<?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 Jack Smith on Medium]]></title>
        <description><![CDATA[Stories by Jack Smith on Medium]]></description>
        <link>https://medium.com/@abeckerman002?source=rss-1ce2583f4c85------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*KqzTVtZJqfOXIGJGTX4ETQ.png</url>
            <title>Stories by Jack Smith on Medium</title>
            <link>https://medium.com/@abeckerman002?source=rss-1ce2583f4c85------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Fri, 20 Nov 2026 05:07:28 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@abeckerman002/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[The life of a Marketing Analytics Team at a SaaS Company]]></title>
            <link>https://medium.com/@abeckerman002/the-life-of-a-marketing-analytics-team-at-a-saas-company-801071674085?source=rss-1ce2583f4c85------2</link>
            <guid isPermaLink="false">https://medium.com/p/801071674085</guid>
            <category><![CDATA[saas-marketing]]></category>
            <category><![CDATA[marketing-analytics]]></category>
            <category><![CDATA[saas]]></category>
            <category><![CDATA[analytics]]></category>
            <category><![CDATA[data-science]]></category>
            <dc:creator><![CDATA[Jack Smith]]></dc:creator>
            <pubDate>Sun, 22 Jan 2023 03:53:17 GMT</pubDate>
            <atom:updated>2023-01-22T03:58:43.865Z</atom:updated>
            <content:encoded><![CDATA[<p><em>The creators of all green performance slides</em></p><p><em>“Half the money I spend on advertising is wasted; the trouble is, I don’t know which half.”</em> — John Wanamaker</p><p>This post is a humorous (and slightly exaggerated) take on Marketing Analytics jobs (or maybe, more accurately, Marketing jobs in general) at SaaS companies.</p><p>For context, I worked at a digital advertising agency for 1.5 years. I then worked at Salesforce on a Marketing Analytics team for over 2 years, and then moved onto a Sales Strategy team at Salesforce for ~4 years.</p><p>This post is supposed to be tongue in cheek, but also hopefully have helpful pearls of wisdom interwoven throughout.</p><p><strong>Background on Marketing at SaaS Companies</strong></p><p>If you’re not familiar with Marketing at SaaS companies, typically Marketing runs campaigns that fall into one of four buckets:</p><ul><li><strong>Digital:</strong> These are campaigns that promote ads that appear digitally on desktop, mobile, or tablet (eg. facebook ads, instagram ads, and youtube video ads etc.)</li><li><strong>Organic:</strong> These are campaigns centered around optimizing a company’s website and SEO (“organically” increasing the number of website visits)</li><li><strong>Events: </strong>These are<strong> </strong>events the company hosts. These can both be large events<strong> </strong>(like <a href="https://colab.research.google.com/drive/1t9pjJiKf2KJ5VQaw0zdA2wK8DCpxeYYs?usp=sharing">Dreamforce</a>), or smaller more personalized events like inviting a customer onsite to hear from a top executive in the hope that this will help close a deal</li><li><strong>Other:</strong> There a few other marketing channels that don’t fit neatly into the above 3 campaign categories. For example, Direct Mailer (sending a personalized gift to a top prospect or customer) and Content Syndication (purchasing leads from 3rd party companies)</li></ul><p>A Marketing department’s main job is to run campaigns effectively and efficiently to accomplish one of two goals: (1) increase the awareness of the company/product while helping build the brand of the company (2) help the sales team close larger and more frequent deals so they can hit their quarterly and yearly revenue targets.</p><p>Ultimately, both of these goals help the company increase its revenue. The first goal of boosting a company’s awareness just has a longer timeline before the increase in revenue materializes.</p><p>Now where does the Marketing Analytics team fit into this? A Marketing Analytics team’s objective is to use data (1) to support the campaign managers in figuring out what campaigns are working and how to make them work even better and (2) support the campaign managers in figuring out which campaigns are <em>not </em>working. However, this is easier said than done. And not for the reasons you might suspect.</p><p>Campaign managers are extremely protective of their campaigns. Not surprisingly, they don’t like being told a campaign is not working as they are afraid it may be interpreted as a reflection on them. Campaign managers can exert large influence over Marketing Analytics teams because usually they are the Marketing Analytics team main stakeholder. Due to this, there is a strong motivation to show campaigns in a positive light for all parties involved.</p><p>As someone who was on a Marketing Analytics team, I occasionally found myself reaching down into the deep barrel of Marketing metrics fishing around to find that one barometer over-performing for a campaign. In one desperate moment, I might find myself responding to a campaign manager with something like “Well it looks like there is a 10% increase in the number of page views since the campaign was run 3 years ago.” I’ll hear them answer back excitedly, “Wow, outstanding insight! Let’s highlight that!”… “And don’t forget to highlight that in green!”</p><p>When working on a Marketing performance slide, it’s not uncommon to hear phrases like, “This is not a green, yellow, and red exercise. Treat this more like a green and yellow exercise.” In other words, a Godfather-esque kind of an agreement. “I don’t want to make things complicated for you, but we understand each other, right, Mr. Corleone? Marketing is in fact indeed firing on all cylinders?” “Yes, yes of course!” The end result? A slide that is essentially all green.</p><p>Sometimes I felt like my role was not to actually help campaign managers optimize their campaigns (as the role is advertised). But, rather, I saw myself as a hired attorney appointed to represent and defend my client’s — in this case the campaign manager’s — marketing decisions. In other words, it was my job to prove — beyond a reasonable doubt — that a campaign manager’s judgement should not be questioned, their marketing campaign was working, and should continue to run exactly as is.</p><p>If I’m skilled at my job, nothing will change. This of course is an exaggeration, but all too often critical questions do not get asked of how campaigns are run, and, in general, healthy skepticism is avoided. It is a running joke between my old marketing analytics co-workers and me that marketing is constantly in a state of “all green” performance where everything is exceeding its goal (regardless of how the company is performing).</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/431/1*rir4aTu0j5IP3lUwBAt97Q.png" /><figcaption>The color palette I’m used to working with when populating Marketing’s performance slides</figcaption></figure><p>By no means is this “all green all the time” state completely Marketing’s fault. There is a real pressure to perform. Sales and the broader company need to give the Marketing department room to fail. This is very important. In fact, the Marketing department should be encouraged to fail and even awarded for it! As Jeff Bezos says, “failure and invention are inseparable twins”, so if you’re not failing you’re certainly not inventing either.</p><p>Some tips on how to avoid the façade of a Marketing department that only produces flawless campaigns and has no room to improve:</p><p><strong>Tie Marketing Performance to Sales Goals.</strong> As much as Marketing is resistant to do this, the truth of the matter is “when Sales wins everyone wins.” Unfortunately this is not true with Marketing. By celebrating Marketing metric wins, you might actually end up commemorating the wrong metric (unfortunately website visits and “hot” leads don’t pay the bills). Marketing metrics may be deceptive and seem “good”, but may not objectively be good for the company. While it’s true tying Marketing performance to Sales performance, is not completely fair to Marketing as they don’t have full control of a deal (the Sales team still needs to accept the opportunity and close the deal) it’s the best alternative.</p><p>By tying the two department’s performance together, Marketing and Sales are much more likely to work as a synchronized team, which is ultimately good for the business. This will also have the added benefit of keeping the Marketing department engaged with the pulse of Sales and revenue metrics giving them a better understanding of when budget should be spent and what campaigns need to be run.</p><p><em>How can this be accomplished?</em> One way is to make the Marketing’s goal the same as one of Sales key metrics. For example, a good metric to have both Marketing and Sales rally behind is the Sales’ team pipeline goal. To be clear, Sales should still be accountable to this metric, and probably still should own the eventual outcome, but having this “joint” accountability between Marketing and Sales allows Sales to put more pressure on Marketing to help (which in the end is good for the business).</p><p><strong>Keep Metrics Simple</strong> Metrics vary by industry, but in SaaS common metrics are pipeline, ARR, ACV. Whatever your metric of choice is, it should be easy for both Marketing and Sales to understand. These joint metrics should not be tangled in a web of attribution. This will also help you achieve the first bullet of corporation across Marketing and Sales. The point of metrics needs to be to motivate behavior in a certain way, this is extremely difficult when the metrics aren’t well understood.</p><p>One of the best ways to do that is to keep the metrics simple and levers to improve these metrics should be obvious. Also making the metrics more “hard” (like revenue) makes it less likely the metric will be gamed or optimized for something that may not in fact matter (like page views).</p><p><strong>Marketing and Sales agree on goals before campaign launch</strong>. It’s important to have this expectation setting where both Marketing and Sales agree what success looks like for a campaign. This protects Marketing from Sales coming back weeks into the campaign demanding more. This also simply makes it very easy to know if a campaign is actually working and exceeding its goal or not. For bonus points, it’s also a good idea for Marketing to provide Sales with frequent updates to keep them informed and also give the teams time to pivot if needed.</p><p><strong>Promotions for Marketing should not be directly tied to campaign performance. </strong>Marketing promotions should be awarded based on key learnings and insights. What are they doing to <em>improve </em>a campaign’s performance?</p><p>It’s not easy to run Marketing effectively but is very much worth it if you can get it right. Make your slides green <em>and</em> believable by truly helping generate more pipeline <em>and </em>closing more business. Good luck!</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=801071674085" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Starting a (failing) business but still having fun and learning along the way]]></title>
            <link>https://medium.com/@abeckerman002/starting-a-failing-business-but-still-having-fun-and-learning-along-the-way-4ca3733b4858?source=rss-1ce2583f4c85------2</link>
            <guid isPermaLink="false">https://medium.com/p/4ca3733b4858</guid>
            <category><![CDATA[hermosa-beach]]></category>
            <category><![CDATA[restaurant-tech]]></category>
            <category><![CDATA[entrepreneurship]]></category>
            <category><![CDATA[startup]]></category>
            <category><![CDATA[business]]></category>
            <dc:creator><![CDATA[Jack Smith]]></dc:creator>
            <pubDate>Fri, 13 Jan 2023 03:34:02 GMT</pubDate>
            <atom:updated>2023-01-13T03:40:52.461Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/59/1*azN82N4UtFY33evrEKia6w.png" /></figure><p>It was the spring of 2021. Hope was in the air. The vaccine was released and people were excited to get back to a life of normalcy. However, COVID-19 still loomed, and people were hesitant to immediately revert back to their old ways.</p><p>At the time, my girlfriend and I just moved to Hermosa Beach (a welcome change from our tiny apartment in cramped NYC). One of our favorite activities was to try the local cuisine. After dining out a handful of times, we began to notice a trend — restaurants frequently took a long time to deliver us a check when we were ready to leave. We hypothesized what the reason might be: limited and less experienced staff (due to COVID-19), laziness, an optimistic staff thinking we would order more food if they waited long enough. Regardless of the reason, we knew there had to be a better way.</p><p>Side note: if you ever find yourself in Hermosa Beach our personal favorite restaurant was a cozy local Italian restaurant called Fritto Misto on Pier Ave. We recommend you come hungry as the portions are gigantic.</p><p>When talking to my parents over the phone one day, they said they just visited a restaurant in Marin County that had installed QR codes to order. They said it was as simple as scanning a code at their table and ordering. I thought, “Wait a minute. All the restaurants around here could use this — in fact they <em>need</em> to be using this.” I thought this could solve 2 problems in one — faster ordering and make the experience touch-free (which was important in the times of COVID).</p><p>I told the idea to my girlfriend and she was immediately in. Our business was born.</p><p><strong>How did we build it?</strong></p><p>Not being developers ourselves, we used Upwork.com to find developers to help build our app. We were courted by many developers across the globe. We met a 23 year old from Ukraine who said he could build the whole app himself, extremely quickly, and charge us a fair price. We were skeptical. Eventually we decided to go with a group of developers from India. We pitched them our idea. They congratulated us on such a novel concept and assured us our idea is destined for greatness. We were pleased.</p><p><strong>What did the app do?</strong></p><p>The app made it easy for restaurants to build out electronic menus that could be shown on a customer’s phone after scanning a QR code. This way, customers could order and pay for their food ahead of time, eliminating the hassle of having to wait for the check when the meal is done.</p><p>We called the company dinz — a short, fun-to-say name that kinda resembles a “speedy dinner” if you squint at it long enough. We decided to get the URL <a href="https://www.dinz.app/">dinz.app</a> (it’s still running today if you like to visit it, at least until September 2023). I had my highschool friend build out the website. This turned out to be one of the best “hire for work” decisions we made. He did a great job for a very reasonable price. If you are looking for a site developer I highly recommend him, and happy to make an introduction if interested.</p><p>The website was connected to the app itself. Therefore after restaurants registered through the website (and agreed to pay the $5/month fee), they would be given access to use the app itself.</p><p>With high hopes, we walked to our local Chase bank and opened up a business account, complete with a fresh new Chase credit card. We celebrated with drinks that night.</p><p><strong>Drama Ensued</strong></p><p>Unfortunately, the good times did not last. QR code menus began popping up at restaurants all across our neighborhood and in broader Los Angeles powered by Toast and Square. Their systems were robust and connected to other systems throughout the restaurant — the display in the kitchen, the point-of-sale system.</p><p>We knew we couldn’t compete, so we quickly started to brainstorm how our app could be used in other ways. Food stands at baseball stadiums. Restaurants at airports. Little league snack shacks. We were dead set on finding the right setting that would have plenty to gain from having its customers order from a QR code, but also would face little competition.</p><p>Eventually we landed on QR codes for hotels. We figured hotels could have many use cases for QR Codes (and not just limited to ordering food). Hotels could use our app to build out a QR code menu at the hotel bar or for room service. They could also use our app to easily post popular workout routines at their hotel gym, or have a QR code stationed in the bedroom to make it easy to order another towel or cot. Based on this, the pivot to hotels was made.</p><p>While initially optimistic, after a few key speed bumps, things quickly took a turn for the worse.</p><p><strong>How the Business Failed</strong></p><p>The business eventually failed from a lack of customers, well 0 customers to be precise.</p><p>Developers were slower to meet deadlines than advertised (who knew India had so many holidays?). After a month of working, the developers were difficult to reach, and went through a few one or two week stretches where zero work got done. This prevented us from pivoting quickly to our new hotel idea and eventually made us settle for a product that <em>barely </em>worked.</p><p><strong>What Would I Do Differently</strong></p><p>There are a few key things I would do differently. In retrospect these reflections seem basic, but still a good learning experience.</p><ul><li><strong>Talk to more customers face to face:</strong> I spent time calling and emailing many potential customers. In hindsight, my time would have been better spent actually visiting and talking to our customers at their storefront or hotel.</li><li><strong>Build and record a solid demo: </strong>We should have built a solid demo to “WOW” potential customers and help land our first customer.</li><li><strong>Understand the competitive landscape:</strong> While I didn’t initially actually <em>see </em>other QR codes at restaurants, if I had done a little more research upfront, I would have discovered that Toast and Square were building out a similar product that would get released soon, and probably would not have rushed into the idea.</li><li><strong>Understand how our product interacted in a larger ecosystem: </strong>While ordering was a piece of the puzzle for restaurants. I should have understood how important it was to have this order system be connected to other systems throughout the restaurant (ie. kitchen display, point-of-sale system).</li><li><strong>Do something that you yourself can contribute to and build on: </strong>I think this is especially important in the early days of a company where you’d much rather trade time for saved money. In our case of building out an app we should have been able to write the code ourselves (or find a co-founder that could) instead of relying on a 3rd party for help. This also gives you more flexibility and the freedom to change course if needed.</li></ul><p>Overall, I learned a lot from the experience and I’m happy I did it. I’d like to try to start a business again someday. If you have a great idea, please feel free to drop me a line!</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=4ca3733b4858" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[SQL Roadmap for Analytical Roles Interviews]]></title>
            <link>https://medium.com/@abeckerman002/sql-roadmap-for-analytical-roles-interviews-d475581e02fd?source=rss-1ce2583f4c85------2</link>
            <guid isPermaLink="false">https://medium.com/p/d475581e02fd</guid>
            <category><![CDATA[analytics]]></category>
            <category><![CDATA[sql-interview-questions]]></category>
            <category><![CDATA[marketing-analytics]]></category>
            <category><![CDATA[sales-strategy]]></category>
            <category><![CDATA[sql]]></category>
            <dc:creator><![CDATA[Jack Smith]]></dc:creator>
            <pubDate>Sun, 08 Jan 2023 23:11:29 GMT</pubDate>
            <atom:updated>2023-01-14T16:34:16.063Z</atom:updated>
            <content:encoded><![CDATA[<p>In this post, I provide a roadmap for learning SQL. After spending countless years grudgingly combing through many SQL learning resources I’m excited to share with you what I think is the best way to learn SQL. I should also note I learned SQL in the context of landing a data analytics supporting business type role (marketing analytics/sales strategy), so if you’re interested in getting into that field this post will be especially helpful!</p><p>This post will focus on: crushing your SQL interviews. For one, landing a job is important and a good reason to learn a new skill. Secondly, I actually think SQL interviews is where your SQL skills are truly tested, so that’s what I’m going to focus on.</p><p>Personally, I’ve used many many SQL resources when trying to learn SQL including online sites, YouTube tutorials, books, and even attended SQL classes. I curated a learning plan for you with the best of the best resources for learning SQL quickly and will cost you less than $25.</p><p>I summarized all my recommendations in a learning roadmap <a href="https://docs.google.com/presentation/d/1h09rzp8T00IsOp4RoRmKELjLmjqRTlUS-Hm0OwekdZY/edit#slide=id.g1cb076f5a8d_0_10">slide</a>.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*f_Val44Yap34aN7cwMFZBw.png" /></figure><p>My first recommendation is the book <a href="https://www.amazon.com/SQL-Minutes-Sams-Teach-Yourself/dp/0135182794/ref=pd_lpo_1?pd_rd_w=rNuI5&amp;content-id=amzn1.sym.116f529c-aa4d-4763-b2b6-4d614ec7dc00&amp;pf_rd_p=116f529c-aa4d-4763-b2b6-4d614ec7dc00&amp;pf_rd_r=DD1PDFYASE1Q1DM4NJFB&amp;pd_rd_wg=N7E9R&amp;pd_rd_r=b4442c35-1e0d-4234-93a0-bb96b22dc96b&amp;pd_rd_i=0135182794&amp;psc=1"><em>SQL in 10 minutes</em></a> by Ben Forta.</p><p>The book is available on amazon. It costs $24. The book packs a ton of value in a very concise format and is absolutely worth the price. You can carry the small book around with you — read it on your commute (or in a hammock). It’s a super quick read and has a lot of great content and will provide you with an excellent SQL foundation.</p><p><a href="https://www.amazon.com/SQL-Minutes-Sams-Teach-Yourself/dp/0135182794/ref=pd_lpo_1?pd_rd_w=rNuI5&amp;content-id=amzn1.sym.116f529c-aa4d-4763-b2b6-4d614ec7dc00&amp;pf_rd_p=116f529c-aa4d-4763-b2b6-4d614ec7dc00&amp;pf_rd_r=DD1PDFYASE1Q1DM4NJFB&amp;pd_rd_wg=N7E9R&amp;pd_rd_r=b4442c35-1e0d-4234-93a0-bb96b22dc96b&amp;pd_rd_i=0135182794&amp;psc=1"><em>SQL in 10 minutes</em></a> was the first SQL book I owned, and it really does a fantastic job of covering all the SQL basics. It provides a background on databases, why the SQL language is so popular today, and an overview of the basic SQL commands and ideas.</p><p>Some SQL concepts you are going to want to take away from this book are:</p><ul><li>Select statements</li><li>Where clauses (including how to use LIKE)</li><li>Order by statements</li><li>Aggregate Functions</li><li>Grouping Data</li><li>Working with Sub-Queries</li><li>Joining Data</li></ul><p>I recommend reading the book from Chapter 1 through Chapter 14. After Chapter 14, the book focuses more on configuring databases like inserting, and deleting rows. This is not relevant to know for someone pursuing a marketing analytics or sales strategy role, so I recommend skipping over these chapters if that is your goal.</p><p>There are two topics that aren’t covered in this book that you’ll want to know to ace your SQL interviews. And one of them is “Window Functions”. For me, this topic came up in all of my SQL interviews. Specifically, I think it is good to understand the below Window Functions concepts:</p><ul><li>RANK()</li><li>DENSE_RANK()</li><li>ROW_NUMBER()</li><li>Using “Over ()” and “Partition by” and “Order By” properly</li></ul><p>This <a href="https://www.geeksforgeeks.org/window-functions-in-sql/">article </a>does a great job summarizing what these concepts are, there are also a lot of helpful YouTube tutorials online as well. Window Functions are more confusing than basic SQL commands, so I recommend devoting at least a few hours in learning this concept to give yourself a chance to thoroughly go over this topic and understand it.</p><p>Secondly, I recommend learning how to use CTEs (Common Table Expressions). This <a href="https://learnsql.com/blog/what-is-common-table-expression/">article </a>helps. CTEs are also known as temporary tables. These tables are very valuable and can transform even the most complex SQL questions into a series of simple SQL tables. The use of these tables also help break down a problem and make it easy to explain the solution to your interviewer. Learning CTEs is a must!</p><p>Okay great! You learned what you needed and now it’s time to practice!</p><p>I visited many websites looking for a good site to practice SQL (it is surprisingly difficult to find a good one). After exploring many sites my go-to site became <a href="https://www.stratascratch.com/">stratascratch.com</a>. Personally, I love this site. It gives excellent instructional problems that are realistic in an interview setting. Part of practicing is making sure you don’t get tripped up on the day of the interview. It’s never fun seeing a question for the first time in an interview. To avoid this letdown make sure you practice, practice, practice!</p><p>Some tips for using the site:</p><ul><li><strong>Start with the free version.</strong> I eventually signed up for the paid version for this site for 2 months. It was $32 a month. It was nice to have access to more problems, but I don’t think it’s necessary. My recommendation is to do all the free problems before considering the paid version.</li><li><strong>Use the filters. </strong>This will get you more targeted questions based on your needs. Some of the top companies actually submit questions to the site. Check out the “Company” filter to see if the company you’re interviewing at submitted questions to the site. The “Topics” filter is also helpful to hone your skills for a particular topic.</li><li><strong>Review the discussions and comments</strong>. This will show you how other people solved the problem. Part of SQL interviews is being able to accurately explain the topic to the interviewer (understand the “why” behind the problems). Understanding how others solved the same problem really helps with this.</li><li><strong>Don’t be afraid to experiment!</strong> Try to solve problems multiple ways. Also, test out different commands in the SQL input window to get familiar with what data gets returned.</li><li><strong>Go through the easy and medium questions first.</strong> The hard problems can be time consuming are there many valuable concepts that can be learned by tackling the easy and medium level difficulty questions first.</li></ul><p>After completing the free and paid practice problems on <a href="https://www.stratascratch.com/">stratascratch.com</a>, if you’re still craving more practice problems, I recommend checking out <a href="https://leetcode.com/">leetcode.com</a>. This site is more popular for traditional programming questions but it has many SQL/database practice problems as well. This site requires a subscription, which I believe is $30 a month.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/225/1*TNyVHHOAyisjhTzOv5huiA.jpeg" /></figure><p>The last thing I want to call out is SQL classes. I’ve personally taken 2 SQL classes — one at UC Berkeley and another through a career type program. I didn’t get a lot of value out of these classes. They tend to focus on more beginning level SQL concepts, and by the time “new” concepts are introduced the class is 75% done. These classes can also be expensive. Again this is just my experience — I think SQL is intuitive enough that the books and online resources are sufficient and will allow you to choose your own learning plan (<em>hint follow the learning plan outlined in this post</em>).</p><p>Interview Tips:</p><ul><li>Review a <a href="https://learnsql.com/blog/sql-basics-cheat-sheet/">SQL Cheat Sheet</a> a day or two before the day of interview to quickly refresh concepts before an interview.</li><li>Do at least 2–3 practice problems (even if you already completed them before) to get you in the right headspace for working on SQL problems. Also practice saying out loud why you’re coding in a certain way. Being able to verbally communicate to your interviewer <em>why </em>you’re coding in a certain way is sometimes just as valuable as <em>what</em> you’re coding.</li></ul><p>So to quickly recap, start with reading the <a href="https://www.amazon.com/SQL-Minutes-Sams-Teach-Yourself/dp/0135182794/ref=pd_lpo_1?pd_rd_w=rNuI5&amp;content-id=amzn1.sym.116f529c-aa4d-4763-b2b6-4d614ec7dc00&amp;pf_rd_p=116f529c-aa4d-4763-b2b6-4d614ec7dc00&amp;pf_rd_r=DD1PDFYASE1Q1DM4NJFB&amp;pd_rd_wg=N7E9R&amp;pd_rd_r=b4442c35-1e0d-4234-93a0-bb96b22dc96b&amp;pd_rd_i=0135182794&amp;psc=1"><em>SQL in 10 minutes</em></a> book to understand the SQL basics, then study and understand window functions and CTEs, then use the free version of <a href="https://www.stratascratch.com/">stratascratch.com</a> to practice.</p><p>If you feel like you need more practice, get the paid version of <a href="https://www.stratascratch.com/">stratascratch.com</a>, and then if you want even more practice, sign up for <a href="https://leetcode.com/">leetcode.com</a>.</p><p>With this SQL interview plan in place, you will crush your SQL interview!</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=d475581e02fd" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[UC Berkeley MIDS (Master in Data Science) Review]]></title>
            <link>https://medium.com/@abeckerman002/uc-berkeley-mids-master-in-data-science-review-f8d104e8832a?source=rss-1ce2583f4c85------2</link>
            <guid isPermaLink="false">https://medium.com/p/f8d104e8832a</guid>
            <category><![CDATA[analytics]]></category>
            <category><![CDATA[machine-learning]]></category>
            <category><![CDATA[uc-berkeley]]></category>
            <category><![CDATA[masters-in-data-science]]></category>
            <category><![CDATA[data-science]]></category>
            <dc:creator><![CDATA[Jack Smith]]></dc:creator>
            <pubDate>Sun, 08 Jan 2023 18:19:58 GMT</pubDate>
            <atom:updated>2023-01-14T16:37:17.559Z</atom:updated>
            <content:encoded><![CDATA[<p>This is a personal review of <a href="https://ischoolonline.berkeley.edu/data-science">UC Berkeley’s MIDS program </a>, which I am currently enrolled in. If you’re considering applying or attending the program, this post is for you!</p><p>For context, I was working full time in the technology industry for most of my time in the program. Some logistics about the program, the program offers 14 courses to choose from. A student needs to complete 8 of these courses in addition to a required capstone project to achieve a degree. The entire program costs about $70k in total. At this price point, this program is one of the more expensive online data science master programs.</p><p>The program is all online. At first this made me apprehensive. I found myself asking questions like <em>will I actually be able to learn online?</em> <em>Will I be able to access the instructors and students for help?</em> However, as the program progressed these worries quickly faded. I actually came to appreciate having the program online and saw it as a benefit. By having the program online, this allowed for learning around my schedule and the additional benefit of getting to interact with students across the US and across the globe. The program does offer an <a href="https://www.ischool.berkeley.edu/programs/5th-year-mids/curriculum/immersion">immersion experience</a>, where 3 times a year students get the opportunity to go to the UC Berkeley campus and meet faculty and students. I haven’t done this yet, but I am looking forward to participating in that this summer.</p><p>Students take between 1–3 classes a semester with the standard being 2 classes a semester. The program also allows students to vary the number of classes you choose to take semester by semester. For example, some semesters I took 2 classes a semester and other semesters, when I felt like my responsibilities outside of class were more demanding, I decided to take only 1 class a semester.</p><p>In terms of workload, the average workload is ~10–15 hours of work per class a week. The number of hours certainly varies by class — there are easier, less time intensive, courses and other classes, which are more time intensive.</p><p>The classes are structured so they are ~14 weeks long and lectures are once a week with about 12–18 students in each lecture. Students watch pre-recorded videos and do readings outside of lecture. During lecture, learnings from the videos and readings are discussed. Each course typically has a slack channel, which allows you to communicate with the other students and instructors throughout the week.</p><p>In terms of job prospects and salary, I found this <a href="https://fortune.com/education/articles/grads-of-uc-berkeleys-data-science-masters-program-can-earn-more-than-155k/">Fortune article</a> interesting which came out in late July 2021. A short excerpt below:</p><blockquote><em>In the U.S., 2021 graduates from Berkeley’s master of information and data science (MIDS) program earned an average annual salary of more than $155,000, according to data provided by the school. The program is ranked №2 among the </em><a href="https://fortune.com/education/information-technology/masters/rankings/best-online-masters-in-data-science/"><em>best online master’s in data science degree programs</em></a><em> in 2022 by Fortune.</em></blockquote><blockquote><a href="https://fortune.com/education/business/mba/rankings/university-of-california-berkeley-haas/"><em>Berkeley’s Haas School of Business</em></a><em> is also recognized as a top-ranked business school, and it landed the №13 spot on Fortune’s ranking of the </em><a href="https://fortune.com/education/business/mba/rankings/best-mba-programs/"><em>best MBA programs</em></a><em>. Haas graduates also report high salaries, with an average annual base salary of nearly $144,000, </em><a href="https://mba.haas.berkeley.edu/careers/employment-report#Overview"><em>according to the school’s 2021 employment report</em></a><em>. Of note: The average salary for Haas MBA graduates is almost $12,000 less than that of MIDS graduates.</em></blockquote><p>In 2021, MIDS graduates had a higher salary on average than Berkeley Haas students. While it’s only one year of data, I think it speaks to the quality of the program and the thriving job market for data scientists.</p><p>The program is what you make of it and part of that is choosing what courses to take. Below I share some of the classes that I found most helpful in landing a data science job:</p><p><strong>Introduction to data science programming:</strong> This is essentially an introduction to python course. Personally, I feel like I’ve learned a ton in this class. I had very little python experience beforehand, and I feel like this course got me to an intermediate level. I’m now comfortable with all basic/intermediate python concepts and also very comfortable with common data science packages like NumPy and pandas.</p><p><strong>Statistics for Data Science:</strong> This class covers typical statistics topics like probability, bayes theorem, hypothesis testing, central limit theorem, sampling techniques but goes deeper into the math than my previous statistics classes. Overall, I feel like this course prepared me well for a stats portion of a data science interview.</p><p><strong>Machine Learning at Scale:</strong> This is by far the most time intensive class that I took in the program — this class averaged ~25 hours a week. Although, this is also the class that I got the most value from. The class covers Big Data topics like Hadoop, MapReduce framework, PySpark, graph and gradient descent algorithms. The class also offers a free license to Databricks, which is great for getting on-hand experience. I personally spoke with 2 students in the program that got data science jobs and both attribute to their success to the skills they learned in this course . So if your goal is to land a data science job, you don’t want to pass this class up!</p><p>The technical tools I learned are:</p><ul><li>Python</li><li>R</li><li>Pyspark</li><li>Tableau</li><li>Latex</li></ul><p>The statistics classes used R but all other classes used python.</p><p>Another huge value add from the program is getting to interact and build relationships with the other students. I feel like the students in this program are truly top notch. Most of the students had impressive backgrounds prior to enrolling in the program. Whether it was working as an engineer at a top tech company or having a masters or PhD in a different subject, the students I had the chance to meet are very well accomplished. In addition to the students generally being very successful, they are all extremely friendly and very willingly to help — something I was grateful for. I learned a lot from other students, and I think this is a huge value add of the program.</p><p>The program also offers access to slack channels where you can get career advice, find students posting jobs and internships at their respective companies, and awareness of job fairs. There is also something called “MIDS for Life”, where as an alumni, you get lifelong access to the course material after you graduate.</p><p>Overall, despite the program being pricey, I think this is an excellent program if you’re hoping to break into the data science field. I’m glad I did it, and I think I got a lot of value out of it. Also, I have been able to make at least 30 connections that I would feel comfortable reaching out for a job referral. The program will give you everything you need to get started on your data science career path and provide you with an excellent foundation for data science interviews.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=f8d104e8832a" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>