The message sent isn’t always the message received. New research reveals why.

David Bowman:

There’s a question worth asking at the start of every internal communications strategy session, and it rarely gets asked: are we actually measuring whether people understood what we sent, or just whether they opened it?

Fresh Intranet commissioned independent research earlier this year with 1,000 professionals and employees across the UK and US, 500 IC professionals and 500 employees, to explore this exact gap. What we found was uncomfortable, important, and honestly, not all that surprising once you see the numbers laid out.

We’re calling it the Employee Attention Recession. 

The read rate that should reframe everything

Only 12% of employees read internal communications in full. That’s the least common active response to receiving an internal communication. More employees are using AI to summarise content (20%) than are reading it in full.

This isn’t a disengagement story. Ninety-one percent of employees say internal communications feel relevant to them always or most of the time. They want to engage. The problem is volume. Eighty-three percent say they receive too much internal content. Thirty-five percent say far too much, they simply can’t keep up.

The single biggest factor in whether an employee reads something isn’t who sent it, how well it was designed, or how relevant the subject matter is. It’s how many other communications arrived before it. Attention is rationed, and your message is competing with everything else.

The gap between sending and understanding

What makes this moment particularly significant is what’s now happening in the space between send and read. Employees aren’t just skimming or ignoring content. Increasingly, they’re handing it to AI.

Eighty-eight percent of employees have used AI to summarise internal communications. Fifty-eight percent do so regularly. And in our sample, AI-generated summaries were the most commonly cited discovery channel for company news, ahead of the intranet, email from leadership, and Teams notifications.

The communication that leaves your hands and the one that arrives with your employees are no longer necessarily the same thing.

We call this the AI Visibility Gap, and it sits at the heart of the Intent Gap. Employees are confident that AI summarises their communications accurately (95% say so). IC professionals are worried about the opposite, 92% are concerned that AI summaries distort the intended meaning or tone of their content.

Neither group has reliable visibility into what’s really happening. Both are operating on assumption.

What IC professionals are measuring, and what’s missing

When we asked IC professionals how they measure content performance, the most common approaches were employee surveys and pulse checks (40%), engagement metrics like likes and comments (35%), AI-generated summaries and dashboards (35%), and intranet analytics (33%).

Notice what’s absent. Direct measurement of whether content was understood. Any visibility into how it performs once it’s been reinterpreted by AI or passed along informally. This mirrors wider sector findings: Gallagher’s 2026 State of the Sector found 70% of IC professionals are still tracking outputs rather than outcomes.

Forty-two percent of IC professionals believe their communications land well but have limited evidence. Only 33% have strong evidence that content is understood as intended. The rest are operating on confidence, not confirmation.

Three things the data suggests you do differently

The research points to three design principles that organisations closing this gap are already moving towards.

The first is subtraction. More content doesn’t solve an attention problem, it compounds one. Before adding a new channel, a new format, a new campaign, audit what you already publish. Treat every piece of content as a bid for a finite resource. Ask honestly whether it earns that bid.

The second is designing for mediation. If a significant proportion of employees will first encounter your content as an AI summary, the question isn’t whether to accept that, it’s how to design for it. Content that survives mediation puts its most important information first, makes intent explicit rather than relying on inferred tone, and uses clear structure so what matters most isn’t compressed away. Notably, the best AI-resilient content is also the best human-readable content.

The third is governing interpretation, not just publication. Nearly all IC professionals (98.6%) know employees are using AI to interpret their content. Only 33% have formal monitoring in place. That gap, near-universal awareness, minimal governance, is the most structurally significant finding in the report. The tools and processes to govern interpretation at scale are still being developed, but the absence of an established playbook isn’t a reason to wait.

Why intranet governance matters more than ever

There’s a dimension to this that sits upstream of communications strategy, and it’s where Fresh Intranet, the organisation behind this research, works most closely with its customers.

The quality of AI-generated summaries isn’t just a function of how good the AI is. It’s a function of how good the source material is. Stale content, unclear ownership, and an ungoverned intranet produce unreliable summaries. That’s a concrete operational risk, decisions made on misunderstood policy, change programmes that stall because the rationale didn’t survive compression.

Fresh is a multi-award-winning AI SharePoint-native intranet platform, built to bring governance, structure and clarity to the digital workplace. We work with medium to large organisations to make SharePoint actually work as an intranet, with clear content ownership, well-structured information architecture, and AI-powered search through FreshMind that helps employees find what they need. In 2026, ClearBox Consulting recognised Fresh with two Choice Awards: AI Innovator and Strong Knowledge & Search.

The future of internal communications isn’t about publishing better content. It’s about governing how that content is understood.

The full Employee Attention Recession report is free to download here.

+++

David Bowman is Product Director for Fresh Intranet, which conducted the Employee Attention Recession research.