I wasted $40M on capability strategies so you don't have to

At Reejig, we're obsessed with workforce redesign. But before we nailed the formula, we made plenty of mistakes. $40 million worth, to be exact. My journey was an expensive education in what doesn't work when building capability strategies. Here's the reality check you need to avoid the same pitfalls.

The costly mistakes organizations are making

Despite the best intentions, many organizations fall into these traps.

• Poor data quality. Extracting or inferring capabilities from job descriptions and adverts led to irrelevant and misleading insights. These sources don't reflect the actual work being done.
• Inconsistent and messy data. Capabilities extracted from multiple systems created a chaotic, ever-changing dataset. AI dynamically updating records without governance made it worse. Pushing them into systems like Workday or SuccessFactors compounded the problem.
• Flawed data sources. Job adverts are one of the worst data sources for inferring capabilities. The same job appears in multiple locations. This skews demand signals. It creates false insights about "emerging" requirements.
• Unharmonized data. Different systems and teams use their own language for capabilities. This results in fragmented datasets. No alignment across the business. Confusion and inefficiencies as each system operates in silos.
• Large language models (LLMs) issues. LLMs provided inconsistent and unreliable outputs. They offered different recommendations for the same job. It depended on how it was described.
• Job architectures. Using job architecture to match work was ineffective. They were rigid. Out of date by the time projects completed. They didn't align with how businesses actually describe or think about work.
• No governance. Without governance, duplications, inaccuracies, and inconsistencies compromised data quality. Maintaining a reliable dataset became impossible.
• AI automates tasks, not capabilities. AI and automation handle specific tasks. Like analyzing data or generating reports. They don't "have capabilities" the way humans do. Recognizing that AI is task-driven makes integration more effective.

Trust me. I tried all of this, and none of it worked. If you want a better outcome, you need a smarter approach.

If I automate chaos, I just scale the chaos.

What works and lessons learned

1. Focus on the work being done. Most organizations don't have a clear view of actual work. That's why we built Work Context, built on 25 industry-specific Work Ontologies. It fills the gaps. It uncovers real insights.

2. People have capabilities. Jobs have tasks. Tasks need capabilities. Once you see work this way, you can't go back.

3. Create one common language of work. Workforce strategy and business leadership need a unified structure. Start with the work being done. Then layer in capabilities. This reduces friction. It builds buy-in for change.

4. Harmonize your data. Pick one system to standardize capabilities across all use cases. Hiring, learning, work movement, and more. This ensures consistency and better alignment.

5. Put a governance layer in place. Don't allow your system to constantly update jobs in your systems of record. Agree on a governance timeline. Approve updates on a schedule.

6. Build strong data foundations. A reliable dataset on your people's capabilities and the work being done is critical. This lets you solve real problems. It supports meaningful strategy work.

7. Match work to workers based on real insights. Go beyond job titles. Understand the tasks being done. Align people with what they're best at.

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

Why this matters

The old methods don't just waste money. They prevent you from realizing your team's potential. Here's how our approach keeps you competitive:

• Uncover AI opportunities. Build clear roadmaps.
• Eliminate inefficiencies with precise task mapping.
• Design meaningful career pathways. Engage your workforce.
• Create a marketplace of opportunity, fueled by accurate data.

From Job Architecture to Work Architecture.

Take action

We've learned the hard way so you don't have to. If you're serious about future-proofing your workforce strategy, let us show you how to get it right.

AI capability is compounding. Work visibility is not.

Change starts with the right data. Together, we build the way the world works.

Siobhan 💜