<?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 Helen Mou on Medium]]></title>
        <description><![CDATA[Stories by Helen Mou on Medium]]></description>
        <link>https://medium.com/@helenmou?source=rss-c726795fa50d------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/2*3UPjhptEFNPw00AUJPXtkA.jpeg</url>
            <title>Stories by Helen Mou on Medium</title>
            <link>https://medium.com/@helenmou?source=rss-c726795fa50d------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Fri, 20 Nov 2026 04:17:44 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@helenmou/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[Setting a foundation for your PM career in the first 6 months]]></title>
            <link>https://medium.com/product-at-shopify/setting-a-foundation-for-your-pm-career-in-the-first-6-months-b5e67c6d3f45?source=rss-c726795fa50d------2</link>
            <guid isPermaLink="false">https://medium.com/p/b5e67c6d3f45</guid>
            <category><![CDATA[career-advice]]></category>
            <category><![CDATA[new-job]]></category>
            <category><![CDATA[apm]]></category>
            <category><![CDATA[shopify]]></category>
            <category><![CDATA[product-management]]></category>
            <dc:creator><![CDATA[Helen Mou]]></dc:creator>
            <pubDate>Wed, 02 Feb 2022 05:01:51 GMT</pubDate>
            <atom:updated>2022-04-12T14:42:37.693Z</atom:updated>
            <content:encoded><![CDATA[<p>Three key areas for first-time PMs to focus on in their first 6 months</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/925/1*Y7LMrLxuLQPBEDXi7Vasww.jpeg" /><figcaption>Starting your very first product management job can be both exciting and overwhelming</figcaption></figure><p>Starting your very first product management job?</p><p>Whether you’re joining a rotational Apprentice Product Manager (APM) program or starting a Product Manager role, it’s both an exciting and potentially overwhelming time.</p><p>In the first 6 months, you’re likely facing a firehose of information, meet-and-greets, and skill-building sessions. The goal is to set a strong foundation for your PM career: this means building your execution experience, building domain expertise, and a roster of cross-functional teammates that trust you.</p><p>Both managers and first-time PMs have a role to play in setting a strong foundation for a product management career that could potentially span decades. For managers, it’s a big responsibility to be supporting someone at the very beginning of their PM journey. For first-time PMs, it can be challenging to maintain and drive an an ongoing career conversation with your manager while you’re drinking from the firehose.</p><p>Here are the three key areas for first-time PMs to focus on in their first 6 months:</p><p><strong>1. Learn on the job: Get execution reps and don’t worry about how to lead a giant/ambiguous/strategic project. Start with projects that are already in the “build” phase.</strong></p><p>Projects that are in the “build” phase will already be resourced and prototyped. Jump in at this point, and you will quickly learn what it takes to ship something, keep stakeholders up-to-date, and ask good questions. You’ll also learn how to evaluate that thing once it ships (through something like a post-launch analysis).</p><p>Since you will likely be the least experienced person from any discipline, you will learn what it’s like to lead from behind—I like to think of this as servant leadership. In a <a href="https://twitter.com/tobi/status/1263483496087064579?lang=en">digital by default</a> (i.e. all-remote) environment, it’s the cherry on top that you can get help from your manager or mentor if you don’t know what to say. Don’t expect that you’ve “found your PM voice” yet — it takes many, many reps to find that.</p><p><strong>2. Actively manage your project portfolio: Make sure there’s always at least 1 project that is 100% under your control.</strong></p><p>6 months is a very short window of time, and you want to ensure that you have impactful achievements when you go to write your self assessment. Work with your manager to ensure there is at least 1 project that is 100% under your control — that means that there are minimal outside forces that will derail the project from being completed as expected. It might have less immediate user impact but would help the team make better decisions, like a competitive landscape analysis or an analysis of merchant feedback. Or it might be super tactical — like improving your help docs, going through the product and identifying bugs or content improvements, or jumping into the support channel or issue backlog.</p><p>For every project that you work on, write down your deliverables and learning outcomes, and ask your manager to share relevant context and resources (I use a template — <a href="https://medium.com/subscribe/@helenmou">subscribe</a> and comment below for a link). Do this at the beginning, middle, and end of working on that project. The dedicated time to write down learning outcomes at each phase of the project will pay dividends when similar challenges arise on future projects. Plus, you can bring this with you to your next rotation or use it for your self-evaluation.</p><p><strong>3. Expose your craft: It might seem scary at first, but try not to operate in the safe world of direct messages (DMs).</strong></p><p>Try to do your work in the project and team channels. This way, you expose your work (and your thought process) to others. Those people are going to become advocates for you when your lead asks 3–5 people for feedback at the end of your rotation or during your review.</p><p>There are definitely times when DMs are appropriate! But, when you’re doing your product work, chances are that your questions and work-in-progress snippets are relevant to things that others are doing. At a minimum, you’ll instill confidence in your teammates that you are making progress and thinking critically about what they’re building. Operating in the “safe” world of DMs will limit your impact. If you’re doing this now, you can start small by first noticing how often you use a DM instead of the project or team channel, and work up to it over time.</p><p>Don’t worry about sitting in on “big meetings,” working on 1 big project from end-to-end, or being the person who knows the answer to every single question. These things will come (except the last one!).</p><p>In the first 6 months, first-time PMs need to build a foundation of that nitty-gritty execution experience, relevant domain expertise, and a roster of cross-functional teammates that trust you. To do that, you will need to learn on the job, actively manage your project portfolio, and expose your craft.</p><p>What did you do during your first 6 months as a product manager to set yourself up for a successful PM career? Would love to hear your thoughts in the comments below.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=b5e67c6d3f45" width="1" height="1" alt=""><hr><p><a href="https://medium.com/product-at-shopify/setting-a-foundation-for-your-pm-career-in-the-first-6-months-b5e67c6d3f45">Setting a foundation for your PM career in the first 6 months</a> was originally published in <a href="https://medium.com/product-at-shopify">Product @ Shopify</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Ten books in 2021]]></title>
            <link>https://medium.com/@helenmou/ten-books-in-2021-ee9681795f88?source=rss-c726795fa50d------2</link>
            <guid isPermaLink="false">https://medium.com/p/ee9681795f88</guid>
            <category><![CDATA[books]]></category>
            <category><![CDATA[reflections]]></category>
            <dc:creator><![CDATA[Helen Mou]]></dc:creator>
            <pubDate>Fri, 31 Dec 2021 23:25:02 GMT</pubDate>
            <atom:updated>2021-12-31T23:25:02.065Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/924/1*iHmEHGG_wOjsWMaoD5XL2A.jpeg" /></figure><p>Ten memorable books from 2021, in no particular order:</p><p><a href="https://www.npr.org/2021/02/06/964567998/a-journalists-eye-enlivens-land-of-big-numbers">Land of Big Numbers</a> by Te-Ping Chen. I like reading books by journalists because of the combination of narrative development and investigative approach — this one is extra special because it’s fiction. Short stories with a dose of social commentary that helped me better understand today’s China. Call center culture, professional video game players, a noodle robot.</p><p><a href="https://www.greenapplebooks.com/book/9780374159276">Fulfillment: Winning and Losing in 1-Click America</a> by Alec MacGillis. Through the lens of e-commerce fulfillment networks, MacGillis explores the role of personally and financially “fulfilling” jobs that have become increasingly difficult to obtain. The book dissects this notion of fulfillment by highlighting the bifurcation of American cities and towns: top-tier cities in a virtuous cycle of prestige and growth (e.g. San Francisco, New York City) vs. “2nd tier” cities and towns (e.g. Baltimore, Dayton) which are trapped in a vicious cycle of low-wage jobs, loss of tax income, and limited opportunity for individual fulfillment.</p><p><a href="https://www.greenapplebooks.com/book/9781982127213">Jackpot: How the Super-Rich Really Live — and How Their Wealth Harms Us All</a> by Michael Mechanic: Dives into the world of the 0.001% (concierge healthcare, nannies trained in combat, bathtubs that cost as much as a car). Many of those interviewed are from the Bay Area. Shows us how the virtuous cycle of the rich getting richer isn’t always so virtuous — it comes at a cost.</p><p><a href="https://www.greenapplebooks.com/book/9780061358302">Tales of the City</a> by Armisted Maupin. A very SF story about a transplant finding her way. I had scheduled a “literary tour of North Beach” for visiting family members and wanted to read something that drew on SF’s rich history. Originally published as regular installments in the San Francisco Chronicle, starting in 1978, and now the inspiration behind the Netflix series. (Bonus: the 2nd book, “More Tales of the City,” was also a great read.)</p><p><a href="https://www.greenapplebooks.com/book/9780593318171">Klara and the Sun</a> by Kazuo Ishiguro: A dystopian science fiction novel showing how artificial intelligence could fit into our lives, and our children’s lives. Powerful, not so far off, and a page-turner. This put a lot into perspective, especially given my experiences building artificial intelligence and machine learning into technology products. Reminded me of Haruki Murakami, whose novels I’ve read extensively over the last decade, starting with 1Q84. I’ll be reading more Ishiguro next year.</p><p><a href="https://www.greenapplebooks.com/book/9780679763888">The Warmth of Other Suns: The Epic Story of America’s Great Migration</a> by Isabel Wilkerson. Part journalism, part storytelling. This book was published in 2010, I read it 11 years later. That’s a miss, because this book has helped me immensely in understanding the roots of our country’s inequality today. I definitely recall learning about the Gold Rush and the Dustbowl Migration in history class, but not the Great Migration.</p><p><a href="https://www.greenapplebooks.com/book/9780062060624">The Song of Achilles</a> by Madeline Miller. I picked this up after reading her 3rd book, <a href="https://www.greenapplebooks.com/book/9780316556323">Circe</a>, last year. I loved learning Latin in high school — the combination of the deterministic structure and the lyrical tradition — and reading, translating, and memorizing The Odyssey. This book gave me another perspective on Odysseus, as a cunningly strategic advisor to Agamemmnon. I also enjoyed reading about the journey of Patroclus, the outsider-narrator, from cast-aside to a physician and partner to Achilles.</p><p><a href="https://www.greenapplebooks.com/book/9781984881755">The Family Firm: A Data-Driven Guide to Better Decision Making in the Early School Years (The ParentData Series #3)</a> by Emily Oster. I waited all year for this. Oster is a professor of economics at Brown. I never crossed paths with her as an undergraduate concentrating (majoring) in economics. Her examination of the actual studies behind conventional wisdom such as eating sushi during pregnancy or the importance of family dinner makes it obvious which “old wives tales” are based on high quality evidence, and where you should use your judgement and personal values. It also makes the lack of scientific priority on women and families very clear.</p><p><a href="https://www.greenapplebooks.com/book/9780062969743">The Cold Start Problem</a> by Andrew Chen. Before I even read the book, I learned a lot just by watching Andrew applying his own growth strategy to his book launch. This book reminded me that it’s not enough to build a killer set of features (<a href="https://andrewchen.com/the-next-feature-fallacy-the-fallacy-that-the-next-new-feature-will-suddenly-make-people-use-your-product/">a la The Next Feature Fallacy</a>), and that the 80/20 rule is everywhere (a minority of users will create disproportionate value and therefore have disproportionate power). The surprise touch for me was the discussion about the <a href="https://en.wikipedia.org/wiki/Allee_effect">Allee effect</a> — something I remember from my Environmental Studies class.</p><p><a href="https://www.greenapplebooks.com/book/9781938298073">What do you do with an Idea?</a> By Kobi Yamada and Mae Besom. Gorgeous illustrations. A classically inspirational story with something resonant for both the adults and kids in our family. A thoughtful gift, from a friend who knows me well.</p><p>I buy all my books from <a href="https://www.greenapplebooks.com/">Green Apple Books</a> and <a href="https://www.foliosf.com/">Folio Books</a> (message them! They’re so responsive and fast), and borrow digital copies from the <a href="https://sfpl.org/">San Francisco Public Library</a>.</p><p>Currently reading <a href="https://www.greenapplebooks.com/book/9781501192852">Mom Genes</a>, which I think will make it to next year’s top 10 list.</p><p>What was in your top 10?</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=ee9681795f88" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Leveling up AI-literacy for PMs]]></title>
            <link>https://medium.com/data-science/leveling-up-ai-literacy-for-pms-dc293b79fe17?source=rss-c726795fa50d------2</link>
            <guid isPermaLink="false">https://medium.com/p/dc293b79fe17</guid>
            <category><![CDATA[data-science]]></category>
            <category><![CDATA[machine-learning]]></category>
            <category><![CDATA[product-management]]></category>
            <category><![CDATA[towards-data-science]]></category>
            <category><![CDATA[artificial-intelligence]]></category>
            <dc:creator><![CDATA[Helen Mou]]></dc:creator>
            <pubDate>Tue, 04 Dec 2018 14:36:00 GMT</pubDate>
            <atom:updated>2018-12-12T13:23:05.917Z</atom:updated>
            <content:encoded><![CDATA[<h4>…and why I’m teaching Data Science for Product Managers</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*WfPCyy_9xi_EDZsd1QEHow.jpeg" /><figcaption>AI isn’t magic.</figcaption></figure><p>Today more than ever, data is at the core of product management. Product managers are increasingly responsible for shipping machine learning-driven product features, making critical product decisions based on machine learning techniques, and developing strong partnerships with their data science counterparts.</p><p>But when I asked data scientists and machine learning engineers about how their PM counterparts are doing, here is what I heard:</p><blockquote>“My PM [doesn’t] know what questions to ask.”</blockquote><blockquote>The PM is “unable to frame what they are trying to solve as an ML problem.”</blockquote><blockquote>“PMs and executives are overconfident in their understanding, and overconfident in what is possible/easy.”</blockquote><blockquote>“Admit when you don’t understand something. Don’t pretend you know more than you do. Don’t make promises to the business without talking to the engineers.”</blockquote><p>Too many organizations are hiring teams of data scientists without simultaneously investing in the data and AI-literacy of everyone else. As a result, there are too many product managers out there who are driving data scientists crazy (and vice versa, too).</p><p>Here’s why product managers need to become AI-literate:</p><p><strong>❄️ 7 billion special snowflakes: Your users demand it. ️️❄️</strong></p><p>The cat’s out of the bag — every individual wants to feel like (wait for it…) an <em>individual</em>. You, me, and all of our users demand intelligent, personalized experiences that make us feel like the special snowflakes that we are. If everyone else is getting the same thing as me, what’s so exciting about that?</p><p>When users give data to companies by registering an account, engaging with the application, and providing feedback, they expect the product to continuously improve and become more intelligent over time. These improvements often require productizing the data generated by the user’s engagement with the product itself.</p><p>AI-enabled products are all around us. Spotify uses your listening history to create personalized playlists, Gmail uses previous email content to finish your sentences, <a href="https://woebot.io/">Woebot</a> remembers past conversations in order to give you a personal therapy session, and <a href="https://get.intelligentx.ai/">IntelligentX’s</a> AI algorithm uses your feedback from the previous batch of beer to create a recipe tailored just for your tastebuds. Yes, that’s right, you can now purchase <a href="https://get.intelligentx.ai/">beer that is brewed by AI</a>.</p><p>Even if it’s just marketing hype, the fact that we are surrounded by all of these exciting developments is the very thing that endlessly increases our expectations and our “needs” (or wants). But for product managers, it’s not just hype — it’s your job to deliver on these expectations.</p><p><strong>🥣 The Goldilocks PM: Not too technical, not too much hype 🥣</strong></p><p>The job of a PM encompasses envisioning what future products are possible, and delivering what is valuable. In order to do this, PMs need to have a <em>just-right </em>amount of<em> </em>knowledge about the current state of data science technologies. This Goldilocks-level understanding enables PMs to bridge the gaps between teams of different expertise — business, technology, and design — without getting too caught up in the AI hype, or too deep into the weeds.</p><p>As a whole, the PM discipline is already well-versed in traditional software development technologies. This skillset is table stakes, across the board. When it comes to data science technologies, though, most PMs don’t have the same level of technical understanding as they already have in business, engineering, and design — and that’s because they haven’t always needed it.</p><p>Until now. As products become increasingly “intelligent” in order to meet the needs of each and every one of those special snowflake users out there, more and more of the application’s decision points will be taken over by predictive models, rather than by heuristics and explicit if-then-else logic.</p><p>Too often, PMs will drive their data science counterparts crazy by either getting too far into the technical weeds (less likely) or getting over-enthusiastic about hyped-up applications (more likely). The sweet spot is somewhere in between. A successful PM needs to understand the basics of data modeling, data infrastructure, machine learning models, and model evaluation metrics. Armed with just the right amount of knowledge, PMs can not only ship products that delight current users, but also deliver a vision of <em>what is possible</em> to create in the future.</p><p><strong>🔮“ML isn’t magic”: Product managers need to understand the ML development process in order to effectively collaborate with the data science team*. 🔮</strong></p><p>In order to build better partnerships (and products) with their data scientist and machine learning engineer counterparts, PMs need to understand major differences in the development process and where he or she can play the most impactful role.</p><p>Specifically, PMs can have the most impact in two areas: the model inputs and the model output. The “output” is the prediction that the ML model will generate. It’s critical that the PM clearly articulates the business value of this algorithm output, as well as the relevant constraints — great model predictions do not guarantee successful business results. The “inputs” are the features (aka attributes, independent variables, columns, etc.) that feed into the model. Garbage in, garbage out. Sharing domain knowledge and context with the data science team can help the team find stronger signal, leading to a higher-quality, more interpretable output.</p><p>As for the rest — step away and leave it to the black box of machine learning, right? Not so fast. As one data scientist said, “PMs need to know that ML isn’t magic.” Beyond the inputs and the outputs, there are many places in the development process where PMs can get tripped up unexpectedly. Here are some potential scenarios:</p><ul><li><strong>Interpretability</strong>: The model that generates personalized coupons sends out 75% off coupons to thousands of customers. The CFO notices gross profit is down this week and asks the PM why so many of these coupons went out, but no one can explain exactly what led the model to output this extremely generous coupon to so many customers.</li><li><strong>Model monitoring</strong>: A personalized video workout app launches with great fanfare and excitement. Three months later, a user writes into the support team wondering why the app recommended the same 15-minute abs routine five days in a row even though he had already completed it twice. The data science team, which had moved on to developing the next new feature, checks and realizes that the model performance has drifted downward.</li><li><strong>Data collection</strong>: The PM believes that working hours per week is a relevant feature input to predict whether a student will earn an online course certificate. However, this information is not part of the student’s profile and is therefore not available to include in the feature engineering process.</li></ul><p>Bias, data cleaning, model evaluation…suffice to say, the development process for machine learning products is full of pitfalls that can easily derail a carefully planned product roadmap. AI-literate PMs can more easily foresee and proactively address these potential scenarios — before they happen.</p><p>To address these challenges, <a href="https://www.usfca.edu/faculty/nicholas-ross">Professor Nick Ross</a> and I developed a 7-week certificate course at the <a href="https://www.usfca.edu/data-institute">University of San Francisco’s Data Institute</a> called <a href="https://www.usfca.edu/data-institute/certificates/data-science-product-managers">Data Science for Product Managers</a>. The certificate covers key data science and machine learning concepts, including analytics, statistics, regression, clustering, classification, NLP, recommendation systems, and deep learning — with a focus on product applications and business challenges.</p><p>By strengthening the AI literacy of PMs, we can build more effective partnerships between data scientists and product managers, and ship more intelligent products for all of the 7 billion special snowflakes out there.</p><p>* I am using “Data Science team” as an umbrella term to refer to the people who are <em>building the product</em>. This person’s title could be Machine Learning Engineer or Data Science Engineer (ideally), but is sometimes called simply “Data Scientist”. Most companies are still figuring out the best way to structure and title their data science teams.</p><p>⭐️ Thank you to my wonderful reviewers: Amelio Vazquez-Reina, Priya Mathew, Vincent Chio, Kona Shen, and Noah Fisher</p><p>📸 Photo by Sarah Trummer from Pexels.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=dc293b79fe17" width="1" height="1" alt=""><hr><p><a href="https://medium.com/data-science/leveling-up-ai-literacy-for-pms-dc293b79fe17">Leveling up AI-literacy for PMs</a> was originally published in <a href="https://medium.com/data-science">TDS Archive</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[3 ways Product Managers can leverage data]]></title>
            <link>https://medium.com/women-in-product/3-ways-product-managers-can-leverage-data-aaa8127190dc?source=rss-c726795fa50d------2</link>
            <guid isPermaLink="false">https://medium.com/p/aaa8127190dc</guid>
            <category><![CDATA[ecommerce]]></category>
            <category><![CDATA[leadership]]></category>
            <category><![CDATA[product-management]]></category>
            <category><![CDATA[machine-learning]]></category>
            <category><![CDATA[analytics]]></category>
            <dc:creator><![CDATA[Helen Mou]]></dc:creator>
            <pubDate>Fri, 17 Nov 2017 23:19:03 GMT</pubDate>
            <atom:updated>2017-11-17T23:19:03.619Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/812/1*o_h-MgNkqnCy3cQme2OByQ.jpeg" /><figcaption>Amplitude Analytics Product Summit, 11/07/17 | Santa Monica, CA</figcaption></figure><p>Last week, I spoke at a “fireside chat” at Amplitude’s Product Analytics Summit in Santa Monica, California.</p><p>The audience: Product Managers, Engineers, and Data professionals — a trifecta that I collaborate with day in and day out at Shopify.</p><p>Here are the three main takeaways (and the only fire) from my talk:</p><p>🔥 <strong>To influence decision-making, you need both ART and SCIENCE. 🔥</strong></p><p>That killer data point, that you spent hours and days digging through the data warehouse for? It’s not enough.</p><p>I often hear the question: “How can I use data to convince my leadership to give me the resources to build XYZ?”</p><p>Or this: “My company is not data-driven enough. I presented my dashboard with my killer data point last week, and then Executive X told an anecdotal story about a random customer interaction, and totally derailed my point!”</p><p>If you want to sway a room of executives, stakeholders, and your peers, then you need more than just a killer data point (the science).</p><p>You need a story (the art).</p><p>Storytelling is an art. Instead of jamming your charts and dashboards down everyone’s throat, help your stakeholders understand the underlying forces that have come together, from many different angles, to create a situation where data point A has come to be the case.</p><p>Instead, I think the right question to ask should be: “How can I own the entire narrative, and incorporate data to make it more powerful?”</p><p>🔥 <strong>To generate ROD, first define your unit of output. 🔥</strong></p><p>Data is an asset. And every asset should generate a “return on data” (ROD).</p><p>But how?</p><p>I work with my data team in two ways, depending on which data skill set I’m working with:</p><ol><li><strong>Product Analysis</strong>: The unit of output from a product analyst’s work is a <em>product roadmap decision, </em>e.g. reprioritize to build product feature D first, before product feature A.</li><li><strong>Data Science (a model)</strong>: The unit of output from a data scientist’s work is a <em>prediction</em>, paired with a <em>confidence score</em>.</li></ol><p>To help product analysts and data scientists do their job, I always make sure to provide three inputs:</p><p>Input 1. The business problem statement— what objective are we solving for</p><p>Input 2. Domain knowledge — is there some unique quality about the problem space that could have implications about analytical scope, feature engineering, etc.?</p><p>Input 3. Prioritization — how important is this output, compared to the other outputs that we have on our roadmap?</p><p>The nexus between product and data is increasingly important, but often overlooked. The more we understand each other, the more ROD we generate, and the more we will succeed with our customers.</p><p>Speaking of customers…</p><p>🔥 <strong>Your “North Star metric” = your CUSTOMER’s North Star. 🔥</strong></p><p>Without fail, on every product roadmap and strategy document, you will invariably arrive at a section called “KPIs”.</p><p>Sometimes, you will see product roadmaps with lists of 5–10 KPI metrics, or more. But we all know that multi-objective optimization is hard. It’s too hard to manage, and it puts your product at risk.</p><p>So, naturally, it’s important to define your “North Star metric” and stick with it, stick with it, stick with it.</p><p>Because you are accountable for your North Star metric, this will influence your overall product decisions, engineering solutions, and design strategy. And it will also influence all of the micro-decisions and marginal tradeoffs that happen everyday, both explicitly and implicitly, and ultimately become your product itself.</p><p>If your North Star is your CUSTOMER’s North Star, then you have a winning formula. Easy, right?</p><p>This requires deep thinking about the thing that deeply motivates and, in some cases, identifies your customer.</p><p>At Shopify, our North Star metric is each merchant’s Gross Merchandise Volume (GMV) — that’s it. Sales.</p><p>What should I write about next? Please comment below with A, B, C, or other.</p><p>A) <strong>Trusting the black box</strong>: How to implement machine learning models as a Product Manager</p><p>B) <strong>A few left turns</strong>: How I went from an Environmental Studies undergraduate major, to a Product Manager at a tech company</p><p>C) <strong>Belly of the beast</strong>: What I learned by moving to Bentonville, Arkansas and working at the Walmart Home Office</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=aaa8127190dc" width="1" height="1" alt=""><hr><p><a href="https://medium.com/women-in-product/3-ways-product-managers-can-leverage-data-aaa8127190dc">3 ways Product Managers can leverage data</a> was originally published in <a href="https://medium.com/women-in-product">Women In Product</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
    </channel>
</rss>