is an acronym for:</p><p><strong>R</strong> = Risk</p><p><strong>O</strong> = Optimization</p><p><strong>A</strong> = Automation</p><p><strong>R</strong> = Re-think</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*53EhK78jo1m5YMKeeiJhYg.png" /><figcaption><a href="http://cxl.com/">Image</a>: ROAR Model to figure out if you have enough data to conduct A/B tests</figcaption></figure><p>As per the ROAR model, if you are below 1000 conversions you cannot run a A/B test. You can use this <a href="https://abtestguide.com/abtestsize/">free A/B test size calculator</a> to find out the minimum sample size and the duration you have to run your tests and more.</p><p>If you have below 1000 conversions and you still run an A/B test, it would be really hard to find a winner with too low data. Even if you find a winner, chances are high that it’s not the real winner, i.e., it’s a false positive.</p><p>Why so? Because, at 1000 conversions your challenger needs to beat the control with 15 percent. A 15% uplift is needed to be recognised as a winner.</p><p>As per the model, when you have 10,000 conversions, you need 5 percent uplift to find a winning A/B test outcome.</p><p><strong>Decide your goal metric</strong></p><p>Next, it is highly important for you to have a goal metric to which everyone in your company is on the same page. Therefore, you need to have an overall evaluation criterion (OEC). For example, for a financial institution, the number of active monthly users of their banking solutions is their overall evaluation criterion.</p><p>Remember Power &amp; Significance rule of thumb</p><ul><li><strong>Power</strong> — Try to test on pages with a high power, i.e., &gt;80% to avoid chances of ending up with <strong>false negatives</strong>. In other words, when there is an effect to be detected you won’t detect them if power is lower.</li><li><strong>Significance</strong> — Try to test against a high enough significance level (90%), otherwise you’ll end up with <strong>false positives</strong>, i.e., you’ll declare winners when in reality there isn’t an effect.</li></ul><p><strong>Generate user behaviour insights with 6V research model</strong></p><p>There’s no point in starting your A/B tests if you’re not trying to solve any problem. Problems that add friction in your customers’ journeys. Understanding user behaviour is crucial as it would help you with inputs to set your hypothesis. The 6Vs in this model are Value, Versus, View, Validated, Verified and Voice.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*eSpkdRx2gwd18iDnub0YrQ.png" /><figcaption><a href="http://cxl.com/">Image</a>: 6V Research Model to get insights on user behaviour.</figcaption></figure><p>Once you are done with all your research, you’re all set to write your hypothesis.</p><p><strong>Write a proper hypothesis</strong></p><p>Wondering why you need a hypothesis? Well, hypothesis gets everyone in your company aligned on why you’re running this growth or research experiment in the first place. A proper hypothesis does 3 things:</p><blockquote>Describes a problem</blockquote><blockquote>Proposes a solution</blockquote><blockquote>Predicts an outcome</blockquote><p>See below for a format for setting a concrete hypothesis.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*PyHZ1j9e3EDa32r06gmhxg.png" /><figcaption><a href="http://cxl.com/">Image</a>: Here’s how to set a proper hypothesis for your A/B test.</figcaption></figure><p>Here’s how this format has been used to generate a hypothesis for addressing self-efficacy among users.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*1rUuhqjbPkEgexrDz6QspQ.png" /><figcaption><a href="http://cxl.com/">Image</a>: Example of a concrete hypothesis.</figcaption></figure><p><strong>Prioritize your A/B tests using PIPE framework</strong></p><p>There are popular frameworks like PIE that stands for Potential, Importance, Ease or ICE that is an acronym for Impact, Confidence, Effort. However, Ton Wesseling recommends not falling trap to use these two frameworks as they lack Power. Therefore, he recommends the PIPE framework as shown below:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1021/1*RhhmwUSr3pxv0HyftOS6Pw.png" /><figcaption><a href="http://cxl.com/">Image</a>: PIPE model includes POWER that the PIE model lacks.</figcaption></figure><p>Here’s an example of how PIPE prioritization looks like:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*rfDxyNKab6cMvCH7qzgIwA.png" /><figcaption><a href="http://cxl.com/">Image</a>: An example of PIPE prioritization.</figcaption></figure><h3><strong>How to execute an A/B test?</strong></h3><p>Now that you are done with all the planning and research, let’s move towards designing, developing and quality assuring your A/B tests.</p><p>Salient points to keep in mind:</p><ul><li>Have one challenger only</li><li>You can do more than one change provided they’re aligned to your hypothesis and your budget permits them</li><li>Consider minimum detectable effect (MDE)</li><li>Don’t use the WYSIWYG code editor</li><li>Consider injecting client side code in the control also unless server-side coding</li><li>Consider QA of browsers and devices</li><li>Calculate the length of your A/B test using the free calculator (already mentioned in this article)</li><li>Monitor your experiment and stop if anything is broken</li><li>Check for Simple Ratio Mismatch (SRM) and stop test immediately</li><li>In case your company is losing money through this experiment then stop the test</li></ul><h3><strong>How to make sense of the result?</strong></h3><p>There can be two possible outcomes of your A/B test — Either no winner (inconclusive) or Winner. Before we move on to the result here’s few things to keep in mind:</p><ul><li>Analyze the result in your site’s Google Analytics or any other analytics tool rather than the test tool</li><li>Avoid sampling</li><li>Analyze users and not sessions</li><li>Analyze users who have converted instead of users and total conversions</li></ul><p>Use this <a href="https://abtestguide.com/calc/">free A/B test result calculator</a>.</p><p><strong>You get a winner</strong></p><ul><li>Implement it as soon as possible</li><li>Dive in segments to understand who caused the effect and use this insight for writing a new hypothesis for your next experiment.</li></ul><p><strong>No winner</strong></p><ul><li>You can still implement the result</li><li>No need to dig in segments to find winners</li></ul><p>So, that’s all for now. This blog is just an introduction to the A/B Testing Mastery course at the CXL Institute. I would recommend enrolling for the course to get a thorough knowledge about A/B testing.</p><p>Please feel free to drop your comments below. See you in the next blog. Until then stay tuned and stay safe!</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=118e7c253903" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Identifying CRO Opportunities and Prioritizing A/B Testing Ideas — Review]]></title>
            <link>https://bhavsar-poli.medium.com/identifying-cro-opportunities-and-prioritizing-a-b-testing-ideas-review-4de49c05217?source=rss-28d4f7b81521------2</link>
            <guid isPermaLink="false">https://medium.com/p/4de49c05217</guid>
            <category><![CDATA[google-analytics]]></category>
            <category><![CDATA[heuristic-evaluation]]></category>
            <category><![CDATA[conversion-optimization]]></category>
            <category><![CDATA[user-research]]></category>
            <category><![CDATA[a-b-testing]]></category>
            <dc:creator><![CDATA[Poli Dey Bhavsar]]></dc:creator>
            <pubDate>Sun, 25 Apr 2021 18:23:05 GMT</pubDate>
            <atom:updated>2021-04-25T20:30:33.488Z</atom:updated>
            <content:encoded><![CDATA[<h3><strong>Identifying CRO Opportunities and Prioritizing A/B Testing Ideas — Review</strong></h3><p>Are you an optimizer wondering “what should I test?” Well, you’re not alone in this. Many people begin with this question when it comes to conversion optimization. Good you landed on this page because you won’t ask this wrong question again and as this blog would tell you exactly how to begin with your optimization efforts.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Kewz0xLvwocuc8INzphzmA.jpeg" /><figcaption><a href="https://unsplash.com/photos/8xAA0f9yQnE">Image</a>: Pondering, “What Should I Test?”. Not anymore with <a href="https://cxl.com/institute/programs/conversion-optimization/">CRO </a>minidegree at CXL</figcaption></figure><p>In this blog, I am going to share (on a high level) my learnings from the <a href="https://cxl.com/institute/programs/conversion-optimization/">Conversion Optimization minidegree at the CXL Institute</a> about how to identify conversion optimization opportunities, prioritize them and then do A/B testing.</p><p>So, there are basically three parts for us to discuss here, research, analysis and testing. Research can help you form a strong foundation for your experimentation. As <a href="https://www.linkedin.com/in/peeplaja/">Peep Laja</a> aptly says,</p><blockquote>“CRO is 80% research and 20% experimentation. You need conversion research to get sustainable results over time. Learn this or fail.”</blockquote><h3><strong>Collect data &amp; stop guess work</strong></h3><p>Therefore, it’s time you stopped guessing and started digging into your <a href="https://bhavsar-poli.medium.com/google-analytics-101-discover-the-how-of-your-website-results-review-e8d8887eb5ea">Google Analytics</a>, <a href="https://bhavsar-poli.medium.com/google-tag-manager-101-get-familiar-with-the-control-centre-for-your-online-data-review-30da0aa456ce">Google Tag Manager</a> to identify areas on your site that are leaking conversions. In case your GA account is not properly set or you’re not sure if you’re getting accurate data from it then I would recommend you to do yourself a favour and enrol in CXL’s <a href="https://cxl.com/institute/online-course/google-analytics-audit/">The Google Analytics Audit</a> course by Fred Pike, Managing Director at Northwoods. This can be complemented with a <a href="https://bhavsar-poli.medium.com/introduction-to-user-research-and-personas-driven-by-fast-rigorous-framework-review-b6e5410de585">qualitative user research</a> to discover the areas in your site that you can improve.</p><h3><strong>Analyze your CRO opportunities</strong></h3><p>Now, let’s move on to the analysis part. Here, I would like to share with you my learnings from the course <a href="https://cxl.com/institute/online-course/conversion-optimization-audits/">Heuristic Analysis Frameworks for Conversion Optimization Audits</a> by Andre Morys, CEO and Founder at KonversionsKRAFT.</p><p>In this course, you would learn about how to conduct proper heuristic evaluation in less than an hour. And you can enhance the benefits of this evaluation with <a href="https://bhavsar-poli.medium.com/introduction-to-user-research-and-personas-driven-by-fast-rigorous-framework-review-b6e5410de585">persona development</a>. Wondering how would personas facilitate heuristic evaluation? Well, they’ll make it easier for you to put yourself in your ideal customers’ shoes and see the website (subject to analysis) through the customers’ eyes.</p><p><strong>About the framework</strong></p><p>This 7-level framework by Andre Morys when compared to existing models not just tells you the why behind the data but also helps you focus on subconscious mechanisms. Yes, “data you measure on your website is the result of user behaviour,” says Andre Morys. “So, if you understand user behaviour, then you can understand the data.”</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*uJ_kPNTCZWsy5Ye3k8uvHA.png" /><figcaption><a href="http://cxl.com/">Image</a>: Heuristic evaluation framework by Andre Morys</figcaption></figure><p><strong>Cognitive biases</strong></p><p>He explains many cognitive biases that if we understand them it would be easier for us to understand why users behave the way they do. Some of the cognitive biases you learn in this course are confirmation bias, decoy effect, contrast effect, social desirability bias and more. Find the exhaustive list of cognitive biases <a href="https://en.wikipedia.org/wiki/List_of_cognitive_biases">here</a>.</p><p><strong>The ROI pyramid</strong></p><p>He also shows us the underlying mechanism that makes a difference between mediocre and great products.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/904/1*bykvYhytmNgolZ6fRE7L5g.png" /><figcaption><a href="http://cxl.com/">Image</a>: Increase the ROI of your website by focusing on user motivation.</figcaption></figure><p>The must-haves in the image are the hygiene factors that people expect in your product or website. They are not fascinated by them. They’re supposed to be there anyway.</p><p>So, as a product owner if you want to create great products that users would feel motivated to buy then you must focus on user motivations instead of the hygiene factors or the must-haves.</p><p><strong>Competitive benchmarking</strong></p><p>One of the best things that you learn from the course is how to get buy-in from the C-suite executives in your company. You can use the framework to analyze your competitors website also along with yours.</p><p>Then show your findings in a form of graph such as why they are better than you and how you can also reach there. Even, instead of pointing out designers or developers, etc. what they’ve not done right, you can simply show them how competitors are doing it better.</p><p><strong>Bring more credibility to your evaluation</strong></p><p>You can work with different raters, say in a team of three raters. Thus, you can raise the validity of an expert evaluation.</p><p>Now, from this evaluation as soon as you have many hypotheses, start prioritizing them.</p><p>Next, let’s move on to testing.</p><h3><strong>Why testing?</strong></h3><p>“Testing is for validation and learning”. So that we no longer rely on gut feeling or intuition.</p><p>A significant fact about A/B testing is that you don’t know why you’ve a winning test. So, don’t fall in the trap of your conscious mind or rational mind to create a story behind the success of your winning test.</p><p>You can use the PXL framework by CXL as it offers the most objective way of prioritization. For A/B testing statistics you can use the A/B testing calculator by CXL. This would help you not fall prey to myths about sample size, statistical significance, etc.</p><p><strong>2 Important criteria to keep in mind while giving a score:</strong></p><ol><li>Ease of implementation</li><li>Opportunity of getting a big lift.</li></ol><p>In the minidegree under the Testing track there is a comprehensive lesson on “How to Run Tests?” I would suggest you enrol for it so that you never ever get stuck with the question, “What Should I Test”.</p><p>Now, let’s put a quick glance on testing strategies!</p><p>Peep Laja gives a very straightforward answer to the eternal question “What to test?”</p><p>“Identify problem not ideas”</p><p>So, he says,</p><ul><li>Perform conversion research to identify problems</li><li>Determine urgent user problems</li><li>Test possible solutions</li></ul><p>Next he answers your other question in the course, “How many changes per test?”</p><p><strong>Strategy #1 — Making many changes.</strong> Peep recommends one thing at a time unless you have millions of visitors. <strong>Just make sure each change that you’re making is directly connected to an identified problem.</strong></p><p><strong>Strategy #2 — Making many changes.</strong> Suppose your hypothesis says “Clarity is obstructing optimization” then you make say 75 changes across the website that would improve clarity. So, <strong>each change on your website supports the same hypothesis</strong>.</p><p>Let’s have a look at a few types of testing.</p><p><strong>A/B testing vs. Multivariate testing</strong></p><p>In a nutshell, A/B testing is ideal for testing dramatic changes whereas multivariate is good for testing interaction effects provided you have more that 100,000 visitors per month.</p><p><strong>Iterative testing</strong></p><p>You can do iterative testing when you want to attribute a specific change to a result. It’s easy to start as it’s low budget. It can help you build momentum for a A/B testing program. Moreover, you can test any random idea, suppose one that comes from the C-suite and you can prove that it might not work all the time by testing. As it’s low budget you can build the test multiple times and learn from hits and misses of each iteration.</p><p><strong>Innovative testing</strong></p><p>When iterative testing is not working, you make larger changes (more likely to have an impact on user behavior) on a portion of your website to move the needle, i.e., have an impact on the conversion rate.</p><p>I have just shared my understanding from the courses; please don’t judge them by this blog. They’re worth pursuing for you as marketer who wants to be among the top 1%. So, don’t miss out. See you in my next blog. Till then stay safe and stay tuned!</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=4de49c05217" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Introduction to User Research, and Personas Driven by Fast & Rigorous Framework — Review]]></title>
            <link>https://bhavsar-poli.medium.com/introduction-to-user-research-and-personas-driven-by-fast-rigorous-framework-review-b6e5410de585?source=rss-28d4f7b81521------2</link>
            <guid isPermaLink="false">https://medium.com/p/b6e5410de585</guid>
            <category><![CDATA[quantitative-research]]></category>
            <category><![CDATA[user-personas]]></category>
            <category><![CDATA[qualitative-research]]></category>
            <category><![CDATA[user-research]]></category>
            <dc:creator><![CDATA[Poli Dey Bhavsar]]></dc:creator>
            <pubDate>Sun, 18 Apr 2021 18:23:13 GMT</pubDate>
            <atom:updated>2021-04-19T13:32:43.059Z</atom:updated>
            <content:encoded><![CDATA[<h3><strong>Introduction to User Research, and Personas Driven by Fast &amp; Rigorous Framework — Review</strong></h3><p>Ever wondered why some products are loved so much by users while some are doomed? The simple answer is user-centric product development begins with user research. A friend in need is a friend indeed — don’t you love your best friend?</p><p>Similarly, when you put efforts to understand your users, their needs and pain points, you’re more likely to build a product that your users find usable and fall in love with your brand.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*4e43PgpgzveKmCqDoLku-Q.jpeg" /><figcaption><a href="https://unsplash.com/@epicantus?utm_source=unsplash&amp;utm_medium=referral&amp;utm_content=creditCopyText">Image</a>: Organizing research data using sticky notes to identify patterns &amp; support insights</figcaption></figure><p>In this blog I am going to share about user research, its types and approaches. We would also have a look at personas driven by fast &amp; rigorous framework. Let’s begin with user research.</p><h3>What is user research?</h3><p>User research is the methodical study of users in order to understand their goals, needs, and beliefs. “We can use this insight to create better experiences and deliver greater business value”, says Megan Kierstead in her course <a href="https://cxl.com/institute/online-course/user-research/">User Research at the CXL Institute.</a></p><p>One of the crucial things that she teaches in the course is that your research must begin with a goal, which in turn would decide what methods you would choose to fulfil the purpose of doing the research.</p><h3><strong>Types of research</strong></h3><p>User research is broadly divided into two types:</p><ul><li>Quantitative research</li><li>Qualitative research</li></ul><p><strong>Qualitative research</strong></p><p>Qualitative research helps you discover the <strong>why</strong> behind your user behaviour and helps you build a deep understanding of why they do what they do. It requires great care as it involves non-numerical data such as user quotes or opinions collected via open-ended questions. You have to take care that you do not influence your users with your opinions.</p><p>Examples of qualitative research methods are interviews, field studies, user testing, etc.</p><p><strong>Quantitative research</strong></p><p>Quantitative data helps you know <strong>what </strong>users do and also test the assumptions formed from qualitative research. This method involves more structured methods such as surveys, polls, etc. to collect measurable data which you can analyze to identify patterns. However, you cannot find deep insights from quantitative research as done via qualitative methods.</p><h3><strong>User research can be split into two approaches:</strong></h3><ul><li><strong>Behavioural</strong> — In this approach, you watch your users’ actions through observational studies such as field research or contextual enquiry, user testing, and more. Nothing can be more valuable than watching your users in their natural environment actually doing their work. However, it’s tough because at times it could be very invasive to follow around your users for a longer period of time.</li><li><strong>Attitudinal </strong>— In this approach you listen to what your users say, e.g., in interviews or surveys. For interviews, Megan recommends not to use scripts. She suggests you let the conversation flow; just ask follow-up questions. Remember the purpose of the interview is to uncover user needs, beliefs. If you stick to topics then you’re focusing on your beliefs, and on your company’s needs.</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*6iyzShvy3Y0GBO_IYDsylA.jpeg" /><figcaption><a href="https://unsplash.com/@wocintechchat">Image</a>: User interviews are one of the best ways to listen what your users have to say.</figcaption></figure><p>For your research to be fruitful, i.e. to get meaningful data out of it, you must have access to the right users. Where to find them? Well, in the User Research course, Megan Kierstead tells you where to look for your users and recruit them. Now, let’s move on to making sense of the data that you have collected.</p><h3><strong>How to derive insights from your data?</strong></h3><p>Now that you have tons of data in the form of inputs, videos, transcripts, etc., how are you going to convert them into something useful that would inform your decisions.</p><p>Megan Kierstead said a very important thing that kind of stuck with me. She said that data-driven decision is a misnomer; it’s in fact insight-driven decision. Because raw data doesn’t guide and is of no use; couldn’t agree more!</p><p>Now categorize and label your data using a synthesis framework. What is a synthesis framework? Megan explains:</p><blockquote>“Structured or semi-structured way to look for patterns in your data.</blockquote><blockquote>Patterns are the direct precursor of insights .”</blockquote><p>You can combine different types of synthesis frameworks (listed below) to generate robust insights:</p><ul><li>Affinity diagramming</li><li>Coding</li><li>2x2s</li><li>Word clouds</li></ul><h3><strong>Creating memorable and usable research artifacts</strong></h3><p>Now that you have synthesized insights from your data, it’s time to create the research outcome useful for your whole organization. You can create personas so that people in your organization who don’t get to interact directly with your users would also get an awareness about them. It would generate a shared understanding and language among them and help them build empathy for your users.</p><p>So, that was just a bird view for a very comprehensive topic User Research.</p><p>Do you have personas in your company? I mean research-driven ones? If you’re avoiding it because you don’t want to put the hard work in then please know that something is better than nothing!</p><p>Do you know that you can create personas without spending much time? Instead of the cooked up personas which are utter useless for your business, you can craft insight-driven personas really fast. How? Well, I am going to tell you next.</p><h3><strong>User Personas Driven by Fast and Rigorous Framework</strong></h3><figure><img alt="The image shows different types of people with different body shapes and features." src="https://cdn-images-1.medium.com/max/1024/1*BDWE48x3KVTIdaPzmxkYng.jpeg" /><figcaption><a href="https://unsplash.com/@gemmachuatran?utm_source=unsplash&amp;utm_medium=referral&amp;utm_content=creditCopyText">Image</a>: Your persona is descriptive “model” of your users.</figcaption></figure><p>Are you familiar with user personas? If yes, then you understand right, how difficult it is to create them. How time consuming they can be and if you compromise with time then quality would suffer. What if I tell you that you can craft personas without compromising any of these two — quality vs. speed?</p><p>In the Fast and rigorous persona course at the CXL Institute, <a href="https://www.linkedin.com/in/stefania-mereu-01402531/">Stefania Mereu</a> and <a href="https://www.linkedin.com/in/eric-taylor-b0057115/">Eric Taylor</a> share their <strong>Fast &amp; Rigorous</strong> framework that defies two traditional variables at play — quality and speed.</p><p>This data-driven framework consists of <strong>three</strong> steps:</p><ol><li>Collect data using quantitative methods</li><li>Identify groups using statistical clustering</li><li>Build archetypes or user personas</li></ol><p><strong>What is a user persona?</strong></p><p>Let’s have a look at different definitions of user persona to understand it better.</p><p>As per the definition of the instructors, Stefania and Eric:</p><blockquote>“A user persona is a model which gives an approximate description of the customer and provides insight into their motivation and goals. A well-constructed persona will provide focus for development projects, create empathy for product designers, give perspective to stakeholders, and add context to design choices.”</blockquote><p>According to Dr. Davis Travis, a UX veteran with over 30 years of experience and currently a User Experience Strategist at Userfocus:</p><blockquote>“A persona is a detailed description of an imaginary person that provides a shared understanding of the users of a system.</blockquote><blockquote>A persona is a <strong>user archetype </strong>— not an “average” user or a real user.”</blockquote><p>Also let’s have a look at the original definition of user persona by Alan Cooper:</p><blockquote>“A precise descriptive <strong>model</strong> of the user, what he wishes to accomplish and <strong>why</strong>”</blockquote><p>Eric Taylor elaborates the meaning of Alan’s definition by picking the words “model” and “why”</p><p>Here’s how simplifies it for you:</p><p><strong>Model </strong>— An approximate description of the user</p><p><strong>Why </strong>— Describes the users’ motivation to buy</p><p>Now let’s move on to Eric and Stefania’s approach to creating persona:</p><p><strong>Step 1: Collecting quantitative data</strong></p><p>They recommend that the questions must be <strong>relevant</strong>, <strong>actionable</strong> and <strong>unbiased</strong> in order to be diagnostic. They exemplify these attributes with the following questions for a car shopping survey:</p><blockquote>Irrelevant: How has been your car driving experience been so far?</blockquote><blockquote><strong>Relevant: How has been your car shopping experience been so far</strong></blockquote><blockquote>Unactionable: What is the MPG on your current car?</blockquote><blockquote><strong>Actionable: What is your desired MPG on your future car?</strong></blockquote><blockquote>Biased: Would you rather buy a car from a dealer or from a stranger?</blockquote><blockquote><strong>Unbiased: Would you buy a car from a salesman or from a private seller?</strong></blockquote><p>You can also lump questions in a survey with proper flow and rank them using plain English such as “not at all important”, “important” instead of likert scale numbered 1–5 as per Eric. You can set up the survey either using paid tools such as Qualtrix, SurveyMonkey or free tools such as Google forms.</p><p>“However, <strong>remember to include open-ended questions</strong> to help with building archetypes down the line,” says Stefania.</p><p>Stefania also demonstrates how you can recruit respondents using tools such as Hotjar or Amazon Mechanical Turk.</p><p>To be sure that you are doing the right things in a right manner, you can do a <strong>pilot survey</strong> first, i.e., <strong>test a small sample</strong> to identify mistakes.</p><p>Once you have the answers to your questions, you must simplify your data with <strong>Exploratory Factor Analyis (EFA)</strong>, i.e., discover the underlying factors reflected in the answers to the questions. One thing that you should know is you must ask similar questions or redundant questions in order to get a stronger measure of underlying factors.</p><p><strong>Factor analysis does these 3 things to simplify your data:</strong></p><ul><li>Identifies questions that are grouped around one factor</li><li>Identifies relationships between theoretical factors</li><li>Assess the weight of importance of each factor by respondent</li></ul><p><strong>Step 2: Identify groups using statistical clustering</strong></p><p>Once your data is simplified you can identify clusters and put your respondents into clusters. You can use tools like IBM SPSS, or you can use the programming language R to analyze the data.</p><p><strong>Step 3: Build archetypes or user personas</strong></p><p>At this point you only know that there is a group of people that has a specific pattern of needs and interests. Later when you add the description of the group, you come to know about the demographics of the group. At this stage, try to limit irrelevant data about the users to avoid fluff in your personas. Also add a picture of your fictitious user that reflects some of the demographics that you added just now, so that you can identify with your users and empathise with them.</p><p>Your user persona is ready now! How long can you refer to them? Not forever, as both people and market change. Also, you would need to keep track of any changes your business is making in terms of its services or value proposition. Remember to iterate your personas to match with them.</p><p>If you like this blog then please clap so that it reaches more readers. You can also leave your comments below and start a conversation. I would love to hear from you. Stay tuned and stay safe!</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=b6e5410de585" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Google Tag Manager 101: Get Familiar with the “Control Centre for Your Online Data” — Review]]></title>
            <link>https://bhavsar-poli.medium.com/google-tag-manager-101-get-familiar-with-the-control-centre-for-your-online-data-review-30da0aa456ce?source=rss-28d4f7b81521------2</link>
            <guid isPermaLink="false">https://medium.com/p/30da0aa456ce</guid>
            <category><![CDATA[tag]]></category>
            <category><![CDATA[trigger]]></category>
            <category><![CDATA[datalayer]]></category>
            <category><![CDATA[variables]]></category>
            <category><![CDATA[google-tag-manager]]></category>
            <dc:creator><![CDATA[Poli Dey Bhavsar]]></dc:creator>
            <pubDate>Sun, 11 Apr 2021 18:18:03 GMT</pubDate>
            <atom:updated>2021-04-11T19:06:13.735Z</atom:updated>
            <content:encoded><![CDATA[<h3><strong>Google Tag Manager 101: Get Familiar with “Control Centre for Your Online Data” — Review</strong></h3><p>Are you a digital marketer who wants to derive more value out of your online data measurement and analysis? Do you want to deploy third-party tracking scripts on your site without relying on a developer? Then it’s time to skill up yourself with the <a href="https://cxl.com/institute/online-course/google-tag-manager/">Google Tag Manager for Beginners</a> course at the CXL Institute.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/734/1*Lp_GsqCV-de-Oi7eFVviwA.png" /><figcaption><a href="https://marketingplatform.google.com/about/tag-manager/benefits/">Image</a>: Google Tag Manager — Your own control centre to get visibility of your users’ behaviour.</figcaption></figure><p>Yes, you guessed it right! Today I am going to share with you my learnings from this course in this blog (8th in the knowledge sharing series) but at a very high level.</p><p>In this blog I would cover the following points:</p><ul><li>Brief introduction to GTM</li><li>What is a tag? Types of Tags</li><li>What are triggers? Different types of triggers</li><li>What is a variable? Two types of variables</li><li>What is a data layer</li><li>6 Noteworthy tips to keep your GTM Organized</li><li>What is Preview mode and why do you need it?</li><li>What is a workflow?</li></ul><p>This blog is limited to giving you an introduction to GTM; so you won’t get a chance to roll your sleeves and get started with the tool.</p><p>To get hands on lessons and master the fundamentals of GTM, I would recommend you to enrol for this course by <a href="https://www.linkedin.com/in/chrismerceronline/">Chris Mercer</a>, Co-Founder of MeasurementMarketing.io and discover the benefits of learning from an expert.</p><h3><strong>Brief introduction to GTM</strong></h3><p>Tag Management Systems is like a control panel that helps you get visibility into different behaviours that are happening on your site. It not just collects data but it also takes various actions based on those behaviours as the customer journeys through your website. And the cool part is that you can sort of train them to do so.</p><p>Google Tag Manager collects information as it comes through and passes on to various platforms like Google Analytics or Facebook.</p><p>A point worth noting here is that the biggest exception to what you can’t use Tag Manager for is split testing scripts. So, if you’re using Visual Website Optimizer (VWO) then make sure to put its script above Tag manager’s script in the header of your site.</p><h3><strong>What is a tag?</strong></h3><p>A tag, basically, is a script that goes on the page and it tells Google Tag Manager <strong>what</strong> to do. You get the script put on your pages via GTM and these tags would get the information if any user is on your site. And when a user converts, tags would instruct GTM to inform the various platforms like Facebook, Hotjar, Google Analytics, etc. whose scripts are on your website pages.</p><p><strong>Types of Tags</strong></p><p>There are two types of tags:</p><ol><li><strong>Built-in</strong> — As the name suggests, you have Google Analytics, Google Ads, Floodlights, Google Optimize and many more built-in tags available</li><li><strong>Custom HTML</strong> — When you can’t find a built-in tag that fits your needs, let’s say, you don’t have a built-in option for Facebook, GTM is flexible enough to allow you to use Facebook. using custom HTML.</li></ol><h3><strong>What are triggers?</strong></h3><p>Triggers tell GTM <strong>when</strong> to fire the tags, while there are different user behaviours happening on your site, in order to enhance your own reporting capabilities.</p><p>Here are different types of triggers that you would find on GTM:</p><ul><li><a href="https://support.google.com/tagmanager/answer/7679319?hl=en&amp;ref_topic=7679108">Page view trigger</a></li><li><a href="https://support.google.com/tagmanager/answer/7679320?hl=en&amp;ref_topic=7679108">Click trigger</a></li><li><a href="https://support.google.com/tagmanager/answer/7679410?hl=en&amp;ref_topic=7679108">Element visibility trigger</a></li><li><a href="https://support.google.com/tagmanager/answer/7679217?hl=en&amp;ref_topic=7679108">Form submission trigger</a></li><li><a href="https://support.google.com/tagmanager/answer/7679322?hl=en&amp;ref_topic=7679108">History change trigger</a></li><li><a href="https://support.google.com/tagmanager/answer/7679411?hl=en&amp;ref_topic=7679108">JavaScript error trigger</a></li><li><a href="https://support.google.com/tagmanager/answer/7679218?hl=en&amp;ref_topic=7679108">Scroll Depth trigger</a></li><li><a href="https://support.google.com/tagmanager/answer/7679323?hl=en&amp;ref_topic=7679108">Timer trigger</a></li><li><a href="https://support.google.com/tagmanager/answer/7679325?hl=en&amp;ref_topic=7679108">YouTube video trigger</a></li><li><a href="https://support.google.com/tagmanager/answer/7679219?hl=en&amp;ref_topic=7679108">Custom event trigger</a></li><li><a href="https://support.google.com/tagmanager/answer/9164222?hl=en&amp;ref_topic=7679108">Trigger group</a></li></ul><h3><strong>What is a variable?</strong></h3><p>A variable is the information GTM needs in order to do its job. And variables can have different values as user behaviour changes and GTM would pass on these respective values to the respective platforms associated with those variables.</p><p>There are basically two types of variable:</p><ol><li><strong>Built-in variables</strong> — The variables that you would find built-in GTM.</li><li><strong>User-defined variables</strong> — The variables that you would set on your own.</li></ol><h3><strong>What is a data layer?</strong></h3><p>Data layer is a place to temporarily store details or values important to GTM in the client that can be used by tags, triggers and variables. There are two parts to a data layer — <strong>keys</strong> and <strong>values</strong>.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/798/1*4UmOslau2MyeN8frlZ3FsA.png" /><figcaption>Image: Screenshot showing Data Layer depicting keys and values.</figcaption></figure><p>As shown in the above image, UniqueEventID, scrollThreshold, ScrollUnits are some of the keys whereas 5, 25, percent, vertical are their respective values</p><h3><strong>6 Noteworthy tips to keep your GTM Organized</strong></h3><ol><li>Check accounts and user roles, that is share the right access with the right people by controlling permissions so that work is done safely and effectively.</li><li>Also check container permission levels which one is appropriate for which type of users — Read, Edit, Approve, Publish.</li><li>As a company, you should always follow a naming convention to avoid the possibility of duplicating things. Because one person won’t realize that the other person has already done it.</li><li>Create folders to stay organized in Google Tag Manager.</li><li>Remember, less is more in GTM, i.e., if you could use one tag to something versus seven tags to that same thing, then try to do it in just one tag. The same holds good for triggers.</li><li>Be aware of at least some basics of RegEx as it will help you with less is more discussed in the previous point.</li></ol><h3><strong>What is Preview mode and why do you need it?</strong></h3><p>Preview is a way for you to check and test all the changes that you have implemented in your container in the tag manager. Whatever you have just updated, you can test it with preview and make sure that it’s working before pushing them live. You can see and test the changes before the rest of the world sees your changes.</p><h3><strong>What is a workflow?</strong></h3><p>Well, in GTM the workflow is like Workspace -&gt; Publishing -&gt; Versions.</p><p><strong>Workspace</strong> — By default you get a Workspace in GTM known as Default workspace. If you are a single user then you can work in this workspace itself.</p><p>However, if there are multiple users being in the GTM container at the same time, then it is quite likely that one person might overwrite the changes of another person.</p><p>Having separate workspaces help to keep those changes separate and siloed and make sure that changes are not overwritten.</p><p><strong>Publishing</strong> — Once you have made changes in your workspace, you preview and test them first and you submit them to publish the actual changes. Once you have published them, you end up with a version.</p><p><strong>Version</strong> — Now you can see your live version that has been published. Please note that you can also save the current state of your workspace as a version so that you can revert back to a previous version anytime, if necessary.</p><p>GTM is an essential skill for any technical digital marketer or analyst. Here’s what Chris Mercer has to say to you,</p><p>“If you want to be at the top of your game, you have to know Google Tag Manager. It’s no longer an optional tool…”</p><p>So, what are you waiting for! Join this course to supercharge your analytics skill with Chris Mercer and get more out of the data that you are currently working with.</p><p>See you next week in my next blog, till then please stay tuned and please feel free to let me know your views on this blog.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=30da0aa456ce" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Google Analytics 4 is Here to Empower You; Are you Ready Marketers? — Review]]></title>
            <link>https://bhavsar-poli.medium.com/google-analytics-4-is-here-to-empower-you-are-you-ready-marketers-review-6e3e4222da5a?source=rss-28d4f7b81521------2</link>
            <guid isPermaLink="false">https://medium.com/p/6e3e4222da5a</guid>
            <category><![CDATA[universa-analytics]]></category>
            <category><![CDATA[google-analytics-4]]></category>
            <category><![CDATA[conversion-optimization]]></category>
            <category><![CDATA[digital-marketing]]></category>
            <category><![CDATA[tracking-code]]></category>
            <dc:creator><![CDATA[Poli Dey Bhavsar]]></dc:creator>
            <pubDate>Sun, 04 Apr 2021 18:21:46 GMT</pubDate>
            <atom:updated>2021-04-04T19:36:51.811Z</atom:updated>
            <content:encoded><![CDATA[<h3><strong>Google Analytics 4 is Here to Empower You; Are you Ready Marketers? — Review</strong></h3><p>Have you been using Google Analytics for the last 10–15 years? Then, over the years you must have been screaming out loud to Google for adding some really important features to Universal Analytics. With GA4 (formerly known as “App + Web”), you can finally breathe a sigh of relief; Google has addressed your grievances to a greater extent.</p><p>GA4 is the next generation of Google Analytics that empowers digital marketers by bringing powerful weapons in their arsenal in the form of enterprise features that were previously only available in Google Analytics 360 such as BigQuery integration, etc.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*17dT8ifFaFyG87xoBQoIxg.jpeg" /><figcaption><a href="https://unsplash.com/@benwhitephotography?utm_source=unsplash&amp;utm_medium=referral&amp;utm_content=creditCopyText">Image</a>: Google Analytics 4 — the future of analytics is here to empower marketers</figcaption></figure><p>To gain an in-depth knowledge about this new and improved Google Analytics platform, I would recommend you to enroll for the <a href="https://cxl.com/institute/online-course/google-analytics-4/">Google Analytics 4 course at tbe CXL Institute</a>. The instructor <a href="https://www.linkedin.com/in/charlesfarina/">Charles Farina</a> is highly experienced with the Google Marketing Platform, speaks at International conferences and has provided training to many Fortune 100 companies.</p><p>In this blog (7th blog in the knowledge sharing series) I would share my learnings from the aforementioned course but at a very high-level. I’ll cover the following points:</p><ul><li>GA4 — All about unification of web and apps</li><li>Why is this new next generation Google Analytics called GA4?</li><li>What are the main differences between UA and GA4?</li><li>9 Key takeaways from the ‘How to find conversion opportunities using analytics’ course</li></ul><h3><strong>GA4 — All about unification of web and apps</strong></h3><p>To have an understanding of GA4, let’s put a quick glance at how Google’s measurement platforms work. The current version of Google Analytics known as Universal Analytics is for web and Google Analytics for Firebase is for mobile apps.</p><p>Google unifies web and app capabilities so that you as a marketer can have a unified view of user behaviour across devices while considering user privacy at a granular level.</p><h3><strong>Why is this new next generation Google Analytics called GA4?</strong></h3><p>Wondering about the naming convention by Google? Here’s a simple explanation:</p><p>Urchin — GA 1</p><p>Classic Analytics — GA 2</p><p>Universal Analytics — GA3</p><p>New Next Gen Analytics — GA4</p><h3><strong>What are the main differences between UA and GA4?</strong></h3><p>GA4 maximizes the benefits of three technologies Google has been working on and developing in the recent past:</p><ul><li>Firebase Analytics</li><li>Google Signals</li><li>Google Site Tag</li></ul><p>In fact, GA4 uses the same data-model as Firebase Analytics; if you upgrade a Firebase Analytics project to GA4, it carries over the historical data. The UI of GA4 is completely new and re-imagined.</p><p>Now, let’s have a look at some of the important distinctive features of UA and GA4:</p><p><strong>Data Model</strong> — Universal Analytics has used a hit based data model, i.e., users are tracked via sessions whereas GA4 uses an event-driven data model. Now, you can get to know what users are actually doing on your site rather than just knowing you have a user on your site.</p><p><strong>Automatic Measurement </strong>— In UA you get only standard page view measurement out of the box, whereas GA4 automatically measures interactions and content on your site such as scrolls, outbound links, site search video engagement and file downloads in addition to page views.</p><figure><img alt="Screenshot showing Enhanced Measurement in GA4" src="https://cdn-images-1.medium.com/max/1024/1*AcqV6tJQrWVhBk-276A9qg.png" /><figcaption><a href="http://cxl.com/">Image</a>: GA4 offers Enhanced Measurement out of the box.</figcaption></figure><p><strong>Reporting</strong> — In UA there are many set reports which allow you to customize.GA4, on the other hand, comes with many default top-notch reports and you can go to the <strong>analysis hub</strong> for specific reports.</p><p>Even the realtime report in GA4 is more interactive and with new user properties or user scope dimensions. Funnel capabilities and a BigQuery export that used to exist only in the enterprise or 360 version of GA, are now available in the basic version of GA4.</p><p><strong>Tracking Code</strong> — There is a clear distinction between UA and GA4 setup. While UA uses a property and view setup, GA4 consolidates your website and app data via Google Analytics’ data streams. Therefore, GA4 stream-id appears like GA-XXXXXXX instead of property-id of UA-XXXXXX.</p><figure><img alt="UA vs. GA4-The basic difference between tracking code" src="https://cdn-images-1.medium.com/max/1024/1*_axcn0dd3bEfke94CogQPQ.png" /><figcaption><a href="http://cxl.com/">Image</a>: UA vs. GA4 — The basic different in setup and tracking ID.</figcaption></figure><p>Eager to roll over your sleeves and implement GA4 on your site? Here’s a <a href="https://support.google.com/analytics/answer/9304153?hl=en&amp;ref_topic=9303319">set up guide</a> for you to refer and get it done within fifteen minutes. As GA4 is relatively new, the best practice is to implement it parallel to UA (dual-tag).</p><p>You can also <a href="https://www.charlesfarina.com/new-bigquery-integration-for-app-web-properties/">setup BigQuery on your site</a>, now that this enterprise grade feature comes for free with GA4.</p><p>Now let me share with you how to identify conversion opportunities with Google Analytics — my learnings from the course Using Analytics to Find Conversion Opportunities by <a href="https://www.linkedin.com/in/jeffsauer/">Jeff Sauer</a> in the <a href="https://cxl.com/institute/programs/conversion-optimization/">CRO Program by the CXL Institute</a>.</p><p><strong>9 Key takeaways from the ‘How to find conversion opportunities with analytics’ Course</strong></p><p>If you have not yet embraced GA4, you can still find conversion opportunities with Universal Analytics:</p><ol><li>Reports that are insightful while looking for conversion opportunities:</li></ol><ul><li>Audience overview — You can get an overall view of who has been coming to your page.</li><li>Behaviour reports — You can see which pages on your site your users visited</li></ul><p>2. Have a look at the Overview report to identify the quality and grouping of your traffic, especially the channels by which people are getting to your website. Use this information to optimize your marketing strategies.</p><p>3. Bounce rate is not always a bad sign although in many cases it’s a negative. It often indicates that you haven’t made clear the next step you want your users to take.</p><p>4. Universal Analytics out of the box doesn’t tell you what’s happening on your page. Bounce rate is often a misleading metric as it doesn’t tell you what action people are taking on your page.</p><p>5. With vanilla UA code you won’t be able to track user interaction on your page; therefore you have to create events — (i) Interaction-based or (ii) Non-interaction based. You can use interaction-based events to reduce bounce rates.</p><p>6. You can get more granular with your data by applying secondary dimensions and advanced segments. You can create your own segment as shown in the image below:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/781/1*eOwakVEVLJ3JO1Yq1eOOjQ.png" /><figcaption><a href="http://cxl.com/">Image</a>: UA provides you options to create your own segment.</figcaption></figure><p>7. Segmentation can help:</p><ul><li>See how users found your page</li><li>Assess marketing campaigns</li><li>Identify traffic source</li></ul><p>8. You can use funnel visualization reports (found under goals in UA) to identify opportunities such as:</p><ul><li>How users arrive at your site</li><li>Where users land</li><li>Structure of landing page</li><li>Each step along the sales funnel</li><li>Identifying shopping cart problems</li><li>For example, the image below shows the “Cart” stage has significant drop-off. You can easily identify the issue (whether technical or tactical) and resolve it</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/678/1*kBhgTwdE5SMNV5IS2VR4Pg.png" /><figcaption><a href="http://cxl.com/">Image</a>: Funnel Visualization of Google Demo Account showing drop-off at the Cart stage.</figcaption></figure><p>9. Use Google Tag Assistant to audit whether you have Google Analytics tag on every page or not and are you using the tags in a right way.</p><p>Please let me know your views on this blog by leaving your comments below. Next, I am going to share my learnings from Google Tag Manager course in my upcoming blog. Till then, stay tuned!</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=6e3e4222da5a" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Research- & Psychology-Driven Optimization Strategy for High-Converting Landing Page — Review]]></title>
            <link>https://bhavsar-poli.medium.com/research-psychology-driven-optimization-strategy-for-high-converting-landing-page-review-24e7f8f0cada?source=rss-28d4f7b81521------2</link>
            <guid isPermaLink="false">https://medium.com/p/24e7f8f0cada</guid>
            <category><![CDATA[neuroscience]]></category>
            <category><![CDATA[psychology]]></category>
            <category><![CDATA[conversion-research]]></category>
            <category><![CDATA[conversion-optimization]]></category>
            <category><![CDATA[landing-page-optimization]]></category>
            <dc:creator><![CDATA[Poli Dey Bhavsar]]></dc:creator>
            <pubDate>Sun, 28 Mar 2021 18:12:46 GMT</pubDate>
            <atom:updated>2021-03-29T12:48:45.204Z</atom:updated>
            <content:encoded><![CDATA[<h3><strong>Research- &amp; Psychology-Driven Optimization Strategy for High-Converting Landing Page — Review</strong></h3><p>What is a landing page? A landing page is an “entrance page” where your users land on after clicking an ad source. Unlike a homepage that presents the user with several offers and navigation options, a landing page is focused on one specific offer and call-to-action.</p><p>Does your landing page shorten your users’ journey from click to conversion? Does it create a clear path to the conversion goal? If your answers are ‘no’, then you are on the right page.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*i6tNai4Ik30GypKELFsYxg.jpeg" /><figcaption><a href="https://unsplash.com/@koushikc?utm_source=unsplash&amp;utm_medium=referral&amp;utm_content=creditCopyText">Image</a>: Your landing page is analogous to the “entrance” of this building.</figcaption></figure><p>In this blog (6th in the series) I am going to share my learning from the <a href="https://cxl.com/institute/online-course/landing-page-optimization/">Landing Page Optimization Course at the CXL Institute</a> by Michael Aagaard. At the end I would be sharing with you the ResearchXL model — the six-step framework for conducting conversion research.</p><p>The points I’ll cover in this blog are:</p><ul><li>Basic psychology principles and cognitive biases</li><li>Wireframing and information hierarchy</li><li>Landing page optimization research</li><li>Landing page copywriting</li><li>Landing page design</li><li>ResearchXL framework</li></ul><p>Before we get started with the landing page optimization steps, let’s take a few moments to understand two important foundational aspects of creating a good landing page experience:</p><ul><li>how the human brain works; and</li><li>how humans make decisions</li></ul><h3><strong>Basic psychology principles and cognitive biases</strong></h3><p>Michael Aagaard teaches a few basic psychology principles and three cognitive biases which would help you to understand what goes on in your users’ minds and if they are behaving in a particular manner, why they are doing so.</p><p>So, let’s get equipped with this knowledge of how human minds work to create better landing page experiences for your users.</p><p><strong>Modes of Thinking and how they affect behaviour</strong></p><p>According to Daniel Kahneman, Cognitive Psychologist, there are two fundamentally different modes of thinking — fast and slow known as System 1 and System 2 respectively. Understanding the difference between fast and slow thinking would help us understand how each type affects your users’ decision making process.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*ysglOCu9QVvfx8TM1E_0Yw.png" /><figcaption><a href="http://cxl.com/">Imag</a>e: Modes of thinking — Fast vs Slow.</figcaption></figure><p>So, as a conversion optimizer, what’s in it for you? Well, if we human beings are faced with a choice of using system 1 versus system 2, then we are certainly going to use system 1 for most of the part.</p><p>Wondering why? Because humans are programmed to avoid pain as much as possible and using system 2 means we are putting conscious effort and thus feel the strain of thinking. Using system 1, on the contrary, is automatic and effortless.</p><p>Therefore, if your users become cognitively strained with higher cognitive load, they are more unlikely to solve a task (for example, fill a form, etc., in this case) and move forward towards the conversion goal.</p><p>Now, let’s move on to three cognitive biases or mental shortcuts which our brain uses in order to prevent us from using system 2 or slow thinking.</p><p><strong>Three Cognitive Biases</strong></p><p><strong>Priming</strong> — “Exposure to one stimulus influences response to a subsequent stimulus.” As a marketer, you would need to think carefully about what you are priming your users to expect.</p><p>For example, if you have primed your users to expect a free offer and then if they find out that they have to use their credit card information to avail that free offer they are most likely to react negatively.</p><p><strong>Framing</strong> — “The way you deliver a message has a direct impact on how it will be perceived.” Therefore, as a marketer you need to be extra careful in the way you are presenting your information. Because if you failed to present the information in a positive way, system 1 would be quick to perceive it as negative.</p><p><strong>WYSIATI</strong> — “What you see is all there is.” This is a very powerful bias and system 1 utilizes it when it hears a story to interpret things, involuntarily put the pieces together and automatically fill the gaps based on your prior experience.</p><p>Let’s move on to put the hat of a neuroscientist and understand two important brain chemicals.</p><p><strong>Two important neurochemicals</strong></p><ul><li><strong>Dopamine</strong> — Aka the “feel-good hormone” keeps us focused on finding rewards that help us survive.</li><li><strong>Cortisol</strong> — Aka the “stress hormone” helps us avoid pain and thus plays an important role in human behaviour.</li></ul><p>Now, with this understanding when you start crafting optimization strategy for your landing page, you would be equipped with empathy which is crucial to user-centered design.</p><p>Let’s move on to deciding the structure of your landing page and the amount of content that would be just right for it as well as the order of presenting the same.</p><h3><strong>Wireframing and information hierarchy</strong></h3><p>Wireframing can be defined as “a visual guide that represents that skeletal framework of a page or a website”. You can whiteboard it or use a piece of paper or use a software tool like Balsamiq to get it done.</p><p><strong>3 Benefits of having a wireframe:</strong></p><ul><li>It helps you visualize the landing page early on</li><li>Helps you prioritize the content and build structure</li><li>Makes it easier for you to align copy and design</li></ul><p><strong>Now, what is information hierarchy and why is it important?</strong></p><p>Information hierarchy is very important in designing landing pages and landing page experiences as it helps you answer two very critical questions:</p><ul><li>What information is most important?</li><li>How much information is necessary?</li></ul><p>The answer to the first question is that the awareness level of your leads or potential customers would help you figure out what overall information you would need to use on your landing page.</p><p>Michael Aagaard answers the second question as follows:</p><blockquote>“It is fine to have a lot of information on your landing page as long as it is relevant and served in a logical and concise way.”</blockquote><p>Let’s move on to landing page optimization research now.</p><h3><strong>Landing page optimization research</strong></h3><p>You would basically need to carry out two types of research</p><ul><li>Quantitative research to find answers to “what” and “where” questions</li><li>Qualitative research to find answers to the “why” questions</li></ul><p><strong>Quantitative research using Google Analytics</strong></p><p>Using GA you would be able to find insights on:</p><ul><li>Overall landing page performance</li><li>Overall device performance</li><li>Overall browser performance</li><li>Traffic, conversion rate, transactions, bounce rate</li><li>Source, second page, exit page, gender, age</li><li>Potential bugs</li></ul><p>Figuring out what device users are primarily on when visiting your landing page would not only help you create the <strong>optimal landing page experience</strong> but also help you <strong>prioritize your efforts and focus</strong> on the most important device. This is important because <strong>user motivation may vary</strong> depending on which device they are using.</p><p>Manual Step-drop analysis</p><p>Michael also taught this super useful method to help you get an overview of each step in the landing page funnel so you can see where users are exiting and <strong>identify the weakest parts.</strong></p><p><strong>Qualitative research</strong></p><p>Qualitative research helps us find answers to the “<strong>why</strong>” questions — the underlying reasons, motivations and emotions that drive user behaviour.</p><p>To find these answers you can use many different methods including:</p><ul><li>Heuristic, full-funnel walkthrough — Identify the weakest points in the landing page experience</li><li>Message mining at customer review sites — Find out what customers are saying after they’ve bought a product or used your service</li><li>Interviews — customers, customer support &amp; sales — To avoid guessing what the audience is thinking or feeling</li><li>Session recording — Get insights on critical aspects of the funnel</li><li>Heat maps — Find out where users are clicking and how far they are scrolling</li><li>Feedback polls — To get clear actionable insights</li><li>Usability testing such as 5-second test and preference test</li></ul><h3><strong>Landing page copywriting</strong></h3><p>Effective landing page copy plays a significant role in preventing bad landing page user experience by shortening their journey from click to conversion. Other crucial aspects of landing page copy are:</p><ul><li>Answering questions</li><li>Reinforcing motivation</li><li>Addressing barriers</li></ul><p><strong>5 Most important landing page copy elements</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/905/1*SOC6DnczqbKEzA2G5KS3TA.png" /><figcaption><a href="http://cxl.com/">Image</a>: Five most important landing page copy elements.</figcaption></figure><h3>Landing page design</h3><p>You can use the wireframe and information hierarchy to guide your design efforts. Here’s how:</p><p><strong>6 Most important aspect of visual hierarchy</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/671/1*-qeDkj4q4XfTuCvRfeXbAQ.png" /><figcaption><a href="http://cxl.com/">Image</a>: Six most important landing page design elements</figcaption></figure><p>Are you pondering, what element is the most important and where you should make your users look first? Well, no worries. Visual hierarchy would come to your rescue.</p><p>Visual hierarchy guides the user understand what to look at and in which order, through the landing page by the use of design elements and <strong>five important principles:</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/827/1*Wi6qZecFetEj6R9KsYo2Hw.png" /><figcaption><a href="http://cxl.com/">Image</a>: Five Visual Hierarchy Principles.</figcaption></figure><p><strong>Form design</strong></p><p>Your forms play a significant role in the overall landing page experience. So, please do remember that forms should make sense on its own independent of the rest of the landing page. It should be like a landing page within a landing page.</p><h3>ResearchXL framework</h3><p>The conversion research course is very comprehensive and I am just sharing the 6 step framework which would help you as an optimizer to kickstart the research process without getting overwhelmed and in the right direction.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/939/1*XQxIQd8YDGBUBWvrGE76UQ.png" /><figcaption><a href="http://cxl.com/">Image</a>: ResearchXL 6-Step Framework for research-driven conversion optimization.</figcaption></figure><p>As shown in the image, the six steps in this framework are as follows:</p><p><strong>Step 1: Heuristic analysis — </strong>To assess user experience on your website</p><p><strong>Step 2: Technical analysis </strong>— To make sure everything is working smoothly from a technical standpoint</p><p><strong>Step 3: Web analytics analysis </strong>— To identify the weakest or money-leaking point of your website</p><p><strong>Step 4: Qualitative research </strong>— To get customer insights and pinpoint areas of friction to work on.</p><p><strong>Step 5: User testing </strong>— To get direct input on how real users use your site</p><p><strong>Step 6: Mouse tracking analysis — </strong>To identify frictions; what is working and what is not</p><p>To understand how to use this framework, please read this blog:</p><p><a href="https://cxl.com/blog/how-to-come-up-with-more-winning-tests-using-data/">How to Come Up with More Winning Tests Using Data [ResearchXL model]</a></p><p>See you with my next blog, next week; till then please stay tuned!</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=24e7f8f0cada" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Google Analytics 101: Discover the “How” of Your Website Results — Review]]></title>
            <link>https://bhavsar-poli.medium.com/google-analytics-101-discover-the-how-of-your-website-results-review-e8d8887eb5ea?source=rss-28d4f7b81521------2</link>
            <guid isPermaLink="false">https://medium.com/p/e8d8887eb5ea</guid>
            <category><![CDATA[google-analytics]]></category>
            <dc:creator><![CDATA[Poli Dey Bhavsar]]></dc:creator>
            <pubDate>Sun, 21 Mar 2021 17:07:09 GMT</pubDate>
            <atom:updated>2021-03-21T17:07:09.916Z</atom:updated>
            <content:encoded><![CDATA[<h3><strong>Google Analytics 101: Discover the “How” of Your Website Results — Review</strong></h3><p>Are you still making marketing decisions based on gut instinct? Do you get intimidated by the sheer amount of data and have no idea what to do with it? Google Analytics (GA) can help you understand what’s exactly happening with your business if you know how to use it.</p><p>So, no more shooting in the dark! Remove your blindfold in the form of ignorance about data analytics by learning the data-driven approach and improve your marketing and strategy decisions.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*fWdqfummiucuK82Wz_uU-A.jpeg" /><figcaption><a href="https://unsplash.com/@kbalobanov?utm_source=unsplash&amp;utm_medium=referral&amp;utm_content=creditCopyText">Image</a>: No more shooting in the dark; let GA remove your data ignorance blindfold.</figcaption></figure><p>In this blog (fifth in the series) I am going to share my learnings from the <a href="https://cxl.com/institute/online-course/google-analytics-beginners/">Google Analytics for beginners</a> course at the CXL Institute, although at a very high-level.</p><p>Before enrolling for this course, I had little knowledge on the inner workings of Google Analytics and I was tired of missing out. This course by Chris Mercer, the Co-Founder of MeasurementMarketing.io is highly comprehensive. Now I would need to keep practicing in order to get better at the tool.</p><p>Here are a few learnings from the course that I would cover in this blog:</p><p>· When and when not to use GA</p><p>· Dimensions vs. metrics</p><p>· Structure of GA</p><p>· Types of reports</p><p>· Standard reports</p><p>· Types of traffic</p><p>· Types of goals</p><p>· Time in GA</p><p>· Analyzing Reports</p><p>I won’t be starting my blog with ‘what is GA?’ You know it right! Let’s begin with how it functions to better understand the instances when it is wise to depend on GA and when it is not.</p><h3><strong>When and when not to use GA</strong></h3><p>As a business owner, you would have so many ways to know the end results about your marketing efforts. However, what you won’t be knowing is “how”. That’s where GA would come to your rescue.</p><p>GA will help you measure the how and it will showcase that in terms of numbers. However, to make the most of this tool, try to understand that GA stores behaviours in the form of numbers.</p><p>For that it collects, stores and reports data. It’s incredible at storing data (behaviors in the form of numbers) however it’s not so good at collecting and reporting. That’s why you should use GA for storing data, the task at which it is adept at.</p><p>However, it’s better not to use GA for collecting data (rather use Google Tag Manager) and for reporting data (rather use Google Data Studio).</p><p>Another reason for not using Google Analytics would be when you want something incredibly visual. For example, heatmaps or session recordings of how people are using your pages. You can rely on tools like Hotjar or TruConversion for such visual outputs.</p><h3><strong>Dimensions vs. metrics</strong></h3><p>Now, let’s put a quick glance on the difference between metrics and dimensions.</p><p><strong>Metrics</strong> are <strong>numbers</strong> in GA and you can perform any sort of arithmetic with them. Whereas, dimensions are the things that help you to break down and metrics and sort them. For example, Page views is a metric while Page is a dimension.</p><p>There are two types of dimensions available in GA: primary and secondary dimensions.</p><h3><strong>Structure of GA</strong></h3><p>Let me give you an overview of how GA structure is organized. You have an account and accounts are able to have properties. Accounts can have multiple properties. Each property have a view, at least one. We can also have multiple views within a property.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*fQQezNMVMlUBgYHu9NYJ-Q.png" /><figcaption><a href="http://cxl.com/">Image</a>: Structure of GA — Account, Property &amp; View</figcaption></figure><p>For GA linking and property moving, visit this <a href="https://support.google.com/analytics/answer/7011855?ref_topic=1009620">support document</a>.</p><p>Also, there’s a lot to understand about the use cases of this structure of GA, which is extensively covered in the course by Chris Mercer.</p><h3><strong>Different types of reports</strong></h3><p>In GA, you will come across mainly three types of reports:</p><p><strong>Overview reports</strong> — Reports that answers general questions, e.g. Home Report.</p><p><strong>Table reports</strong> — Reports in tabular forms which allows you to drill down further by switching dimensions both primary and secondary, e.g. Acquisition Report.</p><p><strong>Flow reports</strong> — Reports that are very visual and radically different, e.g., Goal Flow Report</p><h3><strong>Standard reports</strong></h3><p>By default, GA offers you these <strong>standard</strong> reports which can be categorized as follows:</p><p><strong>Real Time Reports</strong> — Reports that answer “what”. For example, it answers testing questions such as what is working and what is not.</p><p><strong>Audience Reports</strong> — Reports that answer “who” that is “who are my users?”.</p><p>Acquisition Reports — Reports that answer the “where” questions for you such as “Where are my users coming from?” In other words, they answer how your users found you.</p><p><strong>Behavior Reports</strong> — Reports that show you what <strong>actions</strong> your audience is taking in terms of landing and exit pages.</p><p><strong>Conversion Reports</strong> — Reports that give you the answer to the question “what are the results of all of my users’ actions?”</p><h3><strong>Types of Traffic</strong></h3><p>Traffic is categorized mainly under three types:</p><p><strong>Organic</strong> — Traffic coming from search engines.</p><p><strong>Referral</strong> — Traffic coming from other sites (third-party) other than search engines.</p><p><strong>Direct none</strong> — Traffic that GA has no idea from where it’s coming from.</p><h3><strong>Types of goals</strong></h3><p>In order to <strong>measure results</strong>, you can set different types of <strong>goals</strong> such as:</p><p>Let’s have a quick glance at the four types of goals:</p><p>· Destination goals</p><p>· Duration goals</p><p>· Pages per session goals</p><p>· Event goals</p><h3><strong>Analyzing Reports</strong></h3><p>Chris Mercer introduced a very useful analysis framework known as <strong>QIA</strong>:</p><p><strong>Q</strong>uestion — Find the questions you need answers for from GA.</p><p><strong>I</strong>nformation — Figure out the information that you’re going to need (from GA) in order to get those answers.</p><p><strong>A</strong>ction — Ask yourself what actions you would take to get those answers. This helps you to put the information in context.</p><h3><strong>Time in GA</strong></h3><p>Be cautious about time in Google Analytics, it’s a representation not an absolute. So, when Google Analytics shows you someone spending 30 seconds on your page, please understand it’s a representation and not exact time.</p><p>This is because GA calculates time stamp between two different hits. In case, someone lands on a page and they have no other hits then that duration spent on that particular page will be shown as zero. Interesting isn’t? To understand in detail, I would recommend you to check out this GA course, beginner’s edition today.</p><p>For me, understanding how GA tracks time is the most important takeaway from this course. Let me tell you why! At my previous job, I was once told that the content I write will be judged on the basis of ‘Average Time on Page’ and the benchmark would be 4 minutes.</p><p>After completing this course, now I know that this would have been such a wrong metric for judging if my content was engaging or not. Moreover, I also understood that many self-proclaimed ‘Analytics Experts’ have no idea about how GA tracks time.</p><p>I know this blog gives you a very superficial view of GA because I only preach what I practice. So once I have got my hands dirty with the tool to that extent when I have my own insights, you would definitely have a detailed blog on this topic too. Till then stay tuned!</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=e8d8887eb5ea" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>