<?xml version="1.0" encoding="utf-8" standalone="yes"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:media="http://search.yahoo.com/mrss/">
  <channel>
    <title>Engineering Management · Jared Lynskey</title>
    <link>https://jared.lynskey.co.nz/en/tags/engineering-management/</link>
    <description>Short stories of value on software engineering and engineering leadership — performance reviews, hiring and scaling teams, CI/CD, DevOps monitoring and agentic AI. Written from Seoul.</description>
    <generator>Hugo</generator>
    <language>en</language>
    <copyright>© 2026 Jared Lynskey</copyright>
    <lastBuildDate>Tue, 27 Oct 2026 00:00:00 +1300</lastBuildDate>
    <atom:link href="https://jared.lynskey.co.nz/en/tags/engineering-management/index.xml" rel="self" type="application/rss+xml" />
    <item>
      <title>Using the STAR Method for Developer Interviews</title>
      <link>https://jared.lynskey.co.nz/en/posts/2024/2024-12-19-star-interviews/</link>
      <guid isPermaLink="true">https://jared.lynskey.co.nz/en/posts/2024/2024-12-19-star-interviews/</guid>
      <pubDate>Thu, 19 Dec 2024 00:30:03 +0000</pubDate>
      <dc:creator>Jared Lynskey</dc:creator>
      <category>Hiring</category>
      <category>Interviews</category>
      <category>STAR Format</category>
      <category>Engineering Management</category>
      <description>Behavioural interviewing with the STAR method: ask for a real story, then draw out the Situation, Task, Action and Result. The questions I ask developers, leadership ones included, and what I actually listen for.</description>
      <content:encoded><![CDATA[<p>The STAR interview method is a way of asking behavioural questions about what a candidate has actually done, rather than what they say they&rsquo;d do. You ask for a real example, such as &ldquo;Tell me about a time you influenced a technical decision without direct authority&rdquo;, and listen for four parts: the Situation, the Task they were responsible for, the Action they took themselves, and the Result. I use it for developer interviews, and the follow-up questions do most of the work.</p>
<p>Since moving into an engineering management role this year I&rsquo;ve spent a lot more time on the hiring side of the table, and the biggest lesson so far is that letting a candidate talk generally about their CV tells you almost nothing. A resume says what someone worked on. It says very little about how they work.</p>
<p>Behavioural questions in the STAR format are the best tool I&rsquo;ve found for closing that gap: a real story from the candidate&rsquo;s past, steered through those four beats. It sounds like HR boilerplate — it <em>is</em> HR boilerplate — but it forces specifics, and specifics are where the signal is.</p>

<h3 class="relative group">The shape of a good answer
    <div id="the-shape-of-a-good-answer" class="anchor"></div>
    
    <span
        class="absolute top-0 w-6 transition-opacity opacity-0 -start-6 not-prose group-hover:opacity-100 select-none">
        <a class="text-primary-300 dark:text-neutral-700 !no-underline" href="#the-shape-of-a-good-answer" aria-label="Anchor">#</a>
    </span>
    
</h3>
<p>Take a question like &ldquo;Tell me about a time you resolved a complex bug.&rdquo; You&rsquo;re listening for the situation (what was going on and why it mattered), the task the candidate was personally on the hook for, the actions they took — the &ldquo;I&rdquo;, not the &ldquo;we&rdquo; — and the result, ideally with a number attached.</p>
<p>A strong answer sounds something like this one, about technical debt:</p>
<blockquote><p>&ldquo;Our deployment pipeline took 45 minutes, and flaky tests meant developers often re-ran builds several times. I tracked the time lost over a month — roughly 15 hours per developer. I presented that to leadership and proposed a two-sprint investment to parallelise the tests and fix the flakiest ones. Deployments dropped to 12 minutes, and our deployment frequency doubled.&rdquo;</p>
</blockquote><p>Situation, task, action, result, in five sentences, with the candidate&rsquo;s own contribution unmistakable. Put that next to &ldquo;we had a slow pipeline so we fixed it&rdquo; and you can see why the structure earns its keep.</p>

<h3 class="relative group">The questions I keep coming back to
    <div id="the-questions-i-keep-coming-back-to" class="anchor"></div>
    
    <span
        class="absolute top-0 w-6 transition-opacity opacity-0 -start-6 not-prose group-hover:opacity-100 select-none">
        <a class="text-primary-300 dark:text-neutral-700 !no-underline" href="#the-questions-i-keep-coming-back-to" aria-label="Anchor">#</a>
    </span>
    
</h3>
<p>Rather than trying to cover everything, I organise questions around the competencies the team actually needs right now:</p>
<table>
	<thead>
			<tr>
					<th>Competency</th>
					<th>Question</th>
			</tr>
	</thead>
	<tbody>
			<tr>
					<td>Problem-solving</td>
					<td>&ldquo;Tell me about a time you debugged a production issue under pressure.&rdquo;</td>
			</tr>
			<tr>
					<td>Collaboration</td>
					<td>&ldquo;Describe a project where you had to work with a difficult stakeholder.&rdquo;</td>
			</tr>
			<tr>
					<td>Technical leadership</td>
					<td>&ldquo;Tell me about a time you influenced a technical decision without direct authority.&rdquo;</td>
			</tr>
			<tr>
					<td>Learning agility</td>
					<td>&ldquo;Describe when you had to get up to speed on unfamiliar technology, fast.&rdquo;</td>
			</tr>
			<tr>
					<td>Ownership</td>
					<td>&ldquo;Tell me about a time something you built failed. What did you do?&rdquo;</td>
			</tr>
	</tbody>
</table>
<p>That last one is my favourite. How someone talks about a failure tells you more than three success stories, and it&rsquo;s very hard to rehearse.</p>

<h3 class="relative group">Where the interview actually happens
    <div id="where-the-interview-actually-happens" class="anchor"></div>
    
    <span
        class="absolute top-0 w-6 transition-opacity opacity-0 -start-6 not-prose group-hover:opacity-100 select-none">
        <a class="text-primary-300 dark:text-neutral-700 !no-underline" href="#where-the-interview-actually-happens" aria-label="Anchor">#</a>
    </span>
    
</h3>
<p>Candidates rarely give you the full STAR arc unprompted, and that&rsquo;s fine — the follow-ups are the real interview. If someone says &ldquo;we improved performance&rdquo;, ask by how much and how they measured it. If everything is &ldquo;we&rdquo;, ask what they personally did. If a story wraps up suspiciously neatly, ask what they&rsquo;d do differently next time.</p>
<p>The other thing that took me a while to learn: ask the same questions of every candidate and score the answers against a simple rubric. It feels mechanical, but it&rsquo;s the only fair way to compare people. Otherwise you just end up favouring whoever you spoke to most recently.</p>
<p>None of this replaces a technical assessment — you still need to see whether someone can build things. But hypotheticals are easy to bluff and past behaviour is hard to fake. After a year of running these, I trust a well-probed story far more than a polished answer about strengths and weaknesses.</p>

<h3 class="relative group">STAR in four lines
    <div id="star-in-four-lines" class="anchor"></div>
    
    <span
        class="absolute top-0 w-6 transition-opacity opacity-0 -start-6 not-prose group-hover:opacity-100 select-none">
        <a class="text-primary-300 dark:text-neutral-700 !no-underline" href="#star-in-four-lines" aria-label="Anchor">#</a>
    </span>
    
</h3>
<ul>
<li>Situation: what was going on, and why it mattered.</li>
<li>Task: the part the candidate was personally responsible for.</li>
<li>Action: what they did themselves. If it&rsquo;s all &ldquo;we&rdquo;, ask what they did.</li>
<li>Result: what came of it, ideally with a number. If they say they improved something, ask by how much and how they measured it.</li>
</ul>

<h3 class="relative group">Further Reading
    <div id="further-reading" class="anchor"></div>
    
    <span
        class="absolute top-0 w-6 transition-opacity opacity-0 -start-6 not-prose group-hover:opacity-100 select-none">
        <a class="text-primary-300 dark:text-neutral-700 !no-underline" href="#further-reading" aria-label="Anchor">#</a>
    </span>
    
</h3>
<p>On this site:</p>
<ul>
<li><a href="https://jared.lynskey.co.nz/en/posts/2024/2024-07-12-coffee-chats/" >The Coffee Chat: A Precursor to the Hiring Process</a> — getting to know candidates before the formal process starts</li>
<li><a href="https://jared.lynskey.co.nz/en/posts/2024/2024-02-10-prioritizing-onboarding/" >Prioritizing Team Dynamics in Engineering Management</a> — hiring for how someone will fit the team</li>
<li><a href="https://jared.lynskey.co.nz/en/posts/2024/2024-07-01-internships/" >The Benefits of Interning at a Startup</a> — useful context when evaluating early-career candidates</li>
<li><a href="https://jared.lynskey.co.nz/en/posts/2023/2023-09-01-retrospecs/" >Retrospectives: The Power of Feedback Loops</a> — the same feedback habits apply to your interview process</li>
</ul>
<p>Elsewhere:</p>
<ul>
<li><a href="https://www.themuse.com/advice/star-interview-method"  target="_blank" rel="noreferrer">The STAR Method: The Secret to Acing Your Next Job Interview</a> — The Muse&rsquo;s guide, written for candidates</li>
<li><a href="https://rework.withgoogle.com/intl/en/guides/hiring-use-structured-interviewing"  target="_blank" rel="noreferrer">Google re:Work: Structured Interviewing</a> — the research behind structured interviews</li>
<li><a href="https://review.firstround.com/40-favorite-interview-questions-from-some-of-the-sharpest-folks-we-know/"  target="_blank" rel="noreferrer">First Round Review: 40 Favorite Interview Questions</a> — good for stealing</li>
<li><a href="https://www.joelonsoftware.com/2006/10/25/the-guerrilla-guide-to-interviewing-version-30/"  target="_blank" rel="noreferrer">Joel Spolsky: The Guerrilla Guide to Interviewing</a> — a classic that still holds up</li>
</ul>
]]></content:encoded>
      <media:content url="https://jared.lynskey.co.nz/en/posts/2024/2024-12-19-star-interviews/featured.jpg" medium="image" type="image/jpeg" />
    </item>
    <item>
      <title>Project Communication Plan: A Template and Example for Software Teams</title>
      <link>https://jared.lynskey.co.nz/en/posts/2023/2023-06-02-project-frameworks-for-future-cto/</link>
      <guid isPermaLink="true">https://jared.lynskey.co.nz/en/posts/2023/2023-06-02-project-frameworks-for-future-cto/</guid>
      <pubDate>Fri, 02 Jun 2023 11:30:03 +0000</pubDate>
      <dc:creator>Jared Lynskey</dc:creator>
      <category>Project Management</category>
      <category>Communication Plan</category>
      <category>Communication</category>
      <category>Stakeholder Management</category>
      <category>Engineering Management</category>
      <category>Remote Work</category>
      <description>A project communication plan says who needs to know what, how often, through which channel, and who tells them. What goes in one, an example plan for a software project you can copy, and the habits that keep it working.</description>
      <content:encoded><![CDATA[<p>A project communication plan is a short table that says who needs to know what about a project, how often, through which channel, and who is responsible for telling them. Mine fits on one page, with one row per audience, from the dev team up to the CEO. Below is the one I&rsquo;d start a software project with, and the notes I&rsquo;ve been keeping on what makes project communication actually work.</p>
<p>Most of the project messes I&rsquo;ve been pulled into weren&rsquo;t caused by bad code. They were caused by the right information not reaching the right person until it was too late. Since stepping up as project lead at work, I&rsquo;ve been keeping notes on how I&rsquo;d want to run communication if I ever end up in the CTO seat. Here they are.</p>

<h4 class="relative group">Decide who needs to know what
    <div id="decide-who-needs-to-know-what" class="anchor"></div>
    
    <span
        class="absolute top-0 w-6 transition-opacity opacity-0 -start-6 not-prose group-hover:opacity-100 select-none">
        <a class="text-primary-300 dark:text-neutral-700 !no-underline" href="#decide-who-needs-to-know-what" aria-label="Anchor">#</a>
    </span>
    
</h4>
<p>The unglamorous foundation is a communication plan: who needs to be informed, what they need, how often, and through which channel. I&rsquo;d push for one at the start of every project, with names attached — communication that&rsquo;s everyone&rsquo;s job ends up being no one&rsquo;s job. The same goes for stakeholders: work out who actually cares about the project, learn how each of them prefers to be updated, and don&rsquo;t send the CEO the same update you post in the dev channel. Finally, set a few explicit protocols — expected response times, what counts as urgent, which channel is for what. Most &ldquo;communication problems&rdquo; are really just unstated expectations.</p>

<h4 class="relative group">An example communication plan for a software project
    <div id="an-example-communication-plan-for-a-software-project" class="anchor"></div>
    
    <span
        class="absolute top-0 w-6 transition-opacity opacity-0 -start-6 not-prose group-hover:opacity-100 select-none">
        <a class="text-primary-300 dark:text-neutral-700 !no-underline" href="#an-example-communication-plan-for-a-software-project" aria-label="Anchor">#</a>
    </span>
    
</h4>
<p>This is the plan I&rsquo;d start with and then adjust. Each row is an audience, not a meeting:</p>
<table>
	<thead>
			<tr>
					<th>Audience</th>
					<th>What they need to know</th>
					<th>How often</th>
					<th>Channel</th>
					<th>Who tells them</th>
			</tr>
	</thead>
	<tbody>
			<tr>
					<td>Dev team</td>
					<td>What&rsquo;s in progress, what&rsquo;s blocked, decisions that change the code</td>
					<td>Daily</td>
					<td>Stand-up and the team&rsquo;s chat channel</td>
					<td>Tech lead</td>
			</tr>
			<tr>
					<td>Product owner</td>
					<td>Progress against the sprint goal, scope questions, new risks</td>
					<td>Twice a week, and whenever scope changes</td>
					<td>Short call or a chat thread</td>
					<td>Project lead</td>
			</tr>
			<tr>
					<td>Leadership</td>
					<td>One-line status, risks to the date or budget, decisions they need to make</td>
					<td>Weekly</td>
					<td>A written update of one paragraph</td>
					<td>Project lead</td>
			</tr>
			<tr>
					<td>Sales, support and ops</td>
					<td>What&rsquo;s shipping, when, and what changes for customers</td>
					<td>Every release</td>
					<td>Release notes</td>
					<td>Product owner</td>
			</tr>
			<tr>
					<td>Users or clients</td>
					<td>What&rsquo;s new and what&rsquo;s coming</td>
					<td>Every release</td>
					<td>Changelog or an in-app notice</td>
					<td>Product owner</td>
			</tr>
			<tr>
					<td>Everyone</td>
					<td>What was decided, and why</td>
					<td>When it&rsquo;s decided</td>
					<td>A decision log in the issue tracker or wiki</td>
					<td>Whoever made the decision</td>
			</tr>
	</tbody>
</table>
<p>Underneath the table go the protocols, written as plainly as possible: urgent means it blocks a release or a customer, so call rather than message; questions go in chat, decisions go in the tracker; replies within a working day. For the decision log I use <a href="https://jared.lynskey.co.nz/en/posts/2024/2024-12-03-decision-docs/" >decision documents</a>, kept as <a href="https://jared.lynskey.co.nz/en/posts/2024/2024-01-04-githubforprojectmanagement/" >GitHub Issues</a> so they sit next to the work.</p>
<p>Two things make a plan like this work. Every row has a name in the last column, and the leadership update is written rather than spoken, so it can be forwarded without losing anything.</p>

<h4 class="relative group">Tools and rhythms
    <div id="tools-and-rhythms" class="anchor"></div>
    
    <span
        class="absolute top-0 w-6 transition-opacity opacity-0 -start-6 not-prose group-hover:opacity-100 select-none">
        <a class="text-primary-300 dark:text-neutral-700 !no-underline" href="#tools-and-rhythms" aria-label="Anchor">#</a>
    </span>
    
</h4>
<p>Tools matter less than people argue about them. Slack for chat, Trello for tasks, Zoom when a conversation needs faces — pick what fits the team and move on. Rhythm matters more: daily stand-ups if you&rsquo;re working agile, weekly updates on longer cycles, and actually holding to them. And write things down. Decisions, changes, action items — into Notion or Jira, as part of the process rather than an afterthought. Six months later, the doc is the only memory anyone has.</p>

<h4 class="relative group">The soft parts are the hard parts
    <div id="the-soft-parts-are-the-hard-parts" class="anchor"></div>
    
    <span
        class="absolute top-0 w-6 transition-opacity opacity-0 -start-6 not-prose group-hover:opacity-100 select-none">
        <a class="text-primary-300 dark:text-neutral-700 !no-underline" href="#the-soft-parts-are-the-hard-parts" aria-label="Anchor">#</a>
    </span>
    
</h4>
<p>Everything above is mechanics. The harder work is asking the team whether the communication is actually working, and changing it when it isn&rsquo;t. It&rsquo;s training people up, too — teams come from varied backgrounds, and not everyone has seen a well-run project before. It&rsquo;s listening properly instead of waiting for your turn to talk, and dealing with conflict early, while it&rsquo;s still small. None of it stays solved, either: what worked last quarter quietly stops working, so keep checking.</p>

<h3 class="relative group">Working as a distributed/asynchronous development team
    <div id="working-as-a-distributedasynchronous-development-team" class="anchor"></div>
    
    <span
        class="absolute top-0 w-6 transition-opacity opacity-0 -start-6 not-prose group-hover:opacity-100 select-none">
        <a class="text-primary-300 dark:text-neutral-700 !no-underline" href="#working-as-a-distributedasynchronous-development-team" aria-label="Anchor">#</a>
    </span>
    
</h3>
<p>This talk is worth your time if your team is remote or async-heavy:
<div style="position: relative; padding-bottom: 56.25%; height: 0; overflow: hidden;">
			<iframe allow="accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture; web-share; fullscreen" loading="eager" referrerpolicy="strict-origin-when-cross-origin" src="https://www.youtube.com/embed/5lVijFaXaKs?autoplay=0&amp;controls=1&amp;end=0&amp;loop=0&amp;mute=0&amp;start=0" style="position: absolute; top: 0; left: 0; width: 100%; height: 100%; border:0;" title="YouTube video"></iframe>
		</div>
</p>
<p>I don&rsquo;t do all of this perfectly — nobody does. But every project that went smoothly for me had most of these in place, and every one that hurt was missing at least three.</p>
]]></content:encoded>
      <media:content url="https://jared.lynskey.co.nz/en/posts/2023/2023-06-02-project-frameworks-for-future-cto/featured.jpeg" medium="image" type="image/jpeg" />
    </item>
  </channel>
</rss>
