What this will cost you in money, time and internal effort.

We're on site, following the work from enquiry through to invoice, talking to the people who do the work as well as the people who manage the work. You get a map of your processes and an agreed list of what the system has to handle, awkward exceptions included.

A few hours from each of your people, and enough trust to show us the workarounds instead of the tidy version. Your workarounds are where the real requirements live.

We build the system: items, bills of materials, job structures, workflow, pricing rules, user roles and the link to your ledger. You see the build in stages and we ask you to react as we go, so nothing lands as a surprise at the end.

Decisions, mostly, and some of them are ones your business has put off for years. Configuration is where those decisions surface. Expect a handful of calls only an owner or a manager can make.

We extract, clean, map and load, then reconcile the result against what your current system says. Where the data's too poor to bring across, we tell you, and we agree what happens to that data instead.

Somebody who knows your data has to sit with it. Duplicate customers, items nobody's sold since 2014 and stock figures that were never right don't fix themselves in transit. Deciding what gets left behind is a call for your business to make.

We build the test scenarios from the process map, sit with your team while they work through the scenarios, and fix whatever falls over.

Your people put real work through the system, messy jobs included. Testing is the phase most often dropped when time gets tight, and dropping testing is the most reliable way to have a bad go-live.

We train by role instead of by module, so each person learns their own working day rather than taking a tour of the software. We leave written notes behind, pitched at the person who has to remember all of this six weeks later.

Time released for the people being trained, and honesty about who's nervous. The quiet resister will shape your outcome more than the enthusiastic early adopter will.

We handle the final data load, the opening balances and the cutover checklist, and we're there through the first days of live running. Problems in week one get fixed on the spot rather than logged for later.

A short stretch where your people work slower and check things twice. Build reduced output into your plan for the first week and the dip won't hurt you.

We check in regularly while habits form, tune the things that only show up under real load, build the reports your people ask for once they can see what's possible, and answer questions as the same person who built your configuration.

Keep using the system properly while using it still feels uncomfortable. The pull back to the old spreadsheet peaks between weeks two and six, and a half-used system will serve you worse than the one you replaced.