The checklist we run before recommending anything
The features rarely decide it. What decides it is whether the software matches how you already schedule, quote and invoice — and how cleanly it talks to your accounting system. This is the checklist we run before recommending anything, including whether to build.
Write down how a job actually moves through your business before you look at a single product. Enquiry, quote, scheduling, the work, variations, invoice, payment.
Then compare products against that, not against each other. Two products with identical feature lists can fit your process very differently, and the feature list will not tell you which.
Be specific about the messy parts. Where does a job change scope? Who is allowed to approve that? What happens when a customer pays part of an invoice? These are the points where products differ most and where demos never go.
How does it talk to your accounting package, and how well? Not "does it integrate" — every product says yes. Ask what syncs, in which direction, how often, and what happens when the sync fails.
This one question predicts more about your daily experience than the entire feature comparison, because a weak accounting sync means somebody re-types, forever.
Ask for specifics on three things: do customers sync both ways or only out; do part-payments come back; and do credit notes flow at all. Those three cover most of the re-keying we are asked to remove afterwards, and they are all answerable in a sentence by anyone who actually knows the product.
The screen where a job gets accepted and the screen where a variation gets recorded. Everything else is office software, and the office will cope.
Put both in front of the person who will actually use them, on their own phone, outside, with gloves on if that is the job. That five-minute test has killed more shortlists than any spreadsheet.
Demos run on clean data with a happy path. Yours will not. Ask to see: a job that changes scope after acceptance, an invoice that needs partial credit, and a customer with two sites and one account.
Ask what happens on a bad phone signal. Field software that assumes connectivity fails quietly in exactly the places you work.
Ask to drive it yourself rather than watching. Salespeople navigate their own product at a speed nobody else achieves, and watching a fluent demo tells you nothing about the fourth time your apprentice has to find the variation screen in the rain.
Per-seat pricing tends to bite at the point where you add office staff, not field staff. Model it at the headcount you expect in three years, not today’s.
Then check what is extra: extra for the accounting connector, extra for the customer portal, extra for API access. API access being a paid tier is worth knowing before you plan any automation on top.
That last one matters more than it looks. If the API sits behind the top pricing tier, every future integration carries that tier as a hidden prerequisite — which is the sort of thing that turns a cheap product into an expensive one two years in. It is the same trap a vague quote sets, arriving from the other direction.
Run one real job — not a test job — end to end in the trial. A real job brings its own mess, which is the entire point.
Agree beforehand what would make it a no. Deciding the criteria after you have already spent two weeks configuring is how people talk themselves into the wrong product.
Write the criteria down and give them to someone who is not doing the evaluating. Sunk cost is not a character flaw, it is just how people work, and the only reliable defence is having written the test before you started paying for it. If you would rather have someone else run the pilot with you, that is what the Leak Check is for.
Sometimes the shortlist is genuinely wrong for you — but far less often than people conclude at this stage, because by now they are tired of looking.
Before deciding nothing fits, check whether the gap is really at the joins rather than in the product. Custom vs off-the-shelf walks that distinction, and the decision framework gets you to an answer in about an hour.
Seven questions to run against your leading candidate. Answer them about that product specifically.
0 of 7 answered. The result appears once you have answered them all.
Nothing here is sent anywhere. If it comes back short, the missing items are usually the field test and the real-job trial — both free, both a week.
Because "best" depends on your process, your accounting package and your field team, and any list that ignores those is an affiliate page. We will name what we have implemented against and what we have not — but not a winner for a business we have not seen.
About three weeks. A week to write down your process and shortlist, two weeks to pilot one real job. Faster than that and you are skipping the field test, which is the part that predicts regret.
20 minutes. We find where the money's going. No pitch.
Book a free Leak Check