Glossary
Shared language is part of the method. If “friction,” “sample,” and “positioning” slide around, the readout will sound stricter than the notes. The entries below are the usage on dymtrim.cfd. They are not a claim that the wider industry has voted on a definition.
The name dymtrim uses for a qualitative routine: long interviews with people who hold a real role in a B2B purchase, direct observation of touchpoints, written memos, a search for friction and repeating language, and a positioning line that stays inside what those notes can support. Claude listening does not estimate market size, does not claim a representative sample, and does not forecast a return on investment.
A person who can commit budget, block a vendor, or carry a recommendation through the path that ends in a signature. The title on a badge is not decisive. The test is what they could do in the last real purchase. Practice notes are in decision-maker interviews.
The set of jobs involved in one purchase: economic buyer, champion, user, blocker, procurement, and whoever else the case actually contained. It is not a persona and not a committee that meets as a single mind.
The person spending internal reputation to move a file. A valuable witness to effort, and a poor sole witness to doubt. Always pair with another role before you synthesize.
A contact where the purchase is handled: a referral, a form, a review, a pilot, a renewal. Not a stage in the vendor’s funnel. Mapped in touchpoint mapping and watched under the rules in market observation.
The moment responsibility for the file moves from one role to another. Stalls often live here rather than inside either role’s stated opinion.
A structural pause: risk, politics, switching effort, a proof gap, missing ownership, or timing. Not a synonym for “objection,” and not a defect the research is obliged to remove. Types are separated in purchase friction.
A same-day write-up that separates quotation, paraphrase, and inference, and that ends with a sentence on what the case cannot support. The unit of claude evidence.
Repetition of a phrase, stall, or workaround across cases, fenced by at least one negative case. Not a rate. The working rules are in pattern finding.
A case your draft pattern predicts and that the notes do not fit. Kept in the record as a boundary. Deleting it to tidy a theme is a method failure.
The point at which new, deliberately different recruits repeat what you have already written and do not break it. It is not a quota and not proof that the market has been covered.
A draw from a defined frame, designed so that statements about the frame can carry a stated uncertainty. Claude memos are not samples. Do not use “representative,” “most buyers,” or a bare percentage to describe them.
A fenced sentence: for whom, which job, against which alternative, for which reason the notes support. Written in positioning synthesis. If it cannot lose an unfit account, it is not positioning.
Whatever already does the job, including a spreadsheet, a contractor, an incumbent module, or inaction. Often not the competitor named in a battle card.
A conclusion you drew, marked as yours. Inferences are allowed in internal memos and dangerous in customer-facing lines unless a later case tests them. “They hesitated because of price” is often an inference wearing the costume of a fact. The interview guide in interview design is built to slow that costume change: sequence first, theory second.
Both can be honest. They answer different questions and must not share a vocabulary of proof.
| Question | Survey, properly designed | Claude memo |
|---|---|---|
| What it can describe | Distribution of answers in a frame | How a particular purchase unfolded |
| What “more” means | A larger, still-defined draw | Another role, a negative case, a watched scene |
| A responsible number | An estimate with a denominator and a method | A count of cases you actually have, labeled as such |
| A misuse | A leading item written to confirm a slogan | A story told as if it were representative |
If a board asks for incidence, commission a sample and keep its report separate from these pages. Mixing the two genres is how a careful interview becomes a fake chart. The home of the method, including the field conventions that are explicitly not market statistics, is the hub.