Why capability-first failed: the Paris story behind Work Context

Let me take you back to Paris. A city steeped in history and revolutionary ideas.

I was there, just after a customer workshop. Sitting in a café beside the Eiffel Tower. I had been reviewing with our Chief Product Officer why the capabilities we extracted for customers weren't great. And how this impacted the matches we generated.

Truthfully, this realization came after we lost a customer. The matches and logic simply weren't working. We loved this customer. The outcome devastated us. We were determined to learn what went wrong.

As we dug deeper, we realized the job data we received was primarily job adverts. Each recruiter described the same job differently. Most content focused on employment branding. Not the actual work being done. We decided to go back to first principles. Without good people data or work data, AI is not a magician. This led us to focus deeply on the quality of job data. What we found was astounding. No one had any real work data.

No one understood the work being done beyond job titles and adverts.

Determined to understand this better, I called every business leader I knew. I asked how they described work. What I found was consistent. They talked about outcomes and tasks. How long they would take. The types of capabilities needed to achieve those outcomes. Had I gotten it all wrong? Everyone was talking about capabilities first. But no company knew the actual work being done. Workforce strategy tried to label work using solely capabilities. But that's not how the business talks. This explained the pushback on capability-based workforces.

Every enterprise is deploying AI. Almost none can see the work they're deploying it into.

Bold: why we moved away from the capability-first approach

My worry shifted. Why were we all running so fast towards a capability-first approach? This concern explained the underwhelming response to capability-first programs. To prove my point, we built out every customer's jobs with tasks, requirements, and the needed capabilities. That's when it finally clicked. Business leaders validated this was how they described work. On average, 80.5% of the tasks we recommended were accurate. Customers added localized context to the rest.

• Revolutionary insight. People have capabilities. But work has tasks. You need capabilities to do those tasks. This was a turning point.
• Creating a common language. To truly align people with work, we needed Work Context. A common language of work. It includes tasks, outcomes, and necessary capabilities.
• Industry-specific depth. We've now built Work Context on 25 industry-specific Work Ontologies. Each captures the depth and breadth of work and capabilities needed. We sunsetted our previous capability-only model. It wasn't enough.

From Job Architecture to Work Architecture.

Responsible: how this changed our approach

What became crystal clear: we could no longer rely on customer data alone. Now, with every customer, we bring in Work Context. We create this common language of work before going live. We even backdated this for current customers. This fills the gaps of work data in their environment.

• Improved accuracy. Building jobs with a focus on tasks and outcomes dramatically improved match accuracy. Not just capabilities.
• Proactive approach. Developing Work Context before implementation ensures a solid foundation. Customers build workforce strategy on real data.
• Customer-centric. This approach improved results. It strengthened relationships. They see us as true partners.

AI capability is compounding. Work visibility is not.

I truly believe capabilities are the currency that powers your marketplace. However, for this to work, capabilities must connect to the actual work being done. This is the only way to ensure alignment and impact.

I validated this thinking by diving into the work of Ravin Jesuthasan. He was onto this before anyone else. I highly recommend his book Reinventing Jobs. This is your playbook for navigating the new world of work.

Siobhan 💜