Custom CRM development

Custom CRM development

Most businesses should buy a CRM. This page is about the cases where that stops being true — and what to do when you're in one.

Forty years of build-versus-buy →

A free 30-minute call. If buying an off-the-shelf CRM is the better answer for you, I'll say so — that is usually the outcome, and it costs you nothing to find out.

“He has a unique ability to look at a process, understand exactly what needs to be done, and then find a way to make it faster, easier, and more efficient.”

— Zach, Lakeshore Optical

When you need this

  • Your CRM does about 70% of what you need, and the other 30% lives in a spreadsheet nobody fully trusts.
  • You pay per seat for a tool most of the team opens once a week to update one field.
  • The way you qualify a lead is genuinely unusual, and every off-the-shelf pipeline makes you describe it wrong.
  • The CRM and the accounting system disagree about who the customer is, and nobody can say which one to believe.
  • You've outgrown a shared spreadsheet, but every demo you sit through is built for a sales team ten times your size.
  • Reporting means exporting to Excel on the first of the month.

What we do

  • Extending the CRM you already pay for — usually the cheapest answer, and the first one worth pricing
  • A custom CRM built around the pipeline you actually run, with the stages you actually use
  • Two-way sync between your CRM and accounting, invoicing, or payment systems so the two stop disagreeing
  • Migrating years of contacts out of spreadsheets with the duplicates genuinely resolved, not just flagged
  • Role-based access, enforced on the server, so people see the work that is theirs and nothing else
  • Reporting that answers your questions rather than the ones a vendor's default dashboard was designed for
How this gets built now

Building one got cheaper. Knowing whether to build one didn't.

Scaffolding a CRM — contacts, companies, deals, stages, permissions, an audit trail — used to be weeks of unglamorous work, and it is now genuinely fast. That changes the arithmetic, and it is why this page exists at all. What did not get faster is the part that decides whether the project is a good idea: working out which slice of your process is actually yours, and which slice you are about to re-implement because that is how it has always been done. Getting that wrong means paying to rebuild something you could have bought for a few hundred a month. It is a conversation, not a code-generation problem, and it is the first thing worth having.

How I use AI coding agents on client work →
Examples from real engagements

What this looks like

Enrichment before the pipeline

A real-estate investment firm's leads arrive raw. Each record is enriched against third-party data, deduped against existing contacts, classified by a property model specific to how they buy, and pushed into their CRM — so the pipeline starts with a qualified record instead of a name and a phone number.

When the database is the business

A nationwide recruiting firm's participant database is its core asset: tens of thousands of people matched on names, phones and emails that were never linked, with fuzzy matching where exact matching quietly missed people, and a review queue where a person makes the call on every uncertain match.

Extending, not replacing

Where the CRM already works, the job is usually connecting it — qualified leads flowing in on their own, invoices flowing out, and both systems finally agreeing on who the customer is. That is a smaller project than a rebuild and it is often the right one.

Not sure whether to build, buy, or extend what you already have?

Tell me what you're trying to fix — I'll reply within a day.

Want the reasoning rather than the pitch? Forty years of build-versus-buy.