How to work a book from signalinstead of from a list.

A composite score is a single number standing in for thirteen weekly readings. Two accounts can post an identical 62 while one has climbed every week for a quarter and the other has fallen from 94 since April. The number cannot separate them. The shape can.

Four shapes are worth a rep's attention. A staircase — steady week-on-week gains — usually means a funded initiative rather than a single event, and it tends to hold. A spike is a one-week jump with no follow-through; treat it as a reason to research, not to call. A plateau at a high reading often means the evaluation has moved somewhere you cannot observe, which makes committee coverage the priority rather than another sequence. A decline from a high reading is the clearest signal in the set, and it usually means someone else won.

The practical rule: sort by direction first, magnitude second. A 58 climbing for six weeks is a better call than an 81 that peaked two months ago.

Not all triggers are equal. A funding round is public, easy to detect and almost universally acted on by every vendor in the category, which means the window is crowded within days. A reporting-line change is quieter, harder to detect, and far more predictive of a specific purchase.

We group triggers into four bands. Band A — mandate triggers — are events that hand a named person a budget and a deadline: a new executive with a stated remit, a published RFP, a regulatory deadline with a compliance owner. Band B — capacity triggers — are events that break an existing system: an acquisition adding a second ERP, headcount crossing a licensing tier, a location opening. Band C — intent triggers — are observable research behaviour: category browsing, competitor comparison, review-site activity. Band D — context triggers — are ambient facts that qualify rather than time: funding, headcount growth, industry.

Band A and B are worth interrupting a rep's day for. Band C belongs in a weekly digest. Band D belongs in territory design, not in outreach.

Most lost deals are not lost on price. They are lost because a person who could say no was never in the room. Coverage is the number of committee members you have had a real conversation with, divided by the number the record says exist.

Work the gap by role, not by seniority. An unmet economic buyer is an escalation problem and belongs to the AE. An unmet technical evaluator is a proof problem and usually belongs to an SE. An unmet operational owner is the most commonly skipped and the most likely to raise a late objection, because they are the person who will live with the change.

The mechanic that works: ask your champion for a specific introduction framed around a question only that person can answer. 'Who would need to be comfortable with the migration path?' produces a name. 'Who else should be involved?' produces a shrug.

A contract renewal is the only moment an incumbent is genuinely vulnerable, and the window opens far earlier than most sellers assume. By the time a renewal is thirty days out, the budget has usually been re-approved and the switching decision has already been made by default.

Work backwards from the renewal date. At 120 days, the evaluation of alternatives is still politically acceptable. At 90 days, procurement can still be involved without an exception. At 60 days, you are asking someone to delay a renewal, which is a much larger favour. At 30 days, you are asking them to break something.

This is why the renewal window field carries days-remaining rather than a date. The number that matters to a rep is how much room is left, not when the paperwork was signed.

Enrichment vendors commonly present two very different things in the same visual style: a fact observed in a source, and a value assigned because similar companies tend to have it. The first is evidence. The second is a prior.

We separate them. A detection carries the artefact that produced it — the job posting, the MX record, the case study — and a confidence derived from how directly that artefact implies the fact. An inference carries the peer group it was drawn from and is always rendered as an inference in the interface.

The practical consequence is that a rep can tell, in the second before dialling, whether they are about to reference something the company said about itself or something a model assumed. That distinction is the difference between a credible opener and an embarrassing one.

Territories are usually balanced on account count or aggregate revenue potential. Both assume every account is equally addressable, which is the assumption signal data exists to break.

Signal density — the share of accounts in a patch that posted a Band A or Band B trigger in the last two quarters — is a better balancing input. Two patches with identical account counts can differ by a factor of three on density, which means one rep is working a live book and the other is working a directory.

Rebalance on density annually and review quarterly. Moving accounts mid-quarter destroys relationship continuity for a gain that rarely justifies it.