6 min read
Adoption beats capability, and Tuesday afternoon is the test
The best system is the one somebody opens when nothing is being demonstrated. Designing for that is a constraint on the build, not a smaller version of it.
- advisory
- build
Blog
What we think
The reasoning behind what we recommend, including the parts that argue against spending money with us.
6 posts
6 min read
The tool nobody uses is the most expensive one
Software costs what you paid, plus a year of your team working around it. The second number is usually the larger one, and nobody puts it in the comparison.
- advisory
- pricing
5 min read
You already tried a tool for this. That is useful evidence
A tool your team abandoned has recorded the requirement nobody wrote down. Read it before you buy the replacement, because the replacement usually fails in the same place.
- advisory
- build
6 min read
Most people who ask us for AI want automation
The difference is not academic. It usually decides whether a project costs half of what you were expecting, and whether it still works in a year.
- advisory
- automation
- ai
5 min read
Buy, build, or integrate — a decision procedure
Buy when your process is ordinary. Build when the process is the advantage. Integrate when it is both, which it usually is.
- advisory
- build
5 min read
Why nobody will give you a price for custom software
Anyone who quotes before they know four specific things is guessing, and the guess is either padded to cover the risk or low enough to become an argument later.
- advisory
- pricing
Got a version of this problem?
Describe it and get the honest read on whether it needs AI, automation, or nothing at all.