More churn is created in the first thirty days than in the following three hundred. A customer who reaches a real result in week one tends to stay; one who is still “getting set up” in week four usually will not be there at renewal.
This checklist assumes a sales-assisted B2B product. Self-serve teams can run the same sequence with automation in place of the human touchpoints.
Before day one: the handover
Onboarding starts before the customer sees it. The handover from sales to whoever owns onboarding is where most programmes already go wrong.
- Write down the specific problem the customer bought to solve, in their words, not as a product category.
- Record what success looks like to them and by when — a number if possible, a described outcome if not.
- Name every commitment sales made, including roadmap items and integration promises.
- Identify the economic buyer, the day-to-day champion and at least one backup contact.
- Flag anything technical: data to migrate, systems to connect, security review to pass.
Hold a fifteen-minute live handover rather than passing a form. The things that matter are usually the things sales did not think to write down.
Week one: get to one real result
The goal for week one is narrow and non-negotiable: the customer does one real piece of work in the product and sees the output.
- Kickoff call within three business days. Any longer and momentum from the purchase is gone. Confirm the goal, agree the timeline, book the next two calls in the invite.
- Account configured before the call, not during it. Users created, branding applied, obvious settings done. Watching someone fill in settings is a poor use of the only hour of attention you will get.
- Import or connect real data. Sample data is for trials. Real data is what makes the product feel like theirs.
- Complete one end-to-end workflow together. Not a tour of features — the actual thing they bought the product to do, done once, by them, with you watching.
- Send a written recap. What was done, what is next, who owns each item, with dates.
If week one ends without a real result, the plan for week two is to produce one, not to move on to the next module.
Week two: make it a habit
The second week is about repetition without you in the room.
Check usage mid-week rather than waiting for the call. If the champion has not logged in since the kickoff, that is the conversation to have, and having it on day nine is far more useful than discovering it on day thirty.
Add the second user. A single-user account is fragile; a product used by three people survives one of them leaving. This is also where expansion begins, so it is worth pushing even when the customer says they will “roll it out later”.
Deliver one piece of training aimed at the daily user rather than the buyer. The person who evaluated the product and the person who will live in it are rarely the same, and the second one decides whether it sticks.
Weeks three and four: widen and prove
By week three the goal shifts from doing the thing to doing it routinely and showing that it worked.
- Bring in the remaining users or the second team.
- Connect the integrations that were scoped but deferred — these are what make the product hard to remove.
- Set up whatever reporting the economic buyer will want at renewal, now, while someone is paying attention.
- Run a 30-day review against the success definition written at handover. Show what has happened, name what has not, and agree the next milestone.
That 30-day review is the highest-leverage meeting in the whole process. It is where you find out whether the thing the customer bought is the thing they are getting, while there is still time to correct it.
The handover to steady state
Onboarding should end explicitly, not fade out. Mark it with a short document the customer keeps: what was configured, who the ongoing contact is, where the help resources are, and what the agreed next milestone is.
Pass the account to the ongoing owner with the original success definition attached, so the renewal conversation in ten months can be held against what was actually promised. This single piece of continuity does more for renewal rates than most retention programmes.
Signals that onboarding is failing
Four early warnings, all visible before day thirty.
The kickoff gets rescheduled twice. Not a diary problem. It means the project has lost its sponsor or its urgency, and the right response is to escalate to the economic buyer, not to offer a third slot.
Only one person has ever logged in. The account is a single point of failure, and that person leaving ends the contract.
The customer asks for training on everything. Usually a sign they have not identified their own primary use case. Narrow the scope rather than widening the training.
Nothing has been configured that is specific to them. Default settings everywhere at day twenty means the product has not been adopted, only installed.
These four form a reasonable basis for an early-life customer health score.
Scaling it without a CSM per account
Most of the sequence above can be automated for smaller accounts. The parts that must stay human are the handover, the first real result and the 30-day review. Everything else — reminders, training content, usage nudges, the recap email — works as well or better in product and in email.
A practical tiering: high-touch for the top 20% by contract value, a group onboarding webinar plus automation for the middle, pure in-product onboarding for the long tail. Review the thresholds every two quarters as the base grows.
What to measure
Onboarding that is not measured drifts back to whatever the busiest person has time for. Four numbers keep it honest.
- Time to first value — days from contract signature to the first real result. Track the median and the worst decile; the tail is where churn comes from.
- Onboarding completion rate — share of accounts that reach the defined end point within the target window.
- Seats activated at day 30 — as a share of seats purchased. A low figure here predicts both churn and a failed expansion conversation later.
- Retention at day 90 by onboarding cohort — the outcome measure. If changes to onboarding are not showing up here, they are not working.
Review these monthly alongside a read of two or three accounts that went badly. The qualitative read is what tells you which part of the process broke.
The parts customers remember
Teams tend to optimise the parts of onboarding they control and neglect the parts customers actually notice. Three small things punch above their weight.
Responsiveness in week one. A question answered in twenty minutes during the first fortnight sets an expectation that carries the whole relationship. The same question answered in a day, during week one, does the opposite.
Not being asked the same thing twice. Nothing erodes confidence faster than the onboarding specialist asking what the customer told the salesperson in the first call. This is entirely a handover problem and entirely fixable.
A written plan the customer can see. Even a shared document with five milestones and dates. It converts an open-ended project into something with an end, which is what makes people commit time to it.
Onboarding and the trial that preceded it
For products with a self-serve motion, onboarding does not begin at the contract — it began during the trial, and the two are frequently designed by different people with different goals. The activation milestone that matters in free trial conversion should be the same milestone the paid onboarding builds on, not a separate one invented after the sale. Where they differ, the customer effectively starts again on day one, which is a poor use of the goodwill a purchase creates.
Frequently asked questions
How long should SaaS onboarding take?
Thirty days is a reasonable default for mid-market software. Enterprise implementations with data migration and security review run 60–90. The number matters less than having a defined end point with a review attached.
Should onboarding be free or charged?
Charging for implementation increases customer commitment and funds the team, but it slows deals and can stall rollouts when budget is tight. A common compromise is free standard onboarding with paid services for migration and custom work.
Who should own onboarding?
A dedicated owner, whether that is a specialist team or a named CSM. Leaving it with the account executive almost always means it competes with closing next quarter’s deals.
What is the single most important step?
Getting the customer to one real result in week one. Every other step supports that, and no amount of later activity recovers from missing it.
