<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:dc="http://purl.org/dc/elements/1.1/">
  <channel>
    <title>DEV Community: Shohan Shahid</title>
    <description>The latest articles on DEV Community by Shohan Shahid (@shohanshahid).</description>
    <link>https://dev.to/shohanshahid</link>
    <image>
      <url>https://media2.dev.to/dynamic/image/width=90,height=90,fit=cover,gravity=auto,format=auto/https:%2F%2Fdev-to-uploads.s3.amazonaws.com%2Fuploads%2Fuser%2Fprofile_image%2F3725218%2Fb529201e-da05-457f-ac9c-4b549cdb4f5c.jpeg</url>
      <title>DEV Community: Shohan Shahid</title>
      <link>https://dev.to/shohanshahid</link>
    </image>
    <atom:link rel="self" type="application/rss+xml" href="https://dev.to/feed/shohanshahid"/>
    <language>en</language>
    <item>
      <title>Search latency is not a backend detail. It is a conversion constraint. This Field Notes post shows how we kept search fast under load and added a release gate that blocks regressions. What is your search P95 and how do you measure it end to end?</title>
      <dc:creator>Shohan Shahid</dc:creator>
      <pubDate>Thu, 22 Jan 2026 08:30:24 +0000</pubDate>
      <link>https://dev.to/shohanshahid/search-latency-is-not-a-backend-detail-it-is-a-conversion-constraint-this-field-notes-post-shows-1jj2</link>
      <guid>https://dev.to/shohanshahid/search-latency-is-not-a-backend-detail-it-is-a-conversion-constraint-this-field-notes-post-shows-1jj2</guid>
      <description>&lt;div class="ltag__link"&gt;
  &lt;a href="/shohanshahid" class="ltag__link__link"&gt;
    &lt;div class="ltag__link__pic"&gt;
      &lt;img src="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.amazonaws.com%2Fuploads%2Fuser%2Fprofile_image%2F3725218%2Fb529201e-da05-457f-ac9c-4b549cdb4f5c.jpeg" alt="shohanshahid"&gt;
    &lt;/div&gt;
  &lt;/a&gt;
  &lt;a href="https://dev.to/shohanshahid/latency-is-revenue-3j1p" class="ltag__link__link"&gt;
    &lt;div class="ltag__link__content"&gt;
      &lt;h2&gt;Latency is revenue&lt;/h2&gt;
      &lt;h3&gt;Shohan Shahid ・ Jan 22&lt;/h3&gt;
      &lt;div class="ltag__link__taglist"&gt;
        &lt;span class="ltag__link__tag"&gt;#performance&lt;/span&gt;
        &lt;span class="ltag__link__tag"&gt;#webdev&lt;/span&gt;
        &lt;span class="ltag__link__tag"&gt;#architecture&lt;/span&gt;
        &lt;span class="ltag__link__tag"&gt;#systemdesign&lt;/span&gt;
      &lt;/div&gt;
    &lt;/div&gt;
  &lt;/a&gt;
&lt;/div&gt;


</description>
      <category>performance</category>
      <category>webdev</category>
      <category>architecture</category>
      <category>systemdesign</category>
    </item>
    <item>
      <title>Latency is revenue</title>
      <dc:creator>Shohan Shahid</dc:creator>
      <pubDate>Thu, 22 Jan 2026 08:20:35 +0000</pubDate>
      <link>https://dev.to/shohanshahid/latency-is-revenue-3j1p</link>
      <guid>https://dev.to/shohanshahid/latency-is-revenue-3j1p</guid>
      <description>&lt;p&gt;The case&lt;/p&gt;

&lt;p&gt;Trainstop is Australia’s sports recruitment app that connects clubs with athletes. Clubs can search over 5,000 plus player profiles and contact players in app.&lt;/p&gt;

&lt;p&gt;Codezzi built Trainstop from scratch in a staff augmentation model. We shipped as part of the core team, not as an outside vendor.&lt;/p&gt;

&lt;p&gt;The constraint&lt;/p&gt;

&lt;p&gt;As usage and data grew, search stretched into multiple seconds. Users left on the first search because the product failed its first promise.&lt;/p&gt;

&lt;p&gt;Why it matters&lt;/p&gt;

&lt;p&gt;In a recruiting marketplace, discovery drives everything. When search slows down, clubs stop exploring and players stop applying.&lt;/p&gt;

&lt;p&gt;That slowdown reduces messages and weakens the trial motion. Leaders then carry delivery anxiety because the team cannot explain where time goes.&lt;/p&gt;

&lt;p&gt;What we did&lt;/p&gt;

&lt;p&gt;We treated search as a product surface with a budget. We measured the full journey from input to the first usable result.&lt;/p&gt;

&lt;p&gt;Then we made four decisions and enforced them.&lt;br&gt;
Define fast in user terms and write it down.&lt;br&gt;
Reduce work per query and move heavy work off the hot path.&lt;br&gt;
Protect speed under load with worst case tests.&lt;br&gt;
Add a release gate that blocks regressions before users feel them.&lt;/p&gt;

&lt;p&gt;Trainstop offers clubs a one week free trial and lets them connect with an unlimited number of players. That makes performance a revenue constraint, not a nice to have.&lt;/p&gt;

&lt;p&gt;Freedom is the courage gained through wisdom. Measurement creates the wisdom, and the release gate creates the courage to hold the line.&lt;/p&gt;

&lt;p&gt;What you can copy&lt;/p&gt;

&lt;p&gt;Pick one journey and measure it the same way every time. Pick one percentile and use it as the decision metric.&lt;/p&gt;

&lt;p&gt;Set a budget that matches the business value of the action. Block regressions at release, not after churn.&lt;/p&gt;

&lt;p&gt;Reply with your product surface and how you measure today. If you share your current P95, I will reply with a one page measurement plan you can implement this week.&lt;/p&gt;

&lt;p&gt;I build at &lt;a href="//codezzi.com"&gt;Codezzi&lt;/a&gt;&lt;/p&gt;

</description>
      <category>performance</category>
      <category>webdev</category>
      <category>architecture</category>
      <category>systemdesign</category>
    </item>
  </channel>
</rss>
