Welcome to this edition of Business Case, the format in which we take a realistic company, put numbers on a process problem, and calculate what doing nothing costs.
In software businesses, onboarding is the process where revenue, retention and reputation are decided at the same time. A customer who reaches productive use in four weeks tends to renew, expand and refer. One who spends three months in kickoff calls, data migrations and open questions renews reluctantly, if at all — and consumes support capacity throughout. Yet onboarding is rarely treated as a measured process. It sits between sales and customer success, runs partly in the CRM, partly in a project tool and partly in email, and is usually managed by individual effort rather than by design.
That combination — high commercial impact, low process visibility — is exactly where a business case becomes clear. This edition looks at a fictional but realistic software company, quantifies where its customer onboarding fails, and sizes both the cost of inaction and the return of three concrete levers.
Verdalis Software GmbH, based in Karlsruhe, sells field-service management software to mid-market customers across the DACH region. The company has 240 employees, annual revenue of 34 million euros, and roughly 1,400 active customers, which puts the average recurring revenue per customer at about 24,000 euros a year. Customer success and onboarding are handled by a team of 14 people at a fully loaded cost of 85,000 euros per person, or 1.19 million euros a year. Verdalis signs about 180 new customers annually.
The numbers that matter look like this. From signed contract to go-live, onboarding takes 71 days on average, against a peer benchmark of around 35 days for comparable mid-market implementations. Time to first productive use is 46 days. Fifty-eight percent of onboardings miss the go-live date agreed at kickoff — that is 104 of the 180 projects each year. In the first 90 days, a new customer raises 24 support tickets on average, roughly double the benchmark of 12. And 14 percent of customers churn within their first twelve months, against a benchmark of 8 percent.
Nothing in that picture suggests a weak team. Verdalis has good people, an established product and satisfied long-term customers. What it does not have is a described, measured onboarding process — and every number above is a consequence of that.
Problem one: onboarding starts before the prerequisites exist. Kickoff is scheduled as soon as the contract is signed, because momentum feels valuable and the customer is enthusiastic. But the data export is not ready, the technical contact has not been named, the interface to the customer's ERP has not been scoped. So the first three weeks are spent waiting and asking, then re-asking. Each delayed onboarding consumes about 3.5 additional customer-success days in coordination and rework. At 425 euros per day (85,000 euros over 200 working days), the 104 delayed projects cost roughly 155,000 euros a year in internal effort alone.
Problem two: revenue recognition waits for go-live. Verdalis, like most vendors in its segment, starts billing when the customer goes live. Every day of onboarding delay is therefore a day of deferred revenue. Being conservative and counting only the 104 delayed onboardings, and only 30 of their excess days, the arithmetic is 104 projects × 30 days × 65.75 euros per day (24,000 euros of annual revenue divided by 365) — about 205,000 euros of revenue shifted out of the year. Not lost forever, but reliably financing the following year instead of this one.
Problem three: a rough start produces support load and churn. Customers who go live from an incomplete configuration ask twice as many questions in their first quarter. Twelve additional tickets per customer across 180 customers is 2,160 tickets; at 22 euros of handling cost each (half an hour at 44 euros), that is roughly 48,000 euros. The heavier cost is retention. Six percentage points of excess first-year churn on 180 new customers is 10.8 customers, or 259,000 euros of recurring revenue; valued conservatively at a 75 percent gross margin, that is about 194,000 euros. Together, problem three costs around 242,000 euros a year.
The total cost of doing nothing: roughly 600,000 euros a year — 155,000 in internal effort, 205,000 in deferred revenue and 242,000 in support load and churn. That is about 1.8 percent of annual revenue, produced not by a single failure but by a process that nobody owns end to end.
Lever one: a readiness gate before kickoff. Define the small set of conditions that must be true before onboarding starts — data export in the agreed format, named technical contact with allocated time, interface scope confirmed, decision-maker for configuration questions identified — and do not start the clock until they are met. This feels like a delay and is the opposite: it moves waiting from the middle of the project, where it is expensive and invisible, to the front, where it is cheap and negotiable. The condition that makes it work is ownership at the handoff: sales completes the checklist as part of closing, not as a favour afterwards.
Lever two: measure the onboarding process instead of the projects. Verdalis knows how each onboarding went; it does not know how onboarding behaves. The data already exists — the CRM records the signature, the project tool records phase changes, the ticket system records questions, the deployment log records go-live. Joining those timestamps reconstructs the real path and answers the questions that matter: in which state do projects sit longest, how often does a project bounce back to a previous phase, and which variants actually occur. This is standard process mining work, and AI-supported platforms such as noreja add the causal layer — not only where the delay appears but what drives it. The typical finding in this kind of process is that a majority of volume runs through a handful of variants, which makes a standard path for the common case realistic. We described the same reasoning applied to employee ramp-up in the edition on how to halve time to productivity.
Lever three: thresholds and early escalation instead of after-the-fact reporting. Once the process is measured, each onboarding can be watched against a small number of live signals: days in the current phase, number of phase reversals, ticket volume before go-live. Each signal gets a threshold and a named owner, so a project that is drifting gets attention two weeks before the go-live date rather than a post-mortem after it. This also handles the genuine exceptions properly — a complex migration is allowed to take longer, deliberately, instead of hiding inside the average. The related discipline of keeping exceptions from breaking the standard path is covered in the edition on handling exceptions without breaking flow.
The projected impact, deliberately conservative. Assume onboarding drops from 71 to 48 days — an improvement, but still well short of the 35-day benchmark — and that missed go-live dates fall from 58 to 30 percent, so delayed projects go from 104 to 54 per year. Internal rework then falls by about 55 percent, saving roughly 85,000 euros. The 50 fewer delayed projects, at 30 days each, bring forward about 99,000 euros of revenue. A 30 percent reduction in early tickets saves around 14,000 euros, and cutting first-year churn from 14 to 11 percent retains 5.4 customers, worth about 97,000 euros at the same margin. Total annual benefit: roughly 295,000 euros, about half of the identified cost of inaction.
Against that, the investment: approximately 120,000 euros one-off for data integration, process analysis and the redesign of the handoff, plus 45,000 euros a year in platform and maintenance cost. Net annual benefit is therefore about 250,000 euros, and the one-off investment pays back in just under six months. Even if only the internal-effort and ticket savings materialised and every revenue effect were ignored, the payback would still arrive inside two years.
How long does your customer onboarding take from signature to productive use — as a measured median, not an impression?
What percentage of your onboardings miss the go-live date agreed at kickoff, and who sees that number every month?
If billing starts at go-live, how much revenue does your current onboarding duration push into the next financial year?
Which conditions would have to be met before kickoff for the first three weeks to stop being waiting time?
Do your early support tickets get analysed as a signal about onboarding quality, or only resolved as individual cases?
Verdalis does not have a customer success problem; it has an unmeasured process that quietly costs about 600,000 euros a year. That number is not the result of anyone working badly — it is the predictable output of a process that starts before it is ready, is not measured end to end, and is corrected only after the promised date has passed. A readiness gate, a measured process and a small set of thresholds with named owners are enough to recover roughly half of it within a year, with payback in under six months. The first step is not a project: it is measuring what your onboarding actually does, from signature to first productive use.
We invite you to run the same calculation for your own onboarding — take your new customers per year, your average onboarding duration against a realistic benchmark, and your first-year churn, and see what the gap is worth. If you would like a second pair of eyes on the numbers or on the process behind them, we are happy to look at it with you.
Three things at once: internal effort for coordination and rework, revenue deferred when billing starts at go-live, and higher early support load with correspondingly higher first-year churn. In the example above — 180 new customers a year, 24,000 euros average recurring revenue, 71-day onboarding against a 35-day benchmark — those add up to roughly 600,000 euros annually, about 1.8 percent of revenue.
Because it moves waiting to the point where it is cheap. Starting without the data export, a named technical contact or a scoped interface does not remove those dependencies; it hides them inside the project, where each gap costs coordination effort and calendar time. Agreeing the conditions before the clock starts turns invisible mid-project delay into a short, explicit preparation step.
The relevant timestamps already exist across the CRM, project tool, ticket system and deployment logs. Joining them reconstructs the actual onboarding path, showing which phase holds projects longest, how often projects fall back to an earlier phase, and which variants really occur. That turns a set of individual project stories into a measured process that can be improved deliberately.
A short list is enough: median time from signature to go-live and to first productive use, share of onboardings missing the agreed go-live date, days spent in each phase, phase-reversal rate, tickets in the first 90 days, and first-year churn. Each needs a threshold and an owner — a metric without either is reporting rather than control.
In the modelled case, a one-off investment of about 120,000 euros with 45,000 euros of annual running cost produces roughly 295,000 euros of annual benefit, giving payback in just under six months. Even ignoring every revenue and churn effect and counting only internal effort and ticket savings, payback still arrives within two years.