Customer Onboarding

For a long time nobody decided what happened after the signature. Sales opened an account, sent a link to the help center, and that was the end of it. Customers who happened to figure things out settled in on their own. Customers who got stuck stayed stuck, and nobody on the vendor side knew where.

When software was sold once and installed forever, that was survivable. The money had already changed hands. Subscriptions changed the arithmetic. Most of the revenue arrives at renewal — the moment the contract term ends and the customer decides whether to continue — and a customer who never used the thing does not renew.

Pagedeck learned this the plain way. Sofia closed an annual contract with a forty-person company, passed over the handoff notes, and two months later exactly six of the forty seats had ever logged in. The admin on the customer side explained that their old documents had lost their formatting during the import, so the team had quietly gone back to the tool they already had. Nothing was wrong with the product. The account never got through its first two weeks.

Customer onboarding is the decision to stop leaving those two weeks to chance. The period right after signature gets treated as its own stage, with an owner, a deadline, and a written condition for being finished.

It has a start date and an end date, which is unusual for a lifecycle stage. It opens on the contract date and closes when the agreed completion condition is met, typically somewhere between two and eight weeks later.

The second thing to notice is that most of the actual work sits on the customer's side of the table. Migrating documents, creating accounts, assigning permissions, announcing the change internally — the admin at the customer has to do all of it. Progress therefore depends less on how hard the vendor pushes than on how many hours that one person can spare.

Teams usually run it in four steps.

1. Kickoff — confirm what the customer bought the product to fix and write the completion condition down. The gap between what sales heard and what the department that will actually use the product expects shows up here more often than anyone likes.
2. Setup and migration — accounts, permissions, and moving existing material across. Six accounts out of ten stall at this step.
3. Training — walk through the product along the customer's real workflow. Taking one thing the team does every week and finishing it inside the product sticks better than touring the feature list in order.
4. First result — engineer a moment where the customer sees the benefit with their own eyes.

The third characteristic is that failure here is quiet. An account that is fully configured, fully trained, and completely unused generates no complaints. It looks fine right up until someone opens the usage report two months before renewal.

Usage has to survive the vendor letting go. Configuration finished and three training sessions delivered are not the condition. The question is whether the same people come back at the same frequency in a week when nobody from the vendor checked on them.

Two signals are usually read together.

1. Passing the first value milestone — the first moment the customer sees the benefit for themselves. At Pagedeck that is the first meeting where a team writes the notes inside the product and shares them from there.
2. Repeat usage. Pagedeck treats the account as ready when at least 70 percent of invited seats edit a document three or more times a week for two consecutive weeks. On a forty-seat account that is 28 people.

The numbers differ by product, but the way they get chosen doesn't. Look at the customers who stayed past a year, find the usage pattern they showed early on, and draw the line there. A team that already calculates customer health can use the point where that score crosses a threshold instead.

Meeting the condition moves the account into adoption, the stage where the product settles into the team's daily work, with renewal and expansion behind it. If the calendar ran out and the condition didn't land, don't mark the stage complete — write down what blocked it. A completion mark reads to the next owner as "this one is fine," so a mark handed out without the condition costs the same customer twice.

"We're in week four with this account and the migration still isn't done. Their admin keeps getting pulled away."

The deadline is at risk, so either we do the migration for them or we cut the scope.

"Before you mark that complete, check whether they hit the first value milestone."

Every training session on the schedule can be done and it still isn't finished if the customer hasn't seen a result.

"Average time to complete was 31 days this quarter. The target was 21."

The stage is being managed as a thing with a number on it, and somebody wants to know where the extra ten days go.

At Pagedeck, once a contract closes, Sofia puts the customer's admin and the assigned owner in the same kickoff meeting. The completion condition gets written there and confirmed with the customer in the room. Both sides leave holding the same sentence about what has to be true four weeks later.

Mia takes what accumulates here back to the product team. When three of five accounts stall on migration, that is not a sign the owners should try harder. It is a sign the import feature is too weak.

The most common mistake is treating the end of training as the end of onboarding. Training is scheduled, it has a finish time, and that makes it convenient to use as the completion test. But knowing where the buttons are and moving the work onto the product are different things. Plenty of teams sit through three sessions and keep using their old tool. Mark accounts complete on training attendance and the complete list fills up with customers who moved nothing, and the team trusting that list finds out a month before renewal that usage is near zero.

Treating speed as the goal is the other one. Days to complete is easy to measure, so a short number looks like good work. The cheapest way to shorten it, though, is to lower the bar. Trim the migration scope and collapse training into one session and the average drops from 31 days to 18, while the churn rate six months out stays flat or gets worse. The thing worth reducing is not the calendar but the effort the customer spends getting to a first result.