Most bad technology decisions aren't bad ideas — they're good ideas decided badly: bought on a demo, under deadline, without the questions that would have surfaced the real cost. A little structure up front is what separates a purchase you're happy with a year later from one you're quietly working around.

This is the checklist I run before any major technology decision — a new system, a platform switch, a big renewal, an AI rollout. Work top to bottom. If you can answer these honestly, you're deciding on purpose.

Start with the problem, not the product

  • What specific problem are we solving? Write it in one sentence, in business terms — not "we need a new CRM" but "we lose deals because follow-ups slip." If you can't state the problem plainly, no product will fix it.
  • What does success look like in 12 months? Name the outcome you'd point to. Vague goals buy vague tools.
  • What happens if we do nothing? Sometimes the honest answer is "not much," and that's worth knowing before you spend.

Understand the true cost

  • What's the all-in cost, not the sticker price? Licenses, implementation, integrations, training, data migration, and the internal hours to run it. The subscription is usually the smallest line.
  • How does cost scale as we grow? Per-seat pricing that's fine at 10 people can hurt at 40. Model it at your expected size, not today's.
  • What's the cost to leave later? Exit is part of the price. Check data export, contract length, and lock-in before you're committed, not after.

Test the fit

  • Does it solve the actual problem, or an adjacent one? Demos are built to impress. Map each feature you're paying for to a real workflow you have. Ignore the rest.
  • Who has to use it, and will they? The best tool your team won't adopt is worse than a decent one they will. Involve the people who'll live in it daily before you buy.
  • How does it fit what we already run? Integration gaps become manual work forever. Confirm the connections you need actually exist and actually work — don't take "it integrates" on faith.

Check security and risk

  • What data does it touch, and where does it live? Especially for anything customer-facing or regulated. Know what leaves your walls and who can see it.
  • What's the vendor's track record? How long have they been around, how do they handle outages and breaches, and what's their support actually like when something breaks at 4 p.m. on a Friday?
  • For AI tools: where does our data go, and is it used for training? Read this before you turn it on, not after. A governed rollout beats an enthusiastic one.

Decide deliberately

  • Did we compare at least two real options? One quote isn't a decision. Even a quick second option reveals what "normal" costs and where the first one is weak.
  • Is anyone advising us paid to recommend this? If the advice and the sale come from the same place, weight it accordingly. (More on that in 5 Signs Your IT Vendor's Incentives Work Against You.)
  • Can we start small and reverse it? A pilot or a short term beats a three-year commitment on an unproven fit. Buy the option to change your mind.

The short version

Name the problem, price the whole thing (including the exit), prove the fit with the people who'll use it, check where your data goes, and compare at least two options from someone who isn't paid to steer you. That's most of the value of an expensive consultant, on one page.

When the decision is big enough that you want it done right the first time, that's exactly what a technology assessment and roadmap or an ongoing Fractional CIO is for. Or take the IT Decision Scorecard to see how deliberately your technology is being run today — then book a call if you'd like a hand.