Context: the missing layer in GTM AI
An architecture problem dressed up as a prompting problem.
Why prompts and agents fail the same way, and why enablement and revenue strategy should own the context that sellers and agents both draw on.
Most AI adoption in go-to-market is an architecture problem dressed up as a prompting problem. A prompt holds just enough context for one request, agent instructions hold it statically while the market keeps moving, and every team ends up maintaining its own version of the truth. The talk sets out a context architecture built the way enablement is built, around why we sell, how we sell and what we sell, and shows what changes when sellers and agents read from the same governed source: account research measured in thousands of hours removed, and meeting preparation cut from half a day to under an hour.
First given at the Sales Enablement Leaders Conference, London, September 2026.
A structure for their own context library, and a clear owner for it.
- Sep 2026 Sales Enablement Leaders Conference, London · Context: the missing layer in GTM AI Speaker
An example visual
An example of the kind of visual I use to carry the ideas in this talk, drawn one idea at a time. Each talk is shaped to the organisation and the forum, so the visuals on the day will vary.
For a talk, panel or podcast, send the event name, date, format and audience to ian@illogicaldata.com.