Why generic CRMs fail builders
A mainstream CRM models a sales pipeline: lead, qualified, proposal, closed won. A builder's world runs enquiry, survey, quote, job, variations, retention, repeat client. Those are different shapes. Forcing the second into the first is why so many construction firms buy a CRM in January, enter three weeks of data, and have stopped opening it by March.
The tool is not broken. It was built for a business where the deal closing is the end of the story. In building work, the deal closing is where the risk, the paperwork and most of the relationship actually begin.
Why is the CRM empty by March?
Because it demands double entry and gives nothing back for it.
The real record of a job lives in the WhatsApp group, the email thread, the photos folder and the quoting spreadsheet. A generic CRM asks someone to retype a summary of all four into a "deal", by hand, on top of the day job. The moment workload rises, retyping loses, and workload always rises. The pattern is the same one we described in licences installed, habits never formed: the software is present, the behaviour it depends on is not.
There is also a quieter reason. A sales pipeline gives a builder nothing to look at. Once the job is won, the deal card sits in "closed won" doing nothing while the actual work, the variations and the money all happen somewhere the CRM cannot see. People stop opening tools that have nothing to tell them.
What does a sales pipeline get wrong about building work?
Almost every assumption baked into it.
| The pipeline CRM assumes | A builder's reality |
|---|---|
| "Closed won" ends the process | Winning the job starts the risky part |
| One contact per deal | Homeowner, architect, QS and site agent on the same job |
| Revenue is one number, booked once | Valuations, variations, retention held and released |
| Follow-up is a sales cadence | Follow-up is snagging, the defects period and retention release |
| Activity means logged calls and emails | The real activity is on site, in photos and voice notes |
Retention alone shows the mismatch. A generic CRM has no concept of a defects period ending eleven months from now, or of retention release tied to making good under the contract's payment terms, the machinery of which hangs off the Construction Act. Ask any QS how often retention goes unclaimed simply because nobody diarised the date. A tool that cannot hold that date is not managing the customer relationship; it is decorating it.
What does a construction CRM actually need?
Five things, and none of them is a pipeline view.
First, job-centric records. The job is the unit of truth, with contacts, companies, quotes, variations and dates hanging off it, not the other way round. Second, site and office communication flowing into the record without anyone retyping it, which is the argument we make in zero-click automation: the record that fills itself is the only one that stays full. Third, quote follow-up: every quote carries a chase date, and dormant quotes surface on their own instead of dying in a sent folder. Fourth, money dates: retention release, defects period end, final account, each with an owner. Fifth, repeat-client memory: what you did, when, and when to get back in touch.
One legal note while you are at it. The moment you hold past clients' details for keeping in touch, you are a data controller with obligations under UK GDPR. None of it is onerous for a small firm, but check the basics against the ICO's guidance for organisations rather than assuming the CRM vendor handled it. They did not; that responsibility stays with you.
When is a spreadsheet honestly fine?
More often than software vendors would like.
If one person does all the quoting, you run a handful of live jobs at a time, and follow-up fits in a Friday half hour, a well-kept spreadsheet with columns for quote date, chase date, defects period end and retention date beats an unused CRM comfortably. It costs nothing, everyone understands it, and it never demands a login.
The spreadsheet stops being fine at a recognisable point: when two or more people need the same record at once, when the volume of communication outgrows manual logging, or when the cost of a forgotten chase date exceeds the cost of fixing the system. That last threshold arrives earlier than most people think, which is the sum we walk through in the real cost of doing nothing.
So what should a builder actually buy?
Possibly nothing, yet. The order of operations is the point.
Fix capture first: get site traffic, quotes and dates flowing into one place automatically, whether that place is a spreadsheet, a job management tool or something built around your workflow. Then add reminders that fire on their own. Only then judge whether you need anything with CRM on the label. This is why we describe our work as automation rather than software: the fix is rarely another login, and it is never another place to retype what was already written down once.
AI Metric builds job-centric hubs of exactly this kind for small firms, but the test in this post applies whatever you buy: if the record does not fill itself, it will be empty by March, and March always comes.