Product, events and billing
One milestone per product, anchored on the confirmed payment
Conversion, activation and time to first value by cohort
Dashboard, alerts and CRM
What it solves
- Trial conversion is measured where the trial actually exists, and per customer.
- Activation has one definition per product, which does not change when billing rewrites dates.
- New accounts with a high-potential signal reach the team the next day.
Business questions it answers
Ask your agent these questions once the case is live.
What you need
- Product database or events connected, with the trial start date
- Billing connected, with charges
- The list of activation milestones per product, agreed with product and CS
What you get
- A table of trials by cohort, with activation and conversion
- Activation rate and time to first value by product
- A high-potential alert in Slack or in the CRM
- An AI agent that answers with the right activation rule
Where to activate
Conversion and activation by cohort, product and channel, with partial cohorts flagged.
A daily Slack message with yesterday's flagged accounts and a summary on the first day of the month.
An activated trial with no sales contact becomes a task; a stalled free or trial segment goes to the email tool.
Finding the trial source and agreeing on each product's activation milestone takes longer than the calculation. Once defined, the metric stays stable. The detailed view has the business rules, the pitfalls and the acceptance test. Or send everything straight to your agent with the Nekt MCP.
In many SaaS companies the trial does not exist in billing, only in the product database. Whoever measures conversion by Stripe's trialing status finds almost no trials, and whoever measures by cancellation counts unconverted trials as churn. When activation is anchored on the billing cycle date, a card recovery pushes the date forward and undoes activations already earned.
How it is usually done, and the rule
Each row is a mistake that already happened in practice.
Before measuring, find each product's trial source: start and length in the product database (for example, 14 days from signup), the account's test dates, or billing's trial_start and trial_end. In a real case, billing had a single trialing subscription in the whole base, because the trial lived in the product. A product with no trial source stays out of the metric, and that is declared.
Activation is the first event that delivers value, and it changes by product: first reply received in a prospecting tool, first message sent through a connected channel, a creator's first sale on a content platform. Write the list in a single place, with the exact event name. A product with no defined milestone is null, never activated by default.
For post-sale activation and time to first value, the anchor is the first paid charge, not the billing cycle date. Billing rewrites the cycle date forward when a subscription recovers from a failed charge. In a real case, this undid activations already earned and pushed old accounts into the wrong cohort. After the fix, the earlier series was no longer comparable and was flagged.
An account enters the denominator only if the activation event already existed when it started (the event's birth date in the instrumentation) and the observation window has closed. An account that started before the event existed is null. With no eligible account, the rate is null, not 0%.
A converted trial is a customer with a paid charge after the trial start, on any subscription. A trial subscription canceled at the end of the period is not churn: it is an unconverted trial, and it stays out of the MRR movements table. The customer may convert on a new subscription, so measuring per subscription understates conversion.
Conversion and activation are read by trial start cohort (week or month), with a fixed window (for example, conversion up to 30 days after the trial ends, activation within 30 days). The cohort month is not the MRR month. A cohort with an open window shows as partial.
The same person or company opens another account to get another trial. Link accounts by normalized phone or tax ID and treat the second as a reopening of the first, inheriting its history. Deduplicate repeated emails and usernames in the trial base before counting.
In B2B with an onboarding project, activation is the product milestone met within a deadline counted from the sale, for example by the Saturday of the fifth week after the closing week, or connection within 25 days and completion within 90 days of opening the onboarding case. The final stage of the onboarding pipeline and a logged meeting do not prove activation.
An early signal of a valuable account triggers a single alert per account. Real example: among signups in the last 45 days, a first sale within 72 hours of signup, or 10 or more buyers within 2 days of the first sale. Accounts that already have an SDR, closer, CS or CRM record stay in the table but are left out of the message.
From new account to activated customer
Each stage with the source and date that define it.
Definitions
Pitfalls that already happened
The whole base had a single subscription in trialing, because the 14-day trial lived in the product database.
Find each product's trial source before measuring.
The cycle date was rewritten when a failed charge recovered, and activated accounts went back to looking unactivated.
Anchor on the first paid transaction and flag the old series as not comparable.
A rule flagged every subscription under 30 days old as a trial, including customers who paid on day one.
Use the source's trial dates, not the subscription's age.
Trial subscriptions canceled at the end of the period went in as logo churn and nearly doubled the rate.
Separate unconverted trials from paying-customer churn.
The final stage of the onboarding pipeline was set after the meeting, even when the customer was not using the product.
Activation comes from the product event, not the CRM stage.
People already in a sales conversation opened a new account with the same phone and showed up as high potential for another rep.
Link accounts by phone and tax ID and suppress those already with the team.
Acceptance test
The case is done when every item passes.
- Trials started by cohort match the count in the product's trial source, after deduplication.
- No unconverted trial shows up as churn in the MRR movements table.
- Every activated account has a milestone event dated on or after the anchor.
- Reprocessing does not change activation for cohorts with a closed window.
- Cohorts with an open window are flagged as partial, and rates with no eligible account are null.
Agent package
Paste into Claude, ChatGPT or any client with the Nekt MCP connected.
Trial and activation - Trial source per product: <product>: <table and fields>. Do not use billing's trialing status to count trials. - Activation milestone per product: <product>: <event>. Product without a milestone: activation is null. - Anchor: trial start (pre-sale) or first paid charge (post-sale). Never the billing cycle date. - Trial conversion is per customer: a paid charge on any subscription up to <N> days after the trial ends. - An unconverted trial is not churn and does not enter MRR movements. - Read conversion and activation by trial start cohort. A cohort with an open window is partial. - Accounts with the same phone or tax ID are the same person or company; the second is a reopening. - Activation series before <anchor fix date> are not comparable with later ones. Client exceptions: - <client>: on-time activation = milestone by the Saturday of W+5 from the sale week; target <x>%. - <client>: product <product> has no usage telemetry and stays out of the activation rate.
Build the "Trial and activation" case in my Nekt workspace following the rules at https://academy.nekt.com/en/industrias/saas/trial-e-ativacao.md. 1) Find the source tables and confirm the columns the rules use. 2) Create a transformation at the stated grain applying every rule. 3) Run the acceptance test and show me the result of each item. 4) Create the case's context document in the workspace, leaving my clients' exceptions open for me to fill in.
Variations
The cohort is the signup date and conversion is the first paid charge, with no trial end date.
On-time activation by sale week, with a target and a stretch target per person (for example, 60% and 80%).
Cohort by install month, 30-day activation over eligible installs and a null rate with no eligible installs.