đ This is a free edition of the MPL Build Newsletterâresources for running your people team like a product team. Become a paid member to get access to exclusive guides, templates, and case studies.
Like hosting a party or running an event, most well-oiled teams can build an onboarding programme once.
Just throw a heroic week at it, rally the managers, run a brilliant cohort, bask in the glow of the feedback survey, and then realise whatâs against you because the second cohort needs the same heroic week and you do not have another one in you right now. You have OKRs, you have a team with a whole new learning programme to design. âWeâre seriously going to need more headcount to stop this being a one-off is a performanceâ, you think. What you actually want is a machine â something that produces consistently excellent onboarding every month while you are in meetings, on holiday, or doing literally any other part of your job.
This is the difference between building a thing and building the engine that builds the thing, and itâs the part of People Ops that I think we underinvest in most. Onboarding for hundreds of recruiters cannot be an event. It has to be infrastructure.
Onboarding is a subscription, so build it like one
If you take the product framing seriously, onboarding isnât a launch â itâs the first experience product, and you have to repeat it time and time again. Itâs the moment a new joiner decides whether the thing they bought is the thing they were sold. And subscriptions donât run on heroics; they run on systems that deliver the same quality to the thousandth customer as they did to the first.
At Talentfulâs scale, thatâs not a nice-to-have. New people arrived constantly as our client book grew. If onboarding quality was a function of how energetic the People team happened to be feeling that month, weâd end up with a programme whose worst version is the one a real human will inevitably experience. The whole point of the engine is to make the floor high and the variance low.
So once weâd built the individual Sana Live Courses (part two), the job became wiring them into something that runs itself.
The four moves that turned courses into an engine
One: roll everything into a single enrollable programme. Rather than a scattered set of courses a new joiner has to be chaperoned through, all the sessions were rolled up into one training programme in Sana with direct enrolment through our HRIS integration. Someone joins, they get enrolled, and the entire four-week journey unspools for them in order. No one is manually assembling a bespoke learning path per person, which is exactly the kind of invisible admin that silently eats a People team alive.
Two: build a pass/fail rubric into the programme. This is the bit that separates a training programme from a content library you hope people read. We built a submission path and pass/fail rubric directly into Sana, so âdid they actually get itâ stops being a vibe and becomes a measured output, trackable in reporting on an individual and cohort basis, testing them for client readiness. I bang this drum constantly and Iâll bang it again here: People teams should be setting output metrics on their own work. A rubric means you can see, per cohort and per person, whether the programme is doing its job â and when it isnât, you know which session to go and fix rather than vaguely sensing that onboarding âcould be better.â Thatâs the difference between iterating on evidence and iterating on feelings.
Three: build the calendar once, then duplicate it. We built a calendar of events for the live sessions across the four weeks â and then made it a template we duplicate, once a month, for each new cohort. I will happily argue that the most underrated tool in all of People Operations is the duplicate button. Build the month beautifully one time; copy it forever. The marginal cost of cohort number fourteen is a few minutes of cloning, not a fresh build.
Four: one facilitator to MC the whole thing. The live programming each month is run by a single facilitator who hosts the sessions and keeps the cohort moving â part air traffic control, part host. Not a team. One person, running a programme that reaches hundreds of recruiters over time, because the structure does the heavy lifting and the human does the part that genuinely needs a human in the room.
The receipts
Let me put the whole thing on the table, because the numbers are the argument.
Ten weeks to build, end to end.
Four hours of input from the managers who owned the content.
A client readiness rubric giving us a real read on whether it works.
One facilitator running it each month.
Coverage for hundreds of recruiters, on a loop, indefinitely.
Now hold that next to the old way: an L&D hire or three, an enterprise LMS procurement, two quarters of build, and a programme that still would have needed a small standing army to run each month. The gap between those two pictures isnât an efficiency saving. Itâs a different category of what a single People person is now capable of shipping.
And â this is the part I want to be honest about â none of it was the tool being magic (although a lot of features feel this way!). Sana is genuinely good at this, Iâd and Iâd happily (fervently) recommend it, but the leverage came from the sequence. Buy the thinking (Luca and the interviews). Get the architecture right (learning objectives, not vibes). Brief the managers so they react rather than create. Pour it into the platform as a rough first draft and iterate live. Then wire it into an engine with enrolment, a rubric, a duplicable calendar, and one human at the helm. The platform was the easy 20%. The 80% was knowing what to build and being disciplined about other peopleâs time.
What this actually buys you
The thing I keep coming back to is what the engine frees you to do. Once onboarding runs itself, the People teamâs attention is liberated for the work that genuinely needs a beating heart in the room â the coaching, the judgement calls, the human operations that no rubric will ever capture. The systems arenât there to replace the human bit of the job. Theyâre there to protect it, by clearing away the work that didnât need you in the first place.
Thatâs the whole promise of building like a product team. Not more dashboards for their own sake, not automation as an end in itself, but a function that does the systematic things systematically so it can spend its scarce human energy on the things that are irreducibly human.
Ten weeks, four hours of borrowed time, and an onboarding programme that now greets every new recruiter with the same aspirational, well-built first month â whether theyâre cohort two or cohort twenty. Thatâs whatâs possible now, for any of us willing to build it. đ§Ą