Businesses buy CRM software, log in, look at an empty system, and put it aside. Six months later they are still working leads out of an inbox and a notepad, and they have a subscription they are not using.

The software is not the thing. The configuration is the thing, and it is most of the work.

Pipeline stages that match your actual business

Every CRM ships with generic stages — Lead, Qualified, Proposal, Won. They fit almost nobody, and using them means the reporting describes an imaginary business.

Real stages match the path your work actually takes. For a service business that might be: Enquiry, Contacted, Estimate Scheduled, Estimate Completed, Quoted, Booked, Completed, Invoiced. For an agency-type business it looks different again.

This matters because everything else hangs off the stages: the automation triggers, the tasks, the reporting, and the conversion rates you will use to make decisions. Getting them wrong at the start means rebuilding later.

Capture from every source, with the source attached

If leads arrive through a website form, a phone call, a Google Business Profile message, a Facebook lead form, and a chat widget, all five have to land in the CRM automatically, tagged with where they came from.

Any source not connected is invisible. Any source connected without attribution means you can never tell which channel produced the job, which makes all downstream reporting decorative.

The automation layer

This is where the value actually is:

Instant response. Text and email within seconds of any enquiry, plus a task raised for a human.

Missed-call text-back. Automatic, immediate, on every unanswered call.

Assignment and routing. By service area, service type, or round-robin, with escalation when nobody responds.

Follow-up sequences. Multi-touch, by lead type and by stage, running on a schedule rather than on memory.

Appointment confirmations, reminders, and no-show recovery.

Post-job review requests, triggered by completion.

Reactivation and maintenance reminders, based on service date and interval.

None of this is exotic. All of it is the difference between a database and a system.

Compliance, built in rather than added

A2P 10DLC registration, opt-in language on every form, opt-out handling that works immediately and permanently, and consent recorded against each contact.

Adding this after the first campaign is how businesses end up with a list they cannot legally use. It belongs in the build.

Data migration, honestly

Existing customer records, past jobs, and contact history need to come across, deduplicated and cleaned. This is the least interesting part of the project and the part that most often gets skipped or rushed.

A CRM without your history in it cannot run reactivation, cannot show a technician what happened last time, and cannot tell you which customers are overdue. That is most of the value gone.

Reporting that answers real questions

Configured before launch, not bolted on later. It should answer: how many leads by source, what did each source cost per booked job, how fast are we responding, what is the conversion rate at each stage, and where are we losing people.

If a CRM cannot answer those, it is an address book.

What it does not do

It does not replace your office staff, and any pitch that says it does is overselling. It removes the follow-up that never happens in a busy week and frees your people for the conversations that need a person.

It also does not fix a business with no lead flow. A CRM makes existing leads convert better. It does not create them.

How long it takes

Typically a few weeks for a service business, and the timeline is usually driven by access, data, and sign-off rather than by the configuration itself. Anyone quoting a same-day setup is switching software on, not building a system.