Skip to content
Lab notes
LN-00610 Sept 20261 min readBuilding internal tools

Build the missing piece

Why a growing brand ends up with a dozen small apps, and why that is healthier than one big platform.

Every tool I have built this year started the same way: someone asked a question, and answering it took four tabs and ten minutes.

What is today's conversion? Which SKUs sell out this week? Who changed that price? What is ROAS on net revenue across Shopify and Amazon? None of these are exotic. All of them were slow.

The gap is the problem

Off-the-shelf software is built for the median merchant. A brand with retail stores, marketplaces, a paid membership and a custom POS is not the median merchant. The tool almost fits, and the almost is where people spend their afternoons.

The instinct is to buy something bigger. The better move is usually to build something smaller: one screen, one question, one source of truth, gated to the team's Google accounts, done in a week.

Small tools compound

A dozen small apps sounds like sprawl. In practice they share almost everything: the same query layer, the same auth pattern, the same deploy path, the same design language. The fifth one takes a day because the first four paid for the plumbing.

And each one changes behaviour a little. When the sell-out list is a push notification, people act on it. When ROAS is on a phone, the number gets checked before the meeting, not in it.

The rule I use

If a person does the same thing on the same schedule, it is a cron job. If a person asks the same question every day, it is a screen. If a system can't answer it, that is the spec.

internal toolsbuild vs buyShopify
Written by Nikhilendra Pratap Singh Deo. Facts referenced here come from public sources or Nikhil’s own posts.