<?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 Catherine Narae on Medium]]></title>
        <description><![CDATA[Stories by Catherine Narae on Medium]]></description>
        <link>https://medium.com/@catherine.narae?source=rss-2d0313d2795c------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/0*oYTG6ZKLwUAB7Vyw</url>
            <title>Stories by Catherine Narae on Medium</title>
            <link>https://medium.com/@catherine.narae?source=rss-2d0313d2795c------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Wed, 18 Nov 2026 06:42:11 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@catherine.narae/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[How Much Data Analysis Should a PM Know — and How Should They Use It?]]></title>
            <link>https://medium.com/@catherine.narae/how-much-data-analysis-should-a-pm-know-and-how-should-they-use-it-aff0a6815fbe?source=rss-2d0313d2795c------2</link>
            <guid isPermaLink="false">https://medium.com/p/aff0a6815fbe</guid>
            <dc:creator><![CDATA[Catherine Narae]]></dc:creator>
            <pubDate>Thu, 14 Aug 2025 12:10:51 GMT</pubDate>
            <atom:updated>2025-08-14T12:10:51.440Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*PXj1Qg8S0PWf4GwGV7ML6w.png" /></figure><h3>How Much Data Analysis Should a PM Know — and How Should They Use It?</h3><p>If you browse PM job postings these days, you’ll often see “data analysis skills” listed as a requirement.<br> SQL, Python, GA, Amplitude… It can feel like you’re expected to master them all.</p><p>When I first started, it felt overwhelming. But as I worked through real projects, I began to understand what really matters.<br> Here’s what I’ve learned — <strong>what’s more important than the tools, and the level of skill that’s actually useful for a PM.</strong></p><h3>1. Start with what to analyze</h3><p>Before thinking about tools or techniques, ask this:</p><blockquote><em>“Which metric is the most relevant for our product right now?”</em></blockquote><p>The answer depends on factors like the product’s stage (early adoption, growth, maturity), market conditions, and the service’s revenue model.<br> A PM needs the judgment to identify <strong>which metric matters most in the current context</strong>.</p><h3>2. Is that metric actually being recorded?</h3><p>You can’t analyze what you’re not collecting.<br> If the data isn’t being stored, the first step is to set up the environment to capture it.</p><p>Depending on its nature, data might be stored in a database or tracked via event logs.<br> In some organizations, <strong>creating the right environment for data collection</strong> can be part of the PM’s role.</p><h3>3. Data requests are a skill (Part 1)</h3><p>When requesting data, don’t just say, “Please pull this for me.” Give context.</p><blockquote>“Here’s the background, and here’s the decision I need to make.<br> I’m requesting data on [X], but if you see a better way to approach it from an analyst’s perspective, please suggest it.<br> And if you spot areas worth a deeper dive, sharing those insights will help me make a stronger decision.”</blockquote><p>When you do this, analysts can immerse themselves in the problem, and you’re far more likely to get high-quality results.</p><h3>4. Data requests are a skill (Part 2)</h3><p>Once you get the data, always check:</p><blockquote>“Does this match the conditions I originally intended?”<br> “What rules or filters were applied when aggregating it?”</blockquote><p>If the <strong>understanding</strong> isn’t aligned before the request, you can end up with results that are completely off-target.</p><p>To make <em>good</em> requests, it helps to understand <strong>how the data is stored based on the product’s scenarios</strong>.<br> That doesn’t mean the PM must always figure this out alone — sometimes analysts aren’t deeply familiar with a product’s history.<br> In that case, <strong>connect with a backend engineer to clarify the data’s characteristics</strong>, or prepare a clear explanation yourself so the team knows exactly what they’re working with.</p><p>For example:</p><ul><li>Is it continuously updated or fixed after creation?</li><li>Under what conditions is it created, modified, or deleted?</li><li>Does the value change depending on the date range queried?</li></ul><p>Knowing this allows you to <strong>validate whether the dataset fits your analysis goal</strong> and even do light data cleansing yourself.<br> A basic understanding of RDBMS concepts will make this much easier.</p><p>In fact, if you understand this part well, you can often work without writing SQL.<br> You can request raw data extracts and process them yourself.<br> And even if you <em>do</em> know SQL, many companies don’t give PMs direct database access — meaning the ability to make precise requests remains a core skill.</p><h3>5. SQL: the basics are enough</h3><p>The SQL I use most often boils down to: <strong>SELECT, FROM, WHERE, GROUP BY, JOIN, ORDER BY</strong>.<br> You can always look up syntax as needed.<br> This is enough to pull data and process it in Excel.</p><p>When dealing with very large datasets (tens of thousands of rows or more), performance can slow down.<br> Use filters effectively or break the query into smaller date ranges.</p><h3>6. Tools change, but the questions remain</h3><p>GA, Amplitude, Redash — tools vary, and they often change when you move to a new company.<br> What doesn’t change is this:</p><blockquote>“What question am I trying to answer?”<br> “What do I actually want to know?”</blockquote><p>The people who ask good questions are the ones who use data most effectively.</p><h3>The takeaway</h3><p>For PMs, data analysis skills <strong>aren’t just about writing perfect SQL or mastering every analytics tool</strong>.</p><p>It’s about:</p><ul><li>Knowing what data to look at and why</li><li>Making clear, precise requests</li><li>Validating and interpreting results accurately</li><li>Using those insights to make the best possible decisions right now</li></ul><p>Tools can be learned.<br> But <strong>the ability to ask the right questions and connect data to judgment in context</strong> comes from experience and practice.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=aff0a6815fbe" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>