How Verdant Labs runs its new-hire cohorts as one coached Bitir group

Gallup has reported that only 12% of employees strongly agree their organisation does a good job of onboarding new people — and small, fast-growing companies are often the worst offenders, because nobody owns the process. Martha Okonkwo, COO of Verdant Labs, a 15-person climate-tech startup in London scaling towards 30, had exactly that gap. Onboarding lived across Notion, Slack, Linear, and her own head. One private Bitir cohort per intake took day-30 onboarding satisfaction from 6.2 to 8.8 out of 10.

Where she started

Verdant Labs builds carbon-accounting software for mid-sized manufacturers. Martha joined as employee number nine and became COO when the founders realised someone had to own everything that wasn't product or sales. Hiring was the part that scaled fastest and broke first.

The company hired in waves — three engineers and a designer one month, two on the commercial side the next — so onboarding was never a steady trickle she could handle person by person. It came in cohorts, and each cohort hit the same wall.

The setup, if you could call it that, was four tools:

• A Notion workspace with an onboarding page per role, half of it out of date.
• A flurry of Slack invites to a dozen channels, with the actual "start here" message lost by lunchtime on day one.
• A Linear board where someone occasionally remembered to add onboarding tasks.
• Martha's own memory, which was the only thing that actually knew whether a new hire had a laptop, a buddy, and a first task.

None of it told a new starter what to do next, and none of it told Martha who was quietly stuck. A new engineer could spend four days waiting for access to a repository before anyone noticed.

The breaking point

A cohort of four in the autumn made the problem impossible to ignore. Two of them ramped fine. The other two — a capable back-end engineer and a first commercial hire — drifted through their first fortnight without a clear sense of what "good" looked like by week two, and both said as much in their day-30 survey.

One of the two left inside six months. In the exit conversation she was candid: the work was interesting, but the first month had felt like being handed a pile of links and left to guess, and she never quite recovered the momentum.

"Losing someone good because our onboarding was a mess, not because the job was wrong, was the thing that made me stop and rebuild it," Martha says. "We were a company that measured everything about our product and nothing about the first month of someone's job here. I could tell you our activation funnel to two decimal places and I couldn't tell you whether last month's new hire felt set up to succeed."

What she set up in Bitir

Martha rebuilt onboarding around a single idea: treat each intake as a cohort that goes through a structured programme together, not a set of individuals each lost in their own tab. The reasoning follows the distinction between a set of people and an actual team that we set out in team coaching vs team training — the cohort is the unit, and it needs a rhythm, not just a document.

1. One private group per intake. Every new starter in a given month joins one closed Bitir cohort, so they onboard alongside the others hired at the same time rather than in isolation. The buddy and Martha are in the group too.
2. A pinned first-week plan. The "start here" post that used to vanish in Slack is pinned at the top of the group — day-one setup, week-one goals, who to ask for what — so it is the first thing every hire sees and it stays put.
3. A personal 30/60/90 goal per hire. Each new starter writes their own first-90-days outcome as a goal visible only to them, their buddy, and Martha, following the onboarding sequence in our guide to onboarding members into a group.
4. A light daily check-in for the first fortnight. A 30-second recurring check-in — "what did you get done, what are you blocked on" — that surfaces a stuck starter within a day instead of a week. It switches to weekly after the first two weeks.
5. Weekly ramp assignments. Each week of the plan lands as a structured assignment with a due date, so ramping is a followed sequence, not a Notion page nobody revisits.

The daily check-in was the piece Martha was least sure about and now rates most highly. It gave the quieter half of every cohort a low-stakes way to say "I'm blocked" without having to interrupt someone senior in a public channel.

What changed in the numbers

Martha compared the four intakes before the switch with the four after, each a cohort of three to five.

The satisfaction figure is a simple day-30 question the company already asked and mostly ignored; the difference is that Martha now acts on it before the score is set, not after. The time-to-first-contribution number is the one the founders care about, because a nine-day ramp instead of eighteen is, across a year of hiring, weeks of productive work recovered.

What she credits, in order:

1. The daily check-in turned "quietly stuck" into a visible signal, so a blocked hire became a two-minute fix on the same day rather than a lost week.
2. The pinned plan meant no new starter ever had to ask "so what am I actually meant to be doing this week" — the answer was already at the top of their group.
3. Onboarding in a cohort rather than alone gave new hires a peer group from day one, which the exit-risk hires had never had.

The retention figure is the one Martha is most careful about — the sample is small and a startup's retention moves for many reasons — but the direction matched what she saw in the surveys, and no post-switch hire has cited onboarding as a reason for leaving.

What she would tell another operator

Martha's advice to other founders and operators at small companies is blunt.

• Onboarding is a process, not a document. A perfect Notion page that nobody is walked through is a filing cabinet, not an onboarding.
• Run intakes as cohorts wherever you can. A new hire with three peers going through the same first month is far more resilient than one onboarding alone.
• Make "I'm blocked" cheap to say. Most onboarding failures are just a stuck starter no one noticed, and a daily check-in is the cheapest early-warning system you can build.
• Measure the first 30 days like you measure your product. If you can recite your funnel but not your last hire's day-30 score, you are optimising the wrong thing.

She now treats the weekly cohort check-in as a light coaching rhythm rather than pure admin — the same shift from event to rhythm we argue for in the team coaching guide.

What's next

Martha is extending the same cohort structure past the first 90 days into a rolling "new joiners of this quarter" group, so people hired a month apart still have a peer band as the company grows. She is also handing the day-to-day of running each cohort to team leads, with the per-hire view making it straightforward to pass a cohort across without losing sight of who is on track.

Related reading: team coaching vs team training: what the research says, and how to onboard new members to a private coaching group.