I built a four-week onboarding program in 10 weeks
Part one of three: how I built a four-week onboarding programme for hundreds of recruiters in ten weeks, using four hours of anyone else’s time
👋 This is a free edition of the MPL Build Newsletter—resources for running your people team like a product team. Become a paid member to get access to exclusive guides, templates, and case studies.
The most expensive part of building an onboarding training programme used to be the building. The instructional designers, the LMS implementation that often required its own IT team and implementation project, the agency quote with enough zeroes to make any sensible CFO request a quiet word, a six-month timeline that somehow always (always!) became nine. Production was the bottleneck back in the day. When I think of our onboarding programmes we built at McCann, we spent hours working with our delightful L&D team, sadly aware that we’d not be able to effectively test something with so much input and such high stakes. If you wanted something good, you paid for it in money or in months, usually both, and you crossed your fingers that the programme you’d scoped in January was still the programme you needed by the time it shipped in September.
As far as AI is concerned, I’m happy to say that I’m an optimistic and active user, but with a reasonable degree of skepticism about AI-slop, untrue claims, and unrealised potential. When I first started working with AI, I saw it’s potential for L&D but I’d never seen the resolution of that bottleneck I mentioned above fully realised. I’m pleased to say that bottleneck is gone.
The job quietly changed under our feet
For most of the history of our function, “build vs buy” was a genuine fork in the road. You either bought the thing (and inherited someone else’s idea of what good looks like) or you built the thing (and inherited a budget line, a headcount, and a timeline that outlived several reorgs). Building was the heavy, expensive, specialist option. It was the thing you escalated. I talk about this in Built for People quite a bit, the delicate balance between build and buy, and why it’s critical to have a team willing to do both. One of the things I’ve been most excited about the last 6 months is the speed in which we’re finding ourselves all capable of building truly world-class HR tooling on opinionated systems, with rigour, compliance, and near endless flexibility. Sana is one of those platforms.
The tools got good, then they got cheap, then they got fast, and the line between buying and building collapsed into a smudge. Buying a platform is no longer the end of a decision… it’s the start of a build. The moment you’ve bought the canvas, you’re now expected to paint.
Which means every People Leader and People Partner is now (or will soon become) a builder, whether or not that was anywhere in the job description they applied to. You can find this terrifying or you can find it the most fun the function has had in a decade. I’m firmly in the second camp, and I’d gently suggest you join me, because the people who thrive in this version of the job aren’t the ones with the biggest L&D teams around them, I’m afraid, they’re the ones with the biggest blast radius. They’re the ones who can hold a clear idea in their head and reach for the right lever.
The Talentful problem
I wanted to talk you through the work I did in early 2026 that really changed my mind on just how effective a People Team could now become with the right tools, building programmes from the ground up. Talentful onboards a lot of recruiters. Hundreds of them a year, for world-class clients who demand strong performance and a high bar. In a business like that, the recruiters aren’t a support function for the product… they basically are the product (absolutely human, highly appreciated, talented folks who make up the quality of our business). The value of the experience a client gets, the speed at which a new joiner becomes genuinely billable, the consistency of the work across a team that large, all of that lives or dies on how well, and how uniformly, you bring people up to speed.
So the commercial maths is brutally simple. A recruiter who is excellent in week three instead of week eight is not a soft win. It’s a huge performance margin. Multiply a five-week swing in time-to-productivity across hundreds of people and you are no longer talking about an HR programme; you’re talking about a line item the board cares about. Inconsistent onboarding, conversely, is a quiet tax you pay forever… in rework, in client wobbles, in the slow erosion of “the Talentful way” into “whatever your first manager happened to remember to tell you.”
The old way to solve this would have been to hire an L&D team, procure an enterprise LMS, and book out two quarters of build. I had none of those things, and frankly I wasn’t sure we needed them. We wanted something best in class, AI-first, and front-running in the market. What we wanted we had to build, not be lead by status-quo. What I went out to gather was one brilliant collaborator, one tool, and a very small, very precious budget of other people’s time.
Doing more with less, but not in the grim way
“Doing more with less” has been weaponised into a slogan that mostly means “do the same amount, while sad.” That’s not what I’m talking about, and I’m sure there is a whole blog nestled somewhere deep inside my frustration about this phrase in the world of AI, but that’s one for another time.
What I mean is that the leverage available to a single competent People person right now is genuinely absurd by the standards of five years ago. I built a four-week programme (an intensive online “remote classroom” first week, then three weeks of deep-dives and Friday workshops with submitted work) that covers hundreds of recruiters per year and runs on a monthly loop with a primary facilitator. Start to finish, it took ten weeks to build. It cost roughly four hours of input from the managers who actually owned the content. That is not me grinding harder. That is a tooling environment that has fundamentally changed the exchange rate between effort and output.
The constraint has moved. It used to sit on can we build it. It now sits almost entirely on do we actually know what good looks like. When production is cheap, taste becomes the scarce resource. Knowing what to put in the room, what to leave out, what an aspirational version looks like rather than a merely adequate one, that’s the expensive part now, so that’s where I spent my real money. (More on that in part two, where a former Uber onboarding lead proves her weight in gold.)
This is just People Ops as a product, wearing trainers
None of this is a departure from how I think the function should work; it’s the most natural expression of it as I shared in Built for People. If you treat the employee experience as a product, then onboarding is the trial-to-paid conversion moment, the bit where someone decides whether they bought the right thing. You wouldn’t outsource your entire product’s first-run experience to a vendor and never look at it again. You’d build it, measure it, and iterate.
The difference in 2026 is simply that the build is finally within reach of a normal-sized team with a normal-sized budget. The barrier was never imagination. People teams have never been short on ideas about how onboarding should feel. The barrier was that turning the idea into a living, scalable thing required resources we were never given. That barrier has quietly come down, and a lot of us are still standing on the other side of where it used to be, waiting for permission we no longer need.
Over the next two posts I’ll show you exactly how this one got built. Part two is the practical stuff: how we went from a learning brief to living courses in Sana without losing a month to the tool. Part three is the bit I’m proudest of: the automations and workflows that turn a complicated four-week programme into something one person can run every single month without it eating their life.
But the mindset has to come first, because the tooling won’t save you from a fuzzy idea. The question is no longer whether you can build it. It’s whether you know what you’re building, and whether you’re brave enough to make it aspirational rather than adequate.
So, what’s the thing you’ve been waiting for budget or headcount to build, that you could probably just build if you had a tool like Sana?