The twelve questions (printable, one pass)
Skip the vendor deck for an hour. Answer in plain language — enough to decide whether to go deeper, not a procurement finish line.
Fit and the real process
- What is the system of record for the thing that matters (customer, job, order, line item)? If two systems both own the number, a new box will not fix it — map the flow first.
- Can we name who owns the next handoff? Nobody and IT are not job titles. See clear ownership for system work.
- Where do the top three Tuesday exceptions live (credit, split, return, we always fix it in a sheet)? If the package’s happy path has never met them, the demo is theatre.
- Is the process stable enough to automate? If the real process changes every month, bespoke vs script is a moving target; stabilise the human flow or accept maintenance cost.
Product, cost, and integration
- How far does configuration get you (fields, workflows, rules) before you are asking for a code fork or a permanent custom module? That line is the good enough boundary in the next section.
- What must integrate, and is the path vendor-supported (API, supported connector) vs a fragile sync nobody will own at 2 a.m.?
- What does a three-year total cost picture look like — per seat, per feature tier, implementation days, and your internal time? Compare the same time horizon, not a headline per month alone.
- If we outgrow the product, can we leave? Export, contract, and API in writing before signature — not a slide in week four of a rescue.
Risk, timing, and honesty
- What is the cost of it going wrong if this fails at go-live — in cash, margin, and reputation? That sets how much rigour you can afford; tie it to simple ROI levers.
- When is the first week the team is genuinely better — not go-live, but a measurable win on time, rework, or speed to cash?
- Is the gap discipline and data or fit? If the gap is ownership and hygiene, a build is a second broken process — name that before you sign.
- If bespoke: what is the thinnest slice that would win? A thin custom layer, one integration, or a full build — three different budgets and maintainers. Honesty here saves six figures of phase two surprise.
Book a free systems review to work through a slice in a workshop tone — or use our offers for a process audit before a build, custom build when the list says so, and care when something is live. Teesside, the North East, and further afield when the fit is right.
What good enough from a package looks like
Configuration covers your edge density for the next 18–24 months: not every weird history from the last decade, but the repeated Tuesdays the team can script with rules, fields, and a supported integration. The vendor’s roadmap does not need to be your dream roadmap — it needs a predictable, boring path for the next couple of renewal cycles.
Good enough is not compromise in silence; it is a named list of what you are not solving in the box so phase two is a decision, not a surprise backlog after go-live.
When bespoke is the honest answer
Bespoke (or a focused custom layer) is the straight answer when the way you win is in the flow itself — a contract model, a product combination, a regulatory evidence trail, or a rare workflow that is the product — and when off-the-shelf would force a permanent lie in the data, or a re-key every week with no end.
Budget build and maintain — a line item for a person or partner who is not whoever is free next month — and read is custom software worth it? and operational technical debt before you green-light a rescue-shaped project. Book a call to pressure-test the list, or see case studies for what honest delivery can look like.