Field hub
Claude listening for B2B buying decisions
Most positioning lines are drafted before anyone has sat with a buyer long enough to hear how a purchase actually stalls. dymtrim keeps a public shelf of claude field guides for teams who sell to other companies: how to interview decision-makers, watch a real touchpoint, name friction, and write a positioning sentence only after the notes can carry it. A small set of conversations is not a representative sample. A vivid story is not a return on investment.
Field conventions, not market statistics
6–12
conversations before a first claude pattern memo is worth drafting
4
roles inside a buying group that a single champion cannot replace
0
ROI figures allowed inside a claude synthesis, however tidy the story
Context
Why the interview outlasted the score
Serious writing on industrial purchases has, for decades, described a group rather than a shopper. The buying-center literature treated risk, several jobs, and internal argument as the thing to be studied. Account plans later turned that picture into a sales ritual. Voice-of-customer workshops collected language, and survey programs then compressed it into a score that was easy to brief and hard to audit. A claude memo runs the other way. It keeps the buyer’s verbs, the week the file sat still, and the person who never appeared on the organization chart but could stop the signature.
dymtrim uses claude as the name of that routine: a long interview with someone who holds a real role in the purchase, a page of observation from a scene the team actually attended, and a written close that refuses two temptations. The first is to call a handful of conversations representative. The second is to translate a story into a promised return. Start with decision-maker interviews and the limits in pattern finding. Market observation can describe a scene you attended. It cannot describe the market as a whole.
Three instruments
Look at the purchase before you name the offer
01
Decision-maker interviews
Reconstruct one real purchase with someone who can fund it, block it, or carry it through procurement. The hour is for their timeline, not for your deck.
02
Touchpoint mapping
Follow the handoffs where the decision is actually handled: a peer referral, an RFP copied from the incumbent, a security review, a redline, a pilot, a renewal.
03
Positioning synthesis
Write the line last. It names who the offer is for, the job, the alternative already in use, and only the reason the claude notes support.
Working sequence
A plain order of work
A claude sequence is deliberately unadorned. Stand somewhere a purchase is already in motion and write what people handle, who waits, and where talk goes quiet. That is observation, continued in the touchpoint map. Then interview two people who do not share a job title and may not share a conclusion. The shape of that conversation is set out in interview design. Only after both memos exist do you list friction: risk, internal politics, switching effort, missing proof, unclear ownership. Read that list as description, using the distinctions in purchase friction. Positioning is the last page.
Scene
One live touchpoint, written the same day.
Disagreement
Two roles, two memos, no forced consensus.
Line
A positioning sentence that cites the stall.
Shelf
Where to read next
Patterns
Finding patterns without fake statistics
Saturation, language clusters, and the negative case that keeps a claude memo honest.
Terms
Glossary of claude research language
Buying group, friction, handoff, saturation, and the claims these guides will not make.
Craft
Designing an interview that can be believed
Question arcs, probes, silence, and recording consent under United States state law.
Scenes
Observation that knows its limits
What a booth aisle, a lobby, or a renewal call can show — and what it must not be asked to prove.