The startidea notebook
Better questions.
Better decisions.
Ten field notes on choosing tools, testing claims, and finding work worth doing.
- 01Construction ↗
Construction robots: start with the job, not the robot
Layout, material placement, and inspection solve different problems. Put them in different shortlists.
- 02Construction ↗
Layout accuracy is not the same as time saved
A useful field trial measures the path from an approved model to accepted marks—not just printer speed.
- 03Field tests ↗
A robot pilot needs a cost per accepted unit
An illustrative calculation that includes supervision, setup, rework, and the work the robot leaves behind.
- 04Developer access ↗
“Unknown API” does not mean “no API”
Separate missing evidence, an integration claim, and a documented interface before making an architectural decision.
- 05Procurement ↗
Enterprise readiness needs an evidence ladder
A claim, a document, and a tested control answer different questions. Keep them separate.
- 06Procurement ↗
An SSO demo is not an offboarding test
Signing in is the easy demonstration. Removing access is the more revealing test.
- 07Developer access ↗
Read cached research before starting another search
A small client workflow that avoids unnecessary provider work and handles interrupted streams honestly.
- 08Research methods ↗
Read a company map without inventing a ranking
Position can summarize evidence. It cannot tell you which vendor is best for your workflow.
- 09Customer tests ↗
Turn a market gap into a demand test
“Nobody seems to do this” is a research lead. A measurable buyer commitment is a stronger next step.
- 10Research methods ↗
A vendor claim needs an evidence trail
Save the statement, source, scope, and date so the next reviewer can challenge your conclusion.
Written with AI assistance from cited public sources and the startidea implementation. Illustrative exercises are labelled; these notes do not report hands-on vendor testing.