The most expensive AI decisions we see aren’t the failed builds. They’re the custom projects that should have been a $40-a-month subscription, and the subscriptions that quietly never get used. Before you spend a dollar, four questions settle most of it.
The four questions
- Is this core to how you compete, or just plumbing? Buy the plumbing. Consider building only where it’s your edge.
- Does a trusted tool already do 80% of it? If so, start there — you can always extend later.
- Is your data ready? If the information the AI needs is messy or scattered, fix that first; it’s the real bottleneck.
- What’s the cost of waiting three months? If it’s low, wait — the tools get better and cheaper fast.
Most SMBs should buy first
For the majority of small and medium businesses, the right first move is to buy a proven tool, wire it into your workflow properly, and measure the result. Building makes sense when the capability is genuinely your advantage — and when buying has hit a real wall.
Build where it’s your edge. Buy where it’s table stakes. Wait where the cost of waiting is low.
We run this exact assessment in a short engagement: a prioritised shortlist, a build-vs-buy call on each, and a costed roadmap — so your next AI dollar goes to the thing that actually moves the business.
Wondering how this applies to your business? Tell us what you’re solving for — one short conversation, a straight answer.