Every signup, we watch. Not in a creepy way — anonymized event logs, no content, no PII. But we watch what people do in their first hour on the platform, because that hour decides almost everything about whether they're still here in ninety days.
We've now watched the first hour of about 18,000 teams. The patterns are unreasonably stable. There are three onboarding hours that lead to retention, and four that don't. Knowing which one a team is in by minute fifteen lets us intervene before the hour is over.
Most teams sign up with a problem in mind. The ones who solve that problem inside the first hour stay. The ones who get distracted exploring don't.
The three good onboarding hours.
1 · The "one job" hour. A team signs up to do exactly one thing — usually because a single tool just raised its price or broke. They import contacts. They send one email or close one deal. They log out. They come back the next morning and do it again. By day three they have a habit on one module and zero opinions on the rest. Retention curve: nearly flat at 90 days.
2 · The "audit-driven" hour. An ops lead has done the line-item walk on their software bill and arrived with a list. They check four boxes on the module page during signup. They don't try to use any of it on day one. They schedule a thirty-minute call with us, do the imports together, and start with one module the next week. Retention curve: best of the seven cohorts.
3 · The "agency tier" hour. The team signs up to white-label. They go straight to Settings → Branding, upload a logo and pick a color. They spend most of the hour clicking around a fake client account they spun up. They don't touch real customer data for two weeks. They are evaluating, not using. Retention curve: lumpy in the first 60 days, then very durable.
The four bad onboarding hours.
1 · The grand tour. A founder, usually solo, signs up and clicks every module in the sidebar in sequence. CRM, then HR, then accounting, then projects, then POS, then LMS, then bookings. Forty-five minutes of clicking, zero data imported. They feel impressed and overwhelmed in equal measure. They log out. Most don't come back for a week. About 60% never come back at all.
2 · The wrong-module-first. A team's actual problem is their CRM. They sign up and start with the website builder or the LMS, because those felt more fun. By minute forty they've built nothing real. They close the tab. They will not return.
3 · The integration-rabbit-hole. A team spends the entire hour in Settings → Integrations trying to connect Stripe, then QuickBooks, then Mailchimp. They never make it into any actual module. The integrations feel like the work. They are not the work.
4 · The "evaluate against the spec" hour. Almost always a larger team — 30+ people — that has been told to "evaluate" the platform against a 47-item checklist. They open one module, find a feature missing or a label they dislike, and move to the next. Within an hour they have built up a list of grievances and built nothing real. We rarely save these. The decision was made before they signed up.
What we changed about onboarding.
Once we could see the patterns, we changed the product in three small ways:
- The "one job" question, up front. The very first screen after signup, before the dashboard loads, asks: "What's the first thing you'd like to do this week?" Twelve options, plus a free-text box. We use the answer to demote every other module's tile until the chosen one is set up.
- An import-first dashboard. If you said "manage my customer list," the dashboard for the next 14 days leads with the import button, not with a chart of empty data. We removed three slick visualization tiles that surveyed well but predicted churn.
- A 30-minute "audit call" offer for any team over 5 seats. Free, no sales pitch, just an ops engineer running the line-item walk on their existing stack with them and giving them a written switch plan. About 70% take it. About 90% of those become annual-tier customers.
Why this matters for bundles.
The first-hour problem is harder for all-in-ones than for point solutions. A point solution's job is obvious; you signed up for the one thing. A bundle has 150 modules sitting in the sidebar, each of them a possible distraction from the actual reason you came.
The lazy answer is "let the customer explore." The honest answer is that most customers don't want to explore — they want to solve the problem they came in with and feel slightly more competent than they did an hour ago. The job of onboarding is to make sure they leave that hour having done one specific real thing, even if it's small.
The new metric
We replaced our old onboarding KPI — "time to second visit" — with a much simpler one: "did the customer complete one durable action in the first hour?" Sending an email that lands in a real inbox. Importing a contact list that has more than ten people in it. Logging an actual deal with a real dollar value. Issuing a real invoice. The list is short and concrete.
Teams that hit that one-action mark in the first hour retain at four times the rate of teams who don't. Nothing else we have ever measured comes close to that delta. It is, as far as we can tell, the entire job of the onboarding flow.