The real cost of knowledge fragmentation: $14,600 per employee per year

McKinsey's research on knowledge worker productivity found that employees spend 19% of their time searching for and gathering information. That number has been widely cited. What's less discussed is what it actually costs.

The math

Take a knowledge worker earning $80,000 per year. 19% of their productive time — roughly 8 hours per week — is spent searching for information.

That's $15,200 per year in salary cost for non-productive search activity. Adjust for benefits (1.3x multiplier), and you're at approximately $19,760 per employee per year in total compensation cost.

We model the recoverable portion more conservatively. Not all search time is eliminable — some amount of information gathering is inherent to knowledge work. Our estimate: 74% of that search time is eliminable with proper organizational intelligence infrastructure.

That gets you to approximately **$14,600 per employee per year** in recoverable productivity cost.

Where knowledge goes to die

In our research across 50+ enterprise teams, knowledge fragmentation follows a consistent pattern:

• **Decisions**: Made in Slack, documented nowhere, forgotten within 90 days
• **Context**: Embedded in email threads and Notion docs that nobody can find
• **Expertise**: Located in individual heads, not in any system
• **History**: Lost when employees leave

The compounding effect is what makes this expensive. A new engineer who spends 6 weeks ramping up instead of 2 weeks isn't just losing 4 weeks of productivity — they're making worse decisions during that ramp because they're missing context.

What fixing it is worth

At 100 employees, $14,600 per person is $1.46M per year in recoverable productivity. At 500 employees, it's $7.3M.

The right question isn't whether organizational intelligence infrastructure is worth investing in. It's whether the cost of that infrastructure is less than the cost of fragmentation. At current pricing for tools like archzOS, the payback period is typically under 60 days for teams over 50 people.

The harder question is organizational: who owns the problem of knowledge fragmentation? It's not clearly an engineering problem, a people ops problem, or a product problem. It sits at the intersection — which is exactly why it goes unsolved for so long.