This page describes the module as it runs today. Each customer’s usage facts — active days, events, active users, session length, last use and the change over the last two weeks — are live. Usage health — the score from 0 to 100 for each customer — and everything built on it: its tiers, the usage patterns, revenue in healthy usage, customers on track at month 6, and the alerts, is designed and not calculated yet. Your usage data does not yet carry three of the five things the score is built from, and how each part turns into a score is not yet written down. Usage intelligence — how it is calculated sets out every rule the live figures follow.
What it does
A customer can pay every month and barely use what it bought. That usually shows in its use of your product weeks before it shows in billing. Usage intelligence is built to score how well each customer uses your product against customers like it, to name the shape its use is taking, and to flag a stalled start, falling use or a new team early. Today it does the part that needs no score. Once a day, after your product analytics has loaded, it:- Lists every customer that sends usage — a customer in your billing whose account in your product analytics is matched to it — and says how many of your customers that is.
- Reads last full month’s use for each one: how many events, on how many days, by how many people, and the typical session length.
- Notes the last day each customer used the product.
- Compares the last two weeks with the two before: use rising, about the same, falling, or none in four weeks — and how many customers’ use halved.
- Groups the customers — all of them, each size group and each segment — with the middle number of active days in each group, how many customers sit in each band of active days, and their monthly revenue from your billing.
What it produces
Its headline is four figures. None is calculated yet.
Usage pattern links to its definition, which also says how the Data API and MCP will carry it — neither serves it yet. The other three get their definitions when they are calculated.
Behind them it also produces:
The change over two weeks is shown on the page only. It is not a published value: what Beacon publishes about a customer’s direction of use is its usage pattern.
How usage health will be worked out
The figures produced today are counts and sums of what your product analytics and billing record, worked out by fixed, written rules — the same data always gives the same figures. The full rules are in Usage intelligence — how it is calculated. Usage health — designed, not calculated. Each customer is scored on five parts, each from 0 to 100 against customers in the same segment, then weighted and added up. Against its own segment means the same number reads the same everywhere: eight milestones by month 3 can be behind for an Enterprise customer and on track for a small one.
Tiers. Strong — 80 and over. Adequate — 55 to 79. Concerning — 30 to 54. Critical — under 30, passed to Churn risk and Cost to serve. The Critical line at 30 is the same in every growth plan. The Concerning line moves with your growth plan:
Usage patterns. Each customer’s use over the last six months is matched to one of eight shapes: strong start, then builds · spreading to more teams · fast start, then flat · slow and steady · late bloomer · one heavy team · peaked, then falling · stalled onboarding. A new pattern is called only after it has held for 14 days. Under 60% sure, the customer reads No clear pattern yet and goes to its owner. The eight are fixed. See Usage pattern for each shape and the pairs that are easy to confuse.
Usual milestones per segment — the starting values, replaced with your own once 50 customers per segment have finished their first year:
The rules.
- When a customer is scored — after 7 days of usage, once its record passes Data integrity. The first months carry an early label; the score settles from month 3.
- Onboarding stalled — milestones under 40 once a customer reaches month 4. This line is fixed: loosening it needs the VP of Customer Success, the Head of Product and Beacon.
- Changing the weights — at the quarterly review, a change of up to 10% to one part applies by itself; more needs the VP of Customer Success, the Head of Product and Beacon. Milestones stay between 25% and 40%, no part reaches 40%, and the five add up to 100%. One change a quarter.
- No decisions here — Usage intelligence flags; it never acts on a customer. Churn risk, Expansion readiness, Cost to serve and Lifecycle economics use its readings and apply their own rules.
What you see on its page
In Beacon, open Customers under Growth and choose Usage intelligence, under Keep & grow. The banner shows, for each size group, how many customers send usage and their middle number of active days last month; how many of your customers send usage and their monthly revenue; that 2 of the 5 parts of usage health can be read; and how many setup answers scoring still waits for. The page has six tabs:- Overview — what your usage data shows today for each of the five parts; customers sending usage, active days last month and the last two weeks; the six checks of customers to act on and what each waits for; the eight patterns; how usage health is built; and what the module reads, feeds and where it shows up.
- Dashboard — the module’s five charts — usage health by tier month by month, usage health by segment, patterns by segment, stalled onboardings against the usual rate, and new teams starting — each saying what it waits for; then two charts from your real usage: activity over the last two weeks by segment, and active days last month by segment.
- Signals & alerts — what the module watches, its six alert rules and what each waits for, and that none is sending.
- Scorecard — the five parts and their weights, the tiers, how your growth plan moves the Concerning line, the eight patterns, the usual milestones per segment, and the rules.
- Customers — every customer sending usage, filterable by size, industry and the last two weeks, with its segment, active days and active users last month, the change over two weeks and monthly revenue. Usage health and pattern read Not yet calculated. Select a customer for what its usage data shows and what each part can read.
- Setup — the module’s six questions, how your usage connection is matched, and the two company questions it uses. Four answers are saved from this tab.
What it needs
When you connect PostHog, you confirm which PostHog group is your customer and which field it shares with Stripe. That is how each PostHog account is matched to a customer. See Connect your tools for how each connection is made.
Reads from
Which modules read it
Cost to serve, Customer value and Forecast confidence read the live usage facts today, and the Churn risk page shows the change over two weeks. Everything else reads usage health, its parts or its pattern, which are not calculated, so it receives nothing from this module yet. The Modules page in The Brain shows which modules are running for you.
Alerts
Six alert rules are built.
No alert is sent yet. Two wait for your milestones, one for teams in your usage data, one for the score, one for the score and Expansion readiness, and Usage falling waits for the score — its check against the month before needs two full months of usage. A separate signal for a customer whose use has stopped is being added to Signals. Who gets the alerts is picked from your Beacon users in Setup. Usage intelligence never contacts a customer or changes anything in your tools.
Setup questions
- Which events mark a new customer’s milestones? The most important setup decision for this module — best answered with your Head of Product. Recommended: start from Beacon’s list for your kind of product. Other answers: pick events from your PostHog; later, scoring without milestones for now.
- Which features count as using the product fully? Recommended: pick the features from your PostHog events. Other answers: start from the events your customers use most; later.
- How can Beacon tell teams apart inside a customer? Recommended: a team or workspace property on your PostHog events. Other answers: departments on contacts in HubSpot; you sell to one team per customer.
- Where is the number of seats each customer pays for? Recommended: seat quantity on your Stripe subscriptions. Other answers: on the contract, uploaded; you don’t sell by seat.
- How long should a usage pattern hold before Beacon calls it? 14 days. Fixed in Beacon’s method, shown so you know.
- Who gets the usage alerts? Your customer success lead and Head of Product, picked from your Beacon users. Every alert also goes to the customer’s owner in HubSpot. Waits for your users.
What is not calculated yet
- Usage health for each customer, its five parts and its tier.
- Revenue in healthy usage, customers on track at month 6 and usage patterns — patterns also need six months of usage history.
- Milestones, features and teams — not in your usage data until Setup is answered and your events carry them; paid seats are not read.
- Month against the month before — needs two full months of usage.
- Alerts — none is sent yet.
- Usage pattern, active users and last active date through the Data API and MCP — defined on the account object, not served yet.
Related
Modules
The ten module families and which modules are available today.
How it is calculated
Every rule the live figures follow — who counts, the last full month, the two-week windows, active days and the groups.
Usage pattern
The eight shapes a customer’s use can take, and the pairs that are easy to confuse.
Account object
The usage values Beacon publishes about each customer: usage pattern, active users and last active date.
Signals
What changed for each customer — including usage dropping.
Customer fit
Whether this is the right customer — usage like similar customers is part of product fit.