<?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 Ron Ruddock on Medium]]></title>
        <description><![CDATA[Stories by Ron Ruddock on Medium]]></description>
        <link>https://medium.com/@RonRuddock?source=rss-aa9c12ee4324------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*mzBM_byTmzToP89qg0FhGA.jpeg</url>
            <title>Stories by Ron Ruddock on Medium</title>
            <link>https://medium.com/@RonRuddock?source=rss-aa9c12ee4324------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Sun, 22 Nov 2026 11:05:42 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@RonRuddock/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[Misplaced Smartphone: Why it’s become a “near death” experience for some, and how to fix it]]></title>
            <link>https://medium.com/@RonRuddock/misplaced-smartphone-why-its-become-a-near-death-experience-for-some-and-how-to-fix-it-f16a88fbb1a4?source=rss-aa9c12ee4324------2</link>
            <guid isPermaLink="false">https://medium.com/p/f16a88fbb1a4</guid>
            <category><![CDATA[smartphones]]></category>
            <category><![CDATA[personality]]></category>
            <category><![CDATA[productivity]]></category>
            <category><![CDATA[motivation]]></category>
            <category><![CDATA[technology]]></category>
            <dc:creator><![CDATA[Ron Ruddock]]></dc:creator>
            <pubDate>Tue, 06 Feb 2018 23:43:30 GMT</pubDate>
            <atom:updated>2018-02-25T23:28:18.465Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/564/1*RDnBfN2Bp-YtxzV_yLQVuQ.jpeg" /><figcaption>Monlux</figcaption></figure><p><em>I have a problem. It’s small, it’s shiny, and it fits in the palm of my hand. It goes wherever I go — to the office, out to dinner, even to the restroom. No, it’s not that — it’s my smartphone. This device has become incredibly important to me. Whenever I misplace it, I suffer what feels like an acute allergic reaction: my blood pressure spikes, I experience shortness of breath. I’ve come to the conclusion that I’m allowing this little device to disrupt my focus on more important things — like sleep, work, study, workouts, driving, and relationships. Something has to give . . .</em></p><p>Last year, I presented my doctoral research proposal to a group of academic peers. My subject: smartphones, and how their usage patterns can be used to predict a users’ personality and demographic traits. I started my presentation with a set of questions, the goal being to demonstrate how pervasive these devices are. First, I asked, “How many of us in this room use a smartphone today?” Every hand went up. Then I asked if users felt they could function throughout the day without their phone. Most in room said “yes,” so I asked for confirmation. (I felt they were being economical with the truth.) To my surprise, many people seemed to be insulted that I had dared to imply they might have a dependency on their devices.</p><p>So, looking back, I understand my audience’s reaction. I was suggesting that they were suffering from Nomophobia — the fear of being without a mobile phone. Some might argue that I’m just a weak-willed life form — but I would point to the new digital “Attention Economy” and its incessant drive to distract me from what I should be doing. Because so many digital platforms — Google, Snapchat, Twitter, Instagram, FaceBook, to name just the big ones — have advertising at the heart of their business model, the name of the game is to get us to keep returning to do more scrolling, pinching, swiping, and liking.</p><p><strong>Why we do what we do</strong></p><p>When we get on a bus, plane or train, behavioral habits tend to kick in. These spaces have a big influence on our device usage. For example, the restroom is fast becoming another primer. How many of us now take our smartphone with us? I even see men at the urinal, swiping and tapping away with one hand on their device. Context is a driver of our behaviors, and in many instances, these behaviors are so close to us, we have become blind to them.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/277/1*yJwn27VwOETwTjt_KWjT9Q.jpeg" /><figcaption>Monlux</figcaption></figure><p>So, context matters. But, there are other drivers that gnaw away at our conscience. Boredom is one that most of us will be familiar with. Whether we’re lying in bed at night, or waiting to pay at the checkout, our need for stimulation moves us to whip out our device and check something.</p><p>Other feelings drive us, too. Consider anxiety for a moment. We often obsess about how we’re perceived by others. Whether it’s the job we hold, our qualifications, or the way we would prefer to be seen, we can be very critical of ourselves. I accept that we were self-critical long before the smartphone appeared, but now we are under siege, due to our immediate access to email and social media.</p><p><strong>Underlying Dynamics</strong></p><p>For instance, lets lift the kimono on one of my favorite apps. I use Delectable to record and share the wines I taste. At first glance, this seems harmless, until I consider the amount of time I spend on the app. I’m on it when I’m trying to escape from my monotonous bus journey. I’ll click on the app to see who’s drinking what. For whatever reason, I make constant comparisons between the wines others tasted, and those I tasted. I would like to drink that 1982 Lafite that Harry drank, but can’t afford to. I feel some envy when I make these comparisons, but I can’t afford to drink that stuff — but I’d like people to think I can. So, I’m performing this high-wire act, which, in the end, makes me feel anxious.</p><p>It’s easy to get hooked on this technology. Apps are preferred to a native browser — their interface is less cluttered, and therefore easier to navigate. The smartphone is crushing other usage barriers, too. In the past when I played video games, I had to be moored to my couch, but now I can binge-game anywhere, anytime.</p><p>In this new digital age, companies per se won’t acknowledge it, but when was the last time that you felt subtle pressure to immediately respond to that piss-o-gram email you just received, regardless of where you are, or what time it is? Work time is spilling into what was our social time. Mining the divide, it’s implied that we’re expected to be always connected and reachable — assuming you like your job and want to keep it.</p><p>There’s no surprise when Inc. (2017) report that on average, millennials check their phones 150 times per day! Other age groups are not too far behind. Digital enterprises have cultivated a generation of internuts with their click-bait calls-to-action. It’s true that the smartphone has freed many of us from mundane crap. From hailing a cab to finding a date, or locating your next job, all of these daily challenges have been reduced to swipes and taps — what’s not to like here? On the flipside, disruptions, interruptions and push notifications have reached an industrial scale. Because of ease of use and my strong internal drivers, the smartphone exploits my individual motivations, at the expense of my attention.</p><p><strong>Get Your Proverbial Shit Together</strong></p><p>If you think you may have usage issues, download the free Moment app. Created by Kevin Holesh, this app analyzes your smartphone usage by capturing total screen time, app usage, apps most used, device pickups, etc. Once you receive your first usage summary, you’ll probably have the same “oh shit” moment I had…</p><p>After 24hrs, you’ll have access to your usage summary. Figure 1 illustrates a couple of screen-shots from my analysis. Panel 1 shows my overall usage summary of 5h 21m. But with this, I could see exactly where I was spending my attention (red border = bad, green = OK.) Panel 2 shows specific times when I accessed my device. The initial mistake I made was to spend time on this high-res detail in panel 2. Start by focusing on those points in the day when you transition from social to work-time (bus, train, plane, or dare I say it, the car.) That’s where the real value hides.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/604/1*y6Ib2Rnoqo9a5qzftZq7_Q.jpeg" /><figcaption><strong>Figure 1 — Smartphone Usage Data</strong></figcaption></figure><p>Not all smartphone usage is a waste of our attention. For those of us who rely on it for work, study, listening to music or reading a book, it’s useful. On the contrary, tack away from FaceBook, Twitter, Instagram, etc., since among other things, they tend to drive excessive media binging — it’s like that huge bucket of popcorn, we don’t know when to stop eating! Finally, In general, push notifications from the majority of apps should be considered as a third rail. Don’t be afraid to swipe to the left on the push notification on/off button.</p><p>I hear you, as Chip &amp; Dan Heath (2010) note — change is hard, really hard. When I started using the Moment app, I wanted to spend more time on what mattered most to me. Of course, this approach is relatively easy to game; switching to a desktop device will reduce your screen time and frequency of app usage on your mobile device. But you’re only fooling yourself — don’t go looking for loopholes to squeeze into. That being said, this is not a fight or flight scenario, either. Small, meaningful improvements will give you encouragement to continue, and are probably more sustainable.</p><p>We live repetitive, structured lives. Start by monitoring your device usage during those transition periods, then make behavioral adjustments. Good habits need to be planned. And as I learned from Holesh, “don’t even reach for your smartphone without clear intent.”</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/170/1*ijsXdI_sCz-XES8Nt22pqQ.jpeg" /></figure><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=f16a88fbb1a4" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Statistical significance: Why it often doesn’t mean much to marketers]]></title>
            <link>https://medium.com/@RonRuddock/statistical-significance-why-it-often-doesnt-mean-much-to-marketers-d5bec3e1ed4?source=rss-aa9c12ee4324------2</link>
            <guid isPermaLink="false">https://medium.com/p/d5bec3e1ed4</guid>
            <category><![CDATA[data-science]]></category>
            <category><![CDATA[marketing]]></category>
            <category><![CDATA[statistics]]></category>
            <dc:creator><![CDATA[Ron Ruddock]]></dc:creator>
            <pubDate>Tue, 24 Oct 2017 09:24:48 GMT</pubDate>
            <atom:updated>2017-11-01T19:41:44.306Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/329/1*sWaDtkHwyA1bjQagoV8Kjg.jpeg" /></figure><h4><strong>Introduction</strong></h4><p>I know how you feel. You read the title of this article, saw the term “statistical significance,” and tasted vomit! But I promise not to bore you with functions, equations, or formulas. I wanted to aim these thoughts at marketers. In today’s Big Data environment, there is less of an opportunity for marketers to engage target consumers in real time. More and more businesses rely on the Internet and digital capabilities to deliver value. So, while you are not able to communicate with your customers face-to-face, consider asking questions of your digital data instead.</p><p>Traditional marketers are facing a dilemma. In this new hyper-connected world, marketers have access to new data sets that take the form of bits and bytes — billions of them! — generated by digital technologies. Increasingly, marketers are turning to statistics and statistical models to help them transform this data into insights. The intent of this article will broadly explore the differences in predictive and explanatory statistical models, to help you better understand the relevance of statistical significance in each case.</p><p>From a marketer’s perspective, the use of statistics is typically not for the validation of theories, which tends to be the domain of academics. As a marketer, you seek approaches and methods that have immediate commercial utility, or predictive power. In order to generate predictions, you require theories or principles upon which our predictions are based. Statistics and statistical models are the bridge between your theories and predicted outcomes. In essence, these models are an approximate representation of the real-world scenario that you wish to model. It is at this point that marketers tend to get lost in the sauce of what is relevant — theory testing or predictive accuracy. Whether your interest is theory testing or predictions, each has distinguishing features that marketers need to be aware of. (More on this later.)</p><p>Statistics are used extensively across the sciences. In the case of the natural sciences (Physics, Chemistry, etc.), phenomenon are often represented through mathematical formulas or functions (based on fundamental principles). But marketers focus on human behavior — and modeling human behavior is more complex than modeling natural phenomenon. For example, how would you model the likelihood that somebody may choose product A over B? We cannot simply apply a formula to this behavioral scenario, but we can use other models to identify if there is a relationship between given variables within a model.</p><p>As an example, I will use some of my own research to explain explanatory and predictive outcomes, and the relevance of statistical significance within these contexts. My focus is limited to multivariate statistics. This helps to limit the discussion to the most relevant topics in order to explain the concepts involved. I will first outline some important definitions, then I will describe my research goals and connect these to the statistical model. Next, we will briefly review a subset of my results, in order to locate the role of statistical significance. Finally, I will connect measures of statistical significance back to the research goals. I have intentionally over-simplified the process, in order to get to the heart of what statistical significance means within the context of explanatory and predictive statistical models. Assume that my sample data set has been randomly generated, and that my sample size is sufficiently large and balanced equally between male and female subjects. I have also controlled for bias within the statistical model.</p><p><strong>1.</strong> <strong>Let’s Review the Main Definitions and Concepts</strong></p><p>Statistical models are typically used to describe, predict, and explain phenomenon in our world. Academics and practitioners tend to focus in two major domains: Seeking the truth and driving profits. A statistical model is nothing more than a theoretical construction based on something that has been observed in the real world. In the case of explanatory models, using data, statistical models are used to test hypotheses, which you will have constructed as an integral part of the model. The model is made up of variables or factors (<em>X</em>), which are assumed to have an association with the dependent variable (<em>Y</em>). We use this model to test a hypothesis and, thus, explain the phenomenon at question.</p><p>While predictive models use statistics similarly, their goal is to <em>predict</em> new observations. In the example that follows, my goal was to predict the <em>Y</em> — in this case, the gender of a smartphone user based on the apps installed on his or her device. The high-level process I used to construct the research model is identified in figure one.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/586/1*S47ZXNMMLfwxsZ9uOUNfig.jpeg" /><figcaption>Fig 1 — Research Model Build Process</figcaption></figure><p><strong>2.</strong> <strong>The Research Example and Associated Goals</strong></p><p>My context is that I’m a marketer, and that I only have access to smartphone apps (digital data) that my subject has used. I want to segment my clients without asking them questions, so I used the digital data available to me. Based on current behavioral theories, I have established that the usage of certain app categories are gender-based. Figure two shows my conceptual model, with app categories theoretically associated with females. My hypotheses have also been applied to the model.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/315/1*TnL5_9EWKsLOshzTPAwv3Q.jpeg" /><figcaption>Fig 2 — Smartphone App Conceptual Model</figcaption></figure><p>For this article, we will focus on the category of Travel. Since my dependent variable (<em>Y</em>) is a categorical format (male/female), I know that the binary logistic regression model is an appropriate model to use for classification purposes. My independent variables are represented as a binary combination of variables (category of app used or not used). Meaning, my <em>X’</em>s vary between 0 and 1. My research model is predictive; validating the theory (explanatory) is of little importance to me. From a marketer’s perspective, what is most important is the ability to accurately categorize people by gender.</p><p><strong>3.</strong> <strong>Interpreting the Results from the Model</strong></p><p>The conceptual model defined in Figure 2 can be used in both cases. Academics typically use the term “explanatory” when they want to understand how something is working. This is where the hypotheses come in. Here we use the hypotheses to test the relationship between our dependent and independent variables. Since the theory is expressed as hypotheses, this is considered “retrospective,” as the measurement criteria have been defined in advance of the statistical tests. Statistical significance is a quantitative test, because it is expressed as a numerical value. It is used to help reduce or eliminate uncertainty during measurements. You may have heard of the term “<em>p</em>-value,” and a typical threshold of .05, typically defined as the critical value by which measurements are deemed significant or not. For this statistical model, I set my significance level at .1. Let’s review the results of my tests in Table 1 and discuss what these results mean.</p><p><strong>Scenario 1 — Explanatory Model Analysis</strong></p><p>In table one, you can see the measured <em>p</em>-value (highlighted) for this model is .066. Given that I set my significance level to .1, this value is acceptable</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/448/1*xRUNG1CUm3a6evh6UTNkkg.jpeg" /><figcaption>Table 1 — Travel Apps Predicting Gender</figcaption></figure><p>The academic community may not be comfortable with my <em>p</em>-value, but allow me to argue my approach. Statistical significance was the main method by which I could efficiently include/exclude variables from my model. With approximately 25 app categories, I needed a way to reduce the model to only include the variables that demonstrate variance. Theory guided me initially, but I used this measurement to help with the variable selection process. This result suggests that there is a true effect that we have measured. BUT, so what? In terms of theory testing, we are estimating if there is a relationship between the independent and dependent variables. That’s all.</p><p>Here’s the tricky part: we want to assess if there is a probability that we are measuring noise, even though there may not be any real effect. Just because we are confident that we have measured something, it still doesn’t mean that the measurement is accurate, or that we are measuring the right thing. What is more challenging is that statistical significance does not tell us about the strength of the underlying relationship. It simply provides us with a level of confidence in the results we obtained. I concluded that this result did not occur by chance alone. In this case, I validated the hypothesis (rejecting the null hypothesis), and accepted this aspect of the theory. However, as I stated previously, my overarching goal was <em>not</em> testing the theory, although the theory has guided the construction of my statistical model.</p><p><strong>Scenario 2 — Predictive Model Analysis</strong></p><p>When attempting to evaluate the accuracy of our predictions, the researcher needs to first identify which independent variable helps to generate the most accurate prediction. In this case, I generate simulations to identify the variables of interest, and calculate the error rate when predicting gender, using the Travel app. For the sample data set in question, this method assesses the performance of the model, and is measured by the Quality of Fit, or the mean squared error (MSE). When the predicted output is close to the observed data, the MSE value will be small; otherwise, the MSE will be large. The output of the model was cross-validated using data collected from the data donors. You should note that statistical significance has no role to play in this measurement.</p><p>Referring back to scenario 1, I now have a variable that has demonstrated some degree of significance within the model. We can use a little more statistical trickery to help us to understand whether the use of the Travel app is associated with predicting gender. The Chi Square test ( 𝛘² ) can be used for categorical (binary) variables, and provides insights in terms of the relationship between the Travel app, and the dependent variable; gender. In this example, the Chi Square test enabled me to compare the results of who installed the Travel app, in comparison to a theoretical distribution (lets assume that the probability of theoretical distribution is 50% male and 50% female, which is the same as tossing a coin).</p><p>This distribution can be tested using a contingency table and Chi Square distribution table. Based on the output of the contingency table, a Chi Square statistic is generated. This value is then used with a Chi distribution table. The Chi distribution table enabled me to determine the actual probability of my results, and the statistical significance of what I measured — ( 𝛘² = 3.589, <em>p</em> = .058). This result indicates that with one degree of freedom, my Chi Square statistic value is 3.589. Using the Chi Square value and degrees of freedom, I can look up the probability level for the chi statistic. My critical value for significance was set at <em>p</em> &lt; .1. Based on the table, the significance of my result is between .1 and .05, and the result is much closer to .05. This strengthens my confidence that I am truly measuring something, and that it’s not just a fluke. See table 2 below.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/331/1*wfmm5CvUrUZS7ozAzUbpsw.jpeg" /><figcaption>Table 2 — Chi Squared Distribution Table</figcaption></figure><p>Analysis of the contingency table revealed that nearly 70% of users in this sample, who had installed the app, were female. Therefore, there does appear to be a relationship between Travel app and Gender (Since the actual installed distribution was significantly different from tossing a coin). Consequently, I concluded that there was a relationship between the female gender and the Travel app variable. Again, assuming my sample is sufficiently large, all of this work merely suggests that the Travel app variable <em>may</em> be a good variable to use in my statistical model, when trying to make gender predictions. But the model needs to accurately represent the data, and the data must accurately represent the broader population in question, otherwise the model output could be adversely biased. At this point, I have selected my variable and constructed the model. In this scenario, we now present new data to our model in order to establish the performance of the model (calculating the prediction error rate).</p><p><strong>4.</strong> <strong>Statistical Significance — So What, Who Cares?</strong></p><p>Referring to figure 3, I hope you can now see how and where statistical significance is often deployed, and why. In both scenarios, statistical significance was used to test hypotheses. Hypotheses are quantitative endeavors, but sorting people into groups is qualitative.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/841/1*w7k73FxGbw8EH_PjNcKb4g.jpeg" /><figcaption>Figure 3 — Hypothesis Tests Within Explanatory and Predictive Models</figcaption></figure><p>Whether you are trying to prove a theory, or predict a future event, although statistical significance can play a variety of roles, you really need to understand why it is used and what it means in terms of your results. The importance of statistical significance is task-dependent, and is a means to an end. In my example, I could measure a high level of statistical significance within scenario two, but if I’m not able to accurately predict gender from the variables available to me, statistical significance is pretty much irrelevant. No statistical measure is perfect, as the accuracy of the statistical approach depends on many factors, some of which I have chosen not to raise in this article. Remember earlier in my article, I set my critical value of significance to .1. I did that with the understanding that I chose the accept/reject threshold, since I understand the consequences of using the wrong threshold — it’s my risk to assume. For example, if you are completing research within the healthcare environment, statistical significance takes on a higher level of importance.</p><p>Need some help? Give SWBR a call.</p><p>Author: <strong>Ronnie Ruddock</strong> (inspired by many)</p><p><em>Ronnie is the principal and owner of Switching Barriers (SWBR), LLC business research consultancy, where digital technologies and consumers meet. </em><a href="http://www.switchingbarriers.com%C3%82%C2%A0"><em>www.swbr.nyc</em></a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=d5bec3e1ed4" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[An exciting new development in Digital Healthcare Consumer Research…]]></title>
            <link>https://medium.com/@RonRuddock/an-exciting-new-development-in-digital-healthcare-consumer-research-a0d2c994e8ea?source=rss-aa9c12ee4324------2</link>
            <guid isPermaLink="false">https://medium.com/p/a0d2c994e8ea</guid>
            <category><![CDATA[digital-health]]></category>
            <category><![CDATA[marketing]]></category>
            <category><![CDATA[entrepreneur]]></category>
            <dc:creator><![CDATA[Ron Ruddock]]></dc:creator>
            <pubDate>Wed, 02 Mar 2016 04:08:10 GMT</pubDate>
            <atom:updated>2016-03-02T04:08:10.904Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*dIp4iGumOKzk7HXp.png" /></figure><p><em>Switching Barriers is here!</em></p><p>While you, the digital entrepreneur, focus on medical product design and development, let Switching Barriers (SWBR) help you engage and better understand your consumers’ wants and needs.</p><p>What distinguishes SWBR from others in the Consumer Research community is that we focus solely on consumers and the data on their digital healthcare devices. Through cutting-edge, proprietary research, we enable the identification and elimination of barriers in place to bring you closer to the money.</p><p>Give Switching Barriers a call and let’s discuss your commercial barriers. We can help.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=a0d2c994e8ea" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Consumer Insights: Taking the First Steps to Knowing Your Customers]]></title>
            <link>https://medium.com/@RonRuddock/consumer-insights-taking-the-first-steps-to-knowing-your-customers-209017bc60a5?source=rss-aa9c12ee4324------2</link>
            <guid isPermaLink="false">https://medium.com/p/209017bc60a5</guid>
            <category><![CDATA[entrepreneurship]]></category>
            <category><![CDATA[digital]]></category>
            <category><![CDATA[startup]]></category>
            <dc:creator><![CDATA[Ron Ruddock]]></dc:creator>
            <pubDate>Tue, 01 Mar 2016 06:48:57 GMT</pubDate>
            <atom:updated>2016-03-01T06:49:26.392Z</atom:updated>
            <content:encoded><![CDATA[<p>(7 min read)</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/700/1*ODKYUTUoq1bnjceCBQ1jGQ.jpeg" /></figure><p><em>So you’re looking to build a lean startup? Then you have to get your consumer research on</em>…</p><p>It’s no surprise that <em>Entrepreneur</em> magazine lists consumer research as one of the most important practices of successful startups. Indeed, careful, focused research can be the difference between failure and success. Research is one of the most powerful tools for determining <em>where you will play</em> in the market and <em>how you are going to win</em>.</p><p>You may already know you need to do consumer research, but the question still remains: “When is the right time to engage with the consumer?” This article answers this question by exploring the following topics:</p><ol><li><strong>Where are you today?</strong></li><li><strong>The Product/Solution Fit</strong></li><li><strong>Making Your Research Data Actionable</strong></li><li><strong>Build, Validate and Learn</strong></li></ol><h3>1. Where are you today?</h3><p>Startup Genome has completed excellent work around what it defines as the “Internet startup’s stages of growth.” Best practice tells us that if you know where you are, you will have a better idea about which way you should travel. So your first step is to understand where you are relative to the stages of business growth. For the purposes of this article, I’ll assume you are in the Discovery phase, but in order to place this important step in context, I will also touch on the Validation phase (Figure 1).</p><p>Figure 1 — Internet Startup Stages of Growth</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/730/0*fnXw3bLDpevIuvKr.jpg" /></figure><h3>Discovery Phase (First Base)</h3><p>After you have identified an important under-served need in the market, you enter the Discovery phase: the first step to gathering empirical evidence that the consumer has an interest in your nascent solution. You haven’t made a significant investment at this point; you’re just floating a trial balloon. Key characteristics that identify you are in Discovery phase are:</p><ul><li>Some customer interviews have been conducted</li><li>Your startup has one or two employees</li><li>A Minimal Viable Product (MVP) has been created</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/337/0*OIMCXxQwRracz56d.jpg" /></figure><p>Testing and validating your value proposition is what gets you to first base. The output of the customer interviews are intended to create a near-complete picture of your consumers and the problems they are wrestling. But most importantly, these interviews can help you better understand if your proposed value proposition and associated solution solve one or more of your customers’ problems. It’s called the Problem/Solution fit. You accomplish this by analyzing both explicit and implicit customer responses, then organizing the information so themes, patterns, and causes become more visible. This way, you’re not saturated with opportunities (refer to section 3 for more details).</p><h3>Validation Phase (Second Base)</h3><p>The Validation phase helps you confirm that there is demand for your product in the marketplace. You’re in Validation when:</p><ul><li>Your startup has 2–4 employees</li><li>You have secured funding from sources other than friends and family</li><li>You have refined core features of your offering based on output from customer interviews</li><li>You have your first paying customers</li></ul><p>Quantitative research should be deployed to examine if your product is a good fit across your target market (Product/Market Fit); this enables generalizability with the data.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/355/0*kaQHtQ_ma4LiwFbt.jpg" /></figure><p>But qualitative research isn’t enough. In fact, buoyed by a successful qualitative intervention, entrepreneurs often flirt with, and are often lured into the “greedy zone.” With their vision clouded by the fantasy of instant success, they misinterpret the distance they have to go and the obstacles in their way and make a mad dash to steal second base — and are invariably thrown out.</p><p>This stage requires you to validate your proposed solution, with a much larger sample size, before scaling your operation.</p><h3>2. The Product/Solution Fit</h3><p>When you’re in the Discovery phase — i.e., when you’re safe at first base — allow yourself to catch your breath. There’s no rush. The best players are not reckless tacticians; they’re patient strategists. The point is to get to home plate, not to get to home plate <em>now.</em></p><p>At this point, you are still at first base — Discovery phase — and your MVP is, in essence, your proposed solution. Your second customer interview should be designed to test and validate your proposed solution; it needs to focus on your targeted opportunities and how the minimum set of features address the customers’ problem. Additional questions to be addressed should revolve around the following topics:</p><ul><li>Will your target customers pay for the proposed solution?</li><li>What will they pay for the proposed solution?</li></ul><p>I assume that at this point, a prototype of the solution is available for you to demonstrate how you intend to solve your customer’s problem. The prototype needs to be as close to reality as possible, otherwise you risk creating expectations that you may not be able to meet.</p><h3>3. Making Research Data Actionable</h3><p>Within the context of the consumer’s journey, well-structured and executed research provides a clear lens through which to view customer needs and determine potential solutions. These problems may be partially addressed, they may be over-served, or they may not be addressed at all. To illustrate the point, lets take a simple everyday example.</p><p>I recently visited my local pharmacy to buy some eye drops during my lunch-break. We can break down the overall process into a number of steps:</p><ol><li>Locating the eye drops</li><li>Product selection and price</li><li>Checkout</li></ol><p>For me, getting in and out of the shop quickly is essential. Signs were well placed and I was able to locate the shelf quickly. When I arrived at the shelf, there was a bewildering array of eye drops at different prices. After some deliberation, I made my choice and moved to the checkout. There was a huge line of customers waiting to pay. Figure 2 breaks down my analysis of the process. Importance and performance data is captured using a simple survey, which uses a scale of 1–10 (1 = Very Poor and 10 = Excellent). Each step in my journey is then ranked in terms of opportunity, based on importance and performance. See Figure 2.</p><p>Figure 2 — Identifying the Most Critical Needs</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/681/0*ErcGsJuTgDZKSkdi.jpg" /></figure><p>My example shows that locating the product and pricing were satisfactory. There is little room for improvement here. However, at the final step, while rated as very important, the actual performance was very low, resulting in a relatively high opportunity score. A simple formula design by Anthony Ulwick enables the prioritization of opportunities. Opportunity scores can range from 0 to 20 — the closer to twenty, the greater the opportunity. The opportunity rating is a measure of the discrepancy between the perceived importance and actual performance.</p><p>This example illustrates under-served needs; here, the opportunity score is high. Customers are over-served when satisfaction ratings surpass their respective customer importance rating — ripe for cost reduction or disruption. (I didn’t feel over-served at any point in my journey to purchase eye drops; I don’t predict any disruptive innovation coming out of that pharmacy any time soon.)</p><p>Now that you have identified what matters most to your customers, you can build a data-model for delivering solutions and hard-wire it to your product-development road-map. Your data model is fundamental to generating insights and should be considered <em>native</em> to the application or eco-system. Now you also have a baseline from which to measure the performance of your proposed solution. But of equal importance, you can separate the wheat from the chaff and identify the most important customer needs — I need to be able to checkout, <em>fast.</em></p><h3>4. Build, Validate and Learn</h3><p>Best practice indicates that the Discovery phase helps the entrepreneur to validate her approach before scaling to address a proposed solution. Why scale and waste time and money if you have not already confirmed that the market is ready to embrace your solution?</p><p>Once you have identified viable growth opportunities, you need to determine the types of data that you will gather and analyze to validate your MVP. This is why</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/230/0*UkPiMrnyIPXuKtgY.jpg" /></figure><p>successful startups do not begin building the MVP before establishing a clear view of their opportunities. If you make that mad dash to second base, you’re almost guaranteeing yourself a great deal of rework and non value-added costs!</p><p>Targeting the right data sets, along with accelerating data turns will change the game for you. In particular, the data your app (or eco-system) generates can be used to measure customer engagement, along with evaluating technology acceptance, usage and adoption. For example, my Polar Loop tracks my physical inactivity and notifies me that I need to move to keep active. It’s a gentle nudge, but the data has a significant influence on my behavior, thus creating information for me that is immediately actionable.</p><p>Need some help? Give SWBR a call.</p><p>Author: <strong>Ronnie Ruddock</strong> (inspired by many)</p><p>Editor: Kevin Dwyer</p><p>Illustrator: Mark Monlux</p><p><em>Ronnie is the principal and owner of Switching Barriers, LLC business research consultancy, where digital health technologies and consumers meet. </em><a href="http://www.switchingbarriers.com%C3%82%C2%A0"><em>www.swbr.nyc</em></a></p><p><a href="http://www.entrepreneur.com/article/247159">What makes startups succeed — Entrepreneur</a></p><p><a href="https://s3.amazonaws.com/startupcompass-public/StartupGenomeReport1_Why_Startups_Succeed_v2.pdf">Startup Geneome stages of growth</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=209017bc60a5" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>