Welcome to the first issue of Force of Work. I'm Jay, Director of Learning at Docebo.
Over the years I've had hundreds, maybe thousands, of conversations with leaders across L&D, HR, people, and enablement. The stories I remember most are the ones that never make it into the spotlight. They come from people who do the hard work to find an answer, who don't give up when the challenges pile on, and who keep a workforce running while nobody is watching.
So I'm bringing them to you. Every two weeks, each issue goes inside one company you probably know and shows how they hire, train, retain, and now work alongside AI.
First up is GoDaddy, and a failure rate their L&D team plans for on purpose.
Cammy Wood's team has a target for how often they should be wrong. It's 70%.
Not tolerated at 70%, but expected at 70%. Her reasoning is that if you're truly experimenting, you shouldn't have a win more than three times out of ten. The goal is to learn fast so you can iterate fast.
Cammy is the Director of Learning Engagement at GoDaddy, twenty years into a career in sales leadership and L&D. And there's a lot to glean from the way her team operates: how they earned the room to fail, what their experiments look like, and where she draws the line.
So how does a team say that number out loud and still get trusted with the work?
A culture of experimentation
For about four years, GoDaddy has operated on what Cammy calls "an experimentation culture," and it's reshaped how her team works. They often start with pilots, run small tests, and then hold experimentation showcases to uncover what they learned.
This culture is what helps give her team the confidence to say a number like 70% out loud. There's also an element of psychological safety in it that gives her team the creative power to walk into a meeting and say, "I bet you this is what would happen if we did this."
But even still, nobody hands a 70% failure allowance to a function they don't inherently trust.
From order-takers to strategists
Cammy joined GoDaddy to support sales training, which she says was more of a concept than an actual program at the time. Credibility matters for buy-in, and her way of building it was to stop taking requests as written.
Stakeholders used to show up with the solution already picked: what it should look like, how it should sound, which modality to use. That's understandable. They know the business. But deciding the modality is a learning decision, and it's one only people in L&D are trained to make. The learning objective should define the format, not the loudest opinion in the room.
So the job shifted from taking orders to guiding stakeholders toward the right question.
"We've had to kind of flip that so that they understand: give us the what, we will explore with you the how, and we'll come to that conclusion together."
The question she likes to work back from is "From X to Y by Z." What is X? What is Y? And when do we want to achieve it by? Once those answers are clear, the objectives get clear, and the modality follows from them.
This approach also helps teams move from tying their success to completion and engagement rates to aligning their work with business-wide goals. (The kind of goals CFOs both understand and fund.)
What that looks like in practice
One welcome program, twelve months of small changes
On learning through experimentation, Cammy points to GoDaddy's new employee welcome program. They run six welcome sessions per year, and "in between each of those six sessions, we changed something. We tested something. We got feedback on something."
Her team iterated that way for at least 12 months. The version they landed on meets people where they're at (more on that below) and is earning great NPS scores. It combined running experiments with employee feedback, so the team could see what works for the organization and what benefits employees.
That version is now called follow the sun: three days, multiple sessions per day, all aligned with each employee's timezone. Americas gets one track. Rest of world gets identical content on Monday, Wednesday, Friday in their own daylight. Nobody is stuck in onboarding at nine at night after a full workday, or five in the morning before the sun is even up.
Experimenting on people is a different job
L&D can't just borrow the marketing or product team's playbook for experimentation.
"When you're experimenting with people behavior, that is different than experimenting with changing a button color or shifting a product name."
Cammy's team wants the freedom to be creative. They don't want to risk the security and trust of their employees. So for every experiment they run, proprietary information and employee data stay protected, with enough runway to play and explore inside that line.
Three things worth stealing
Ask for the what before the how. When working with stakeholders, ask them to outline their “X to Y by Z,” then propose the best path forward.
Set your guardrails first. Decide what's off limits so experiments can run within the appropriate confines.
Get comfortable with failure. Meaningful change rarely arrives on the first try.
Good learning is structured struggle. People get things wrong before they get them right, and our job is to make room for that. Experimenting with our own programs is a different activity, but it runs on the same logic: try it, see what happens, adjust. If learners get room to be wrong on the way to right, L&D programs should too.