[drop_cap]T[/drop_cap]he card is charged. The account is provisioned. Somewhere in your product right now sits a customer who logged in once, poked at the one feature they signed up for, and hasn’t been back in eleven days. Sitting right beside them, in a different account, is a customer who has built three workflows into your product, taught two teammates how to use it, and would need a genuinely bad reason to ever cancel. Both paid you the same amount. Only one of them is still your customer next quarter.
This is Stage 8 of the GROW band: EDUCATE. Everything before this stage got the customer to say yes. This stage decides whether saying yes turns into staying. Get it right and the product becomes load-bearing in someone’s actual work — the thing they’d have to replace, not just cancel. Get it wrong and you’ve built a leaky bucket that ATTRACT has to keep refilling, forever, at whatever it costs you to acquire a customer in the first place.
Onboarding Gets You In the Door. Adoption Is the Whole House.
Onboarding is the first successful run: the account gets created, the first project gets set up, the first result comes back before the customer’s attention runs out. It matters — a trial or a new paid account that never reaches a first working result churns almost immediately, and time-to-first-value is the metric that measures exactly this. But onboarding is a single event. It happens once, and then it’s over.
Adoption is what happens after. It’s the second use case, then the third — the point where a customer stops using your product for the one thing they bought it for and starts routing other parts of their work through it too. A customer on one use case is a customer who is one budget review, one new hire, one “let’s consolidate our tools” conversation away from being cancelled. A customer on three or four is one whose workflow would break without you, and a broken workflow is what actually stops a cancellation, not a discount.
Here is the part almost every SaaS company gets backwards: most of the org’s energy goes into the first event and almost none goes into the second. There’s a whole onboarding flow, a welcome email sequence, maybe a product tour. Then the customer is left alone to discover the rest of the product by accident, if they discover it at all. Onboarding gets treated as a project with an end date. Adoption never gets treated as anything, because nobody owns the day after “welcome.”
The Adoption Ladder
Map the specific sequence of capabilities your product wants a customer to climb, in the order they’re actually able to climb it — each rung earned by the one before, not by how impressive it looks in a deck. A customer who hasn’t finished rung one has no business being shown rung three; that’s not adoption, that’s noise.
A working ladder for most B2B software looks something like this, adapted to what your product actually does:
- First value — the one thing they signed up for works, once, unassisted.
- Repeat use — they come back and do the same thing again without being prompted.
- A second use case — they route a different piece of their work through the product.
- Team use — a second seat gets added, or a second person on their team touches the account.
- Configuration — they’ve set up something specific to how they work: a saved view, an integration, a custom rule.
- Advocacy — they refer someone, leave a review, or ask you for a feature, all signs they’ve mentally filed you as “ours.”
Your own product’s ladder will have different rungs and a different order — the point isn’t to copy this one, it’s to write down yours, with a concrete, observable action defining each rung, so “how adopted is this account” has an answer that isn’t a guess.
Where an Account Sits, and How You’d Know
An adoption ladder is only useful if you can read it off real usage, not off a renewal date on the calendar. Every rung needs a signal that fires from product data — an event, a count, a threshold — not a survey you’d have to ask the customer to fill in.
| Adoption rung | The capability | The signal it happened | The nudge if it didn’t |
|---|---|---|---|
| First value | Completes the core action once | Event fires: first [X] completed | In-app checklist item; a day-2 email naming the exact next click |
| Repeat use | Returns and repeats the core action | Second session with the same action, within X days | A “you left this half-finished” reminder tied to their actual last state |
| Second use case | Uses a distinct feature area | First event from feature area #2 | In-product prompt at the moment their first use case naturally completes |
| Team use | A second user is added or active | Seat count > 1, or a second distinct user ID logs an event | Ask the champion directly: “who else on your team should have access?” |
| Configuration | Sets something up specific to their use | A saved view / integration / rule is created | A template gallery shown right where a blank state currently sits |
| Advocacy | Refers, reviews, or requests a feature | Referral link used, review submitted, feature request logged | Ask at a peak-usage moment, never at signup and never at a support ticket |
Read this table as a diagnostic, not a scoreboard. An account stuck at rung 2 for two months isn’t a curiosity — it’s the account you call before it churns, not after.
The Single-Champion Account Is a Ticking Clock
Here is a specific, felt version of the adoption fear worth naming plainly: an account where exactly one person at the customer knows how to use your product is not a healthy account, no matter how much they’re using it. That person is one resignation, one internal reorg, one parental leave away from being gone, and when they go, the knowledge of how your product fits into their company’s work goes with them. The renewal conversation six weeks later happens with someone who has never logged in and doesn’t know why they’re paying you.
Watch for this pattern directly: one active user on an account that could plausibly support more, especially at a company past some size threshold of your own choosing. When you see it, the response isn’t to sell harder — it’s to get a second person’s hands on the product before the first person’s departure becomes your problem. Ask your champion who else touches this part of the workflow. Offer to walk a second teammate through the setup. Make “who else should have access” a standard question at every check-in, not a crisis question asked only after someone already left.
Documentation Is a Growth Surface, Not a Cost Center
Most companies file documentation under support — a cost to be minimized, staffed reluctantly, judged by ticket deflection. That’s the wrong ledger. Good documentation is how a customer discovers rung 3 and rung 4 of the adoption ladder without you having to personally walk them there. It is doing sales and retention work, quietly, at zero marginal cost per customer.
The test for whether your docs are pulling this weight: can a customer who has only ever used rung 1 find, on their own, the page that shows them rung 3 exists and is worth trying? If your documentation is organized around features — an alphabetical reference of every button — the answer is usually no, because a customer at rung 1 doesn’t know what to search for. Organize it instead around outcomes: “how to do X,” where X is the thing a customer at the next rung is trying to accomplish. A reference section still earns its place for people who already know what they’re looking for; it just isn’t the front door.
In-Product Education vs. the Email Course
Both have a job. Neither replaces the other.
In-product education — tooltips, empty-state prompts, a checklist that tracks real progress — wins when the customer is already in the product and the friction is “I don’t know this is possible right now.” It’s contextual, it’s low-effort to consume, and it reaches people who never open your emails at all.
The email course — a scheduled sequence over days or weeks — wins when the lesson requires the customer to not be mid-task: a concept that needs explaining before they’d think to look for it, a workflow that spans multiple sessions, a case study that builds a case for trying something they haven’t yet considered. It also reaches people who’ve gone quiet in the product, which in-app prompts by definition cannot.
Use in-product nudges for “you’re here, and here’s what’s one step away.” Use an email sequence for “here’s why you’d want to be somewhere else in the product at all.” Sending an email about a feature the customer is currently looking at, or an in-app tooltip about a workflow they haven’t touched in a month, is the two getting swapped.
Release Notes Are the Cheapest Retention Asset Most Teams Waste
Every shipped feature already comes with a natural, no-cost touchpoint: the release note. Most teams write it for themselves — a changelog entry, terse, past tense, read by nobody but the engineering team that shipped it. That’s a wasted door.
A release note that earns its place does three things: names the customer problem it solves, not the internal mechanism (“find last month’s invoices in two clicks,” not “added created_at index to reporting query”); links straight to the feature, so reading the note and trying the feature are the same action; and goes out on a cadence customers can rely on, so the changelog becomes something they check rather than something they scroll past. Treat every release as an unpaid adoption nudge for whichever rung it happens to unlock, and the changelog quietly becomes one of your highest-leverage, lowest-cost retention channels.
AI Earns Its Place Here
This is exactly the kind of structured, repeatable drafting a model handles well, provided you feed it your actual product and actual data instead of asking it to invent either. Give it your adoption ladder — the real rungs, the real signals — along with a specific account’s current usage pattern, and ask it to draft the next nudge: the in-app prompt, the release note, or the “who else should have access” message, tuned to exactly where that account sits.
Treat every draft as a first pass, not a send-ready message. The model can hold the shape of a good nudge — problem-first, one clear next action, tied to something the customer actually did — but only you know whether the feature it’s pointing to is really live for that account, whether the timing is right, and whether the tone matches how you actually talk to customers. Verify before it goes out; a nudge pointing at a feature the customer can’t access does more damage to trust than sending nothing at all.
What’s Next
An adopted customer — one who’s climbed past the first rung, brought a teammate in, configured the product around their own work — isn’t just retained. They’re primed to become your best acquisition channel. Stage 9, SHARE, is where that depth of use gets turned into the referral, the review, and the case study that do for the next stranger what your original hook did for this one.