---
name: what-the-account-already-told-you
description: "Load when an account manager or CSM pastes a customer's record (footprint, usage, tickets, stakeholders, catalogue, any upstream risk tier) and asks for expansion documents: a whitespace brief plus, where asked, a mutual action plan or buyer page. Ranks only what the customer has asked for or repeatedly absorbed, clears printed prerequisites first, puts repair before expansion on a wounded account, and returns questions instead of tiers when the record is thin. Not for new-logo prospecting, renewals, or save plans."
---

# The customer has already told you

Almost everything you need is already on the record, said or done by the people who work there. Somebody asked for a thing. Somebody kept absorbing the same failure instead of raising it. A whole department went a year without mentioning anything at all. An expansion recommendation is that testimony, gathered up and given a consequence — which is why it holds when it is read back in the room.

What does not hold is a page built from things nobody said. Staff numbers, sector, contract value, the shape of a reporting line, the price list itself: all useful for reading the record, none of them grounds for a claim. They are also plentiful, while real grounds are scarce, so the pull towards them is constant — and the resulting page is wrong in the way a customer notices immediately, because they know their own company and you are guessing about it.

## Four ways the record speaks, in order

**They asked for a capability, in writing.** Somebody in the team that would run the product put the request on the record. The scarcest thing in any account, and the only one that carries a recommendation by itself; name them and quote them. Weigh the words: a person who wants better visibility, more engagement, more scale has named a destination rather than a capability, and it stays a destination even when the price list holds something built to reach it. Between a named capability and a named destination, the capability wins — and it wins even when the destination came from someone more senior, because the person who runs the job is the one whose request can be acted on. A fresh request beats a stale one on the same terms. And a customer telling you that *their* customers keep asking for something has reported a decision taken at another company: it corroborates, it does not request.

**They kept absorbing the same failure.** Repeated tickets on one job, from the team that owns it, against a product built for that job. Behaviour rather than intention — the cost is on the record and nobody has asked to spend. It carries the opportunity that stands behind the recommendation, written as the problem it really is, with the job named and the one question that would turn behaviour into a request.

**It joins onto something already installed.** A small-commitment link whose worth follows from what they run today: no new sponsor, no new job, no budget conversation. Say it exists; never open on it.

**A title, a team, a staff count, a sector, a contract value.** None of these is anything the customer told you about your product. A name on a stakeholder list establishes that a person exists — not that they want something, control a budget, or hold a view. A function that every company this size must have is not a function with a problem. Everything here is an open question rather than a rank: name the product, name what is missing, name what would settle it and who could say. No confident adjective may put the claim back in by the side door.

## The org chart is not a source

A stakeholder list looks like grounds and is mostly seating. Somebody who spoke to you lately about their own work is a witness, and what they said carries its date, because trouble moves faster than a reporting line. Somebody who approves spending but has not sat with you in the better part of a year is a limit rather than a backer: decisive for timing and for the size of the ask, and not a person whose intentions you are entitled to supply. Somebody whose team appears in the org chart and nowhere in the usage, the tickets or the notes is an open question, and the honest form of it names what would have to be found and who could find it. A title with a likely need reasoned onto it is nothing at all, however senior.

One test settles all four: could you say the sentence to that person's face? *You asked for this in July* holds. *Someone in your seat usually needs this* does not, and a page built out of sentences like it falls apart in the meeting it was written for.

Public news about the account (a funding round, a hire, an acquisition) reads the same way: context for the record, never a request from anyone in it.

## Only what was ordered

**Fix the list of documents from the request before the account is read.** A plan gets written when a plan was asked for, and a page for a buyer when that was asked for. An economic buyer appearing in the stakeholder list is not a request for a page addressed to them, and *brief only* means brief only, however good the opportunity turns out to be. If you are unsure whether something was asked for, it was not.

**A heavier record does not buy a longer answer.** Three contacts or thirty, the reader has the same few minutes and needs the same few things; a heavy record means more that stayed off the page, not more page. One recommendation, no more than two ranked behind it, every remaining catalogue line dismissed in a sentence. Walking a reader through eight products is a price list with opinions attached.

## Cleared before ranking

**Printed prerequisites.** An executive sponsor, a champion from a named function, a dependency, an installed base above some size — cleared first, never footnoted afterwards. A product whose printed condition has nobody behind it is not rankable at any level, and the unmet condition is the finding. It cuts both ways: something built for a scale this customer is nowhere near is out now, on the price list's own terms.

**What their use of what they bought is telling you.** Take-up far under what is being paid for, escalations open past excuse, satisfaction sliding two quarters running, a sponsor gone, a renewal in doubt. Where any of that holds this is not an expansion cycle, and the answer's first line says so rather than its last. Something that would fix the problem is repair, described as repair and not sold as growth. A genuine request stays genuine when the timing is wrong: hold it as the first conversation once the account is steady. Where the damage is serious, refuse the expansion plan outright, name what takes its place, and give the two or three things that must become true first — attaching a warning and shipping the plan anyway fails, because the warning does not travel once the document is forwarded. Where the risk is contained, one product often does both jobs; say so, and put the repair first.

Where usage, contacts and history are all thin, rank nothing: open by saying which documents cannot honestly be produced and why, and let a short set of questions be the deliverable, each naming who answers it and what the answer would settle. A ranking hedged with low confidence is still a ranking. Where the catalogue already sits with the teams that would use it, name that coverage as the finding and rank only what genuinely remains; adding a line so the page has three is the org-chart mistake in better clothes. Where a signal reads two ways — a workaround they built and maintain, heavy use alongside complaints about the same area — put both readings where the reader meets the signal, then settle it with a reason, often that both hold at once and the position is repair and expand together.

## What a plan is for

It belongs to the customer, for an outcome they want, written so their own side can hold their own people to it. Every step is something they will have rather than something you will do, and the objective is stated as their business result. One recommendation, one plan.

Size it by counting what has to happen — settle what is wanted, put it in, sign for it — then a step for each real layer: money held by somebody absent from the room for months means the case has to be made to them; anything that must be proved before a signature would be credible is another; an open escalation is another, and it goes first. Beyond half a dozen, steps are being invented to look diligent.

Every step carries a date, a named owner on each side, and something a person could go and look at: a written confirmation, a signed scope, an output built from their own data, a figure reached inside a system. *Alignment achieved* and *the sponsor is on board* are hopes in the costume of criteria; where no name on your side is given, use the role rather than inventing a person. A declared position on spending — a freeze, a cost-control stance, a wait for the next cycle — is a date and not a verdict: it belongs in the plan as a dated step naming whoever declared it and what they will have seen by then, never as background and never as an open question at the foot of the brief, because a constraint you have scheduled around is handled, while the same constraint raised as a question reads as a deal nobody has thought through. Unknowns get named rather than filled, and where an answer decides whether an opportunity exists at all, that question is the plan's first step.

A buyer's page, where one was asked for, is written from that chair: the problem their organisation has, then what changes, what it costs, what they are asked to do. None of your internal vocabulary appears in it.

## Three answers, at size

Write as a manager briefs a colleague about to stake their own standing on it: plain, ordered by consequence, everything said once, prose instead of a lattice of labels. Headings belong to the account; every figure comes from the record. The answer starts at the heading of the first document that was ordered, with nothing in front of it — no sentence announcing what is coming, no repetition of the request back to whoever wrote it.

The working never reaches the page. No section called analysis, no march through the four ways the record speaks, no run down the catalogue clearing prerequisites one at a time, no separate verdict on account health, no inventory of signals ahead of the documents, no list of what the customer did not say. That is the shape of the thinking, and what the reader wants is its result, carried inside the documents where each claim sits beside the thing it is about. When a risk tier, QBR brief or save plan from an earlier step is on the record, that work is finished: use its tier and its open faults as given and do not re-derive them.

Size follows from that. The recommendation takes two sentences; anything ranked behind it takes one; anything declined takes one naming it and the prerequisite it fails; an open question takes one. Where a plan or a buyer's page was ordered as well, those all drop to a clause each, because whatever follows the brief has to reach the same reader inside the same page.

Three answers below, all invented and deliberately spare — carry the proportions, never the sentences. The first ordered a brief and nothing else, so the brief is the whole of it.

> ## Ridgemont Poultry — one opportunity, one question
>
> Nadine Ashby, who runs the Fenwick packing hall, asked in August for chill-chain readings to be logged against each crate as it moves rather than written up at the end of shift — the only capability anyone here has asked for.
>
> **Crate Logging — recommended.** Ashby named it and she runs the hall. Ridgemont gets a temperature record their retail auditors accept, and an end to the two hours a night the shift leads spend copying sheets.
>
> **Route Sequencing — behind it.** Eleven tickets since May about loads picked in the wrong order, nobody asking to buy; one question to Ashby settles which it is.
>
> **Not ranked.** *Yard Booking* — its entry needs a logistics sponsor, and none is named. *Retail Portal* — built for suppliers above forty accounts; Ridgemont serves eleven.

The second ordered all three: a brief, a plan for the recommended opportunity, and a page for the buyer. Different account, and it is not the longer of the two — notice what the brief gives up so that what follows it arrives.

> ## Kingsley Ceramics — recommendation and plan
>
> Ivo Brandt, who runs the glaze line, asked in September for glaze batches to be held against each firing instead of in the kiln log — the only capability asked for, and seven tickets about firings re-run from memory say the same in behaviour. **Batch Control — recommended**, and he owns the line. *Behind it:* Shade Matching, on those same tickets, once batches are held. *Not ranked:* Distribution Planning, needing a commercial sponsor Kingsley has not named; Group Analytics, built above four plants against their two. *Open question:* whether the Marden plant fires to the same glazes, which decides one rollout or two.
>
> **Plan — glaze batches held against firings before the spring run**
> - 14 Oct · Brandt and our solutions lead settle the batch list and what counts as a firing — written scope, signed by Brandt.
> - 4 Nov · Three glazes running against Kingsley's own 2025 firings — output reviewed on the call.
> - 25 Nov · Brandt and Finance Director Peter Vane see the re-firing hours gone — a figure both sides accept.
> - 12 Dec · Vane signs the order.
>
> **For Peter Vane.** Kingsley's glaze record lives in a kiln log that goes missing, so firings get re-run from memory and you cannot show a customer why a shade moved. Holding batches against firings closes that at £26,000 a year against roughly two hundred glaze-line hours now spent repeating work; you are asked to decide in December, once those hours have come out of your own 2025 figures.

The third ordered a brief and asked what should happen next. That account was in trouble, so the honest answer refuses the plan it was angling for.

> ## Farrow & Deane Joinery — repair before anything else
>
> Two faults reported in March are still open, both raised again in May; the survey score has fallen two quarters running, and the manager who signed the contract left in April.
>
> **Hold the renewal first.** Get both faults closed to the satisfaction of the supervisors who logged them, and meet Ines Barral, who took over operations, before the September review — no contact so far, and the budget is hers.
>
> **What is not happening.** No plan for Edge Banding, even though Barral's shift lead asked for it in February. That is the conversation to have once the faults are shut, and raising it while they stand open is what costs the renewal.

Two failures end this work, neither of them analytical: a page carrying documents nobody ordered, and a page that outran its reader, so that whatever came last was delivered to nobody.