> ## Documentation Index
> Fetch the complete documentation index at: https://docs.beaconrevenue.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Customer value

> A 0-100 modelled measure of what an account is worth to you across revenue, margin, strategic weight and risk, its four dimensions, and the five tiers it classifies into.

## What it is

`customer_value` is a 0-100 measure of what an account is worth to you, where a higher number means greater value. It is forward-looking — it describes the value ahead of you, not the revenue already booked — and it is published for every active account.

It is **modelled**, not measured. It weighs four different kinds of worth against each other, and it is calibrated against what comparable accounts turned out to be worth in reality.

**The composite is never the whole answer, and Beacon does not present it as one.** Two accounts can both score 70 and be entirely different businesses to you — one carrying strong revenue on thin margin, the other the reverse. The four dimensions behind the score are published alongside it, and it is meant to be read with them.

The shape you get depends on where you read it. On the read interface and MCP tools you get the exact score, with its as-of and the version of the model that produced it. In your CRM it lands as a band carrying the tier rather than the number, because a CRM field strips the definition, confidence, version and freshness that make a modelled number safe to read.

**This value carries a higher ceiling than the rest of the account object.** It is available on the read interface and MCP tools under a grant that covers it, and it is **off by default for writing into your CRM** — turning it on takes an opt-in that states who becomes able to see it. That is a rule about where the value may travel, not about who inside your company may look at it. Account owners see it on their own accounts.

## How it is calculated

Modelled. Four dimensions are scored 0-100 and combined at fixed weights.

| Dimension       | Weight | What it reads                                                                                                                                  |
| --------------- | ------ | ---------------------------------------------------------------------------------------------------------------------------------------------- |
| Forward revenue | 0.30   | Recurring revenue expected ahead, expansion revenue weighted by readiness, and the quality of the contract structure underneath it             |
| Forward margin  | 0.30   | Lifecycle economics, cost to serve inverted so that cheap-to-serve counts positive, and realised gross margin once finance is connected        |
| Strategic value | 0.20   | Segment standing, breadth of adoption across teams, marketing signal, and a place for a judgement your leadership makes rather than the system |
| Risk            | 0.20   | Deductions for churn risk, stalled or cyclical progression, worrying adoption patterns, irregular payment, and a wide confidence band          |

The four weights always sum to 1.00, and hard limits hold them apart: no single dimension may carry 0.45 or more, strategic value may never exceed 0.30, and forward revenue, forward margin and risk each have a floor. Your growth plan then applies a light overlay to the result, within a few percent either way.

| Tier      | Score                                                              |
| --------- | ------------------------------------------------------------------ |
| Strategic | 90-100                                                             |
| Premium   | 70-89                                                              |
| Standard  | 45-69                                                              |
| At risk   | 30-44                                                              |
| Critical  | Below 30 — **or** a risk dimension below 30, whichever comes first |

**The critical floor has two conditions, and either one is enough.** An account at 45 with its risk dimension at 25 is critical on risk, not on the composite. This is deliberate: a composite alone can average away a structural problem, and the risk floor stops it doing so.

**Where the tiers sit moves with your growth plan, and two points never move.** Premium sits at 65, 70 or 75 depending on how aggressive your plan is, and the at-risk line moves with it. **Strategic at 90 and the critical floor at 30 are fixed under every plan.**

Every score is published with a confidence label — high, medium or low — taken from the weakest input weighted by how much it contributes. Confidence is not decoration: a wide confidence band is itself one of the risk deductions, so a thinly-evidenced account scores lower for that reason and says so.

## Where it comes from

Customer value is composed from values Beacon already holds rather than read from a source system, so what it needs is other parts of Beacon rather than another connector.

| Dimension       | What has to be working                                                                             |
| --------------- | -------------------------------------------------------------------------------------------------- |
| Forward revenue | Renewal and expansion forecasting, and contract records                                            |
| Forward margin  | Lifecycle economics and cost to serve. Realised margin joins once finance is connected             |
| Strategic value | Segment classification and product usage. Marketing signal joins later                             |
| Risk            | Churn risk, lifecycle progression, usage patterns, and payment behaviour once finance is connected |

**It does not run at all until the customer side of Beacon is live.** Before that there is no partial score and no estimate — the value is simply absent, and segment-level comparisons fall back to cross-segment reference patterns, labelled as such. An account is admitted to scoring once its identity is resolved at a confidence of 70 or above.

Beacon names what is missing rather than quietly narrowing the calculation. A dimension short of its inputs is marked as partial; a domain not yet switched on is marked as awaiting activation; margin before finance is connected is marked as partial margin; an unresolved disagreement between sources is marked as pending reconciliation. Where confidence has dropped below the line, the score is still shown — what is held back is the automatic action it would otherwise drive.

## How fresh it is

Recomputed in full on the first business day of each month for every active account, and partially within the hour whenever a material upstream value moves — a change in lifecycle economics, a forward-revenue shift above 15%, a cost-to-serve tier change, a churn or expansion tier change, or a lifecycle stage transition.

**A tier change is not declared the moment the score crosses a line.** A crossing must hold for 30 days before the move is announced, in both directions, because a tier change on this value reaches leadership. Crossing the critical floor is the exception and is announced immediately.

Each monthly recomputation is committed and held until the next one. The dimension weights, the tier thresholds and the disagreement thresholds can change only at the quarterly calibration — that is the single window for it. A change of growth plan re-seals the tier thresholds outside that window, since it re-classifies every account at once.

## Currency and rounding

Neither applies. The score has no unit and holds no money value. Money figures feed it, in your single reporting currency, but the value that comes out is an index.

## What changes it

Movement in any of the four dimensions. A change of segment changes the baseline the score is read against. A change of growth plan moves the tier lines and applies a slightly different overlay, which can move the tier without anything changing at the account.

**When two parts of Beacon disagree about the same account, the disagreement is published rather than averaged away.** Strong lifecycle economics against a critical cost to serve, strong forward revenue against a high churn risk, expansion readiness against concerning usage, or two dimensions more than 30 points apart — each of these raises a flag that names the conflict, the dimensions involved and who should look at it. The flag never suppresses the score and is never folded into it. It tightens the confidence band, marks the account as pending reconciliation, and pauses any automatic expansion candidacy while the disagreement is being resolved.

**The model is recalibrated quarterly against what actually happened.** Accounts that closed are measured on what they were really worth over their first 18 months, and the tier baselines and dimension weights are reviewed against that. A small weight adjustment can be applied on review. A larger one takes joint sign-off from the customer, revenue and finance leads. The strategic threshold and the critical floor change only with your CEO, revenue and finance leads together. Nothing retunes itself, and no retuning happens more than once a quarter.

Between calibrations the model is watched for drift — whether it answers differently on unchanged inputs, whether it moves while its inputs have not, whether it starts contradicting other parts of the system, and whether it recomputes off schedule. A review can clear the value, hold it, or escalate it. A hold pauses consequential use; it does not change the score. Separately, a named owner — never an agent — can override the value used for one decision, with the reason recorded. That changes what the decision consumes; it never changes the model.

## What it is not

* Not lifetime value. Lifetime value is what an account turned out to be worth, measured after the fact. This is what it looks worth from here. Beacon holds both, and the realised figure is what recalibrates this one each quarter.
* Not recurring revenue, and not a size ranking. Revenue is one input inside one of four dimensions. A large account that is expensive to serve and structurally at risk scores below a smaller one that is neither.
* Not [`fit_score`](/definitions/fit-score). Fit is how closely an account matches who you sell best to; value is what it is worth to you. They move independently, and where a high-fit account sits at a weak value tier, that mismatch is itself one of the flagged disagreements.
* Not [`health_score`](/definitions/health-score) or [`churn_risk`](/definitions/churn-risk). Both are measured and reproducible from your data, and both feed the risk dimension here. This value is modelled and covers more than risk.
* Not comparable across segments. A 70 in one segment and a 70 in another are measured against different distributions. Compare an account with its own segment, not across.
* Not a customer-facing score. This is your view of the account, not a rating shown to the customer, and it is never written into anything the customer reads.