← Research

Your engagement score is a spoiler with no story

Most teams collect more data than they read. The number isn't the problem — it's what it replaces.

Beltmar·Aug 3, 2026·3 min read
Share

A 72 walks into a pipeline review. Nobody asks where it came from.

That number — call it an engagement score, a health score, a lead score — did what it was designed to do. It compressed weeks of behavior into something a rep can glance at between calls. But compression has a cost, and the cost is the story. The sequence of pages visited, the pause before the pricing click, the return visit three days later from a different device. All of that became 72. And 72 doesn't remember any of it.

We've been looking at how teams actually use their scoring systems, and there's a pattern worth naming. The score gets built with genuine care — weighted events, decay functions, sometimes a real data scientist involved. Then it ships. And within a quarter, almost nobody on the team interacts with the underlying signals anymore. They interact with the number. The dashboard becomes a leaderboard: who's hot, who's not, who crossed the threshold. The journey that produced the score disappears behind it.

This isn't a failure of the score. It's a failure of attention. Scores are supposed to be shorthand. They become the whole language.

Here's what gets lost. Two accounts can both sit at 72 and mean completely different things. One visited your docs six times in two days — a technical evaluator going deep. The other opened every marketing email for three months but never once looked at a product page. Same number, opposite narratives. One is pulling toward a decision. The other is passively consuming content the way someone watches cable news in an airport. Treating them the same isn't just inefficient. It misreads what's actually happening.

The deeper issue is that over-collecting data doesn't fix this. Most teams we observe are capturing far more behavioral signals than they read. They have the page-by-page journey, the session duration, the scroll depth. It's all in the warehouse. But the workflow routes through the score, and the score is the only thing that triggers action. So the raw signal sits there, technically available, practically invisible. More collection, less reading. The ratio gets worse every year as tracking gets cheaper and attention stays expensive.

What would change if teams treated the journey as the primary artifact instead of the score? Not as a replacement — you still need a way to sort and prioritize — but as the thing you actually look at before you act. A rep who reads a two-week journey before picking up the phone has a different conversation than one who sees "72, qualified." They notice what the person lingered on. They notice what the person skipped. They walk into the call with a hypothesis instead of a script.

That takes more time, obviously. Not every lead deserves a close reading. But the ones near the threshold — the 68s and the 74s, the ones where the score is ambiguous — those are exactly the accounts where the compressed number fails hardest and the raw story matters most. The irony is that scoring was supposed to help teams focus. In practice, it often replaces the focus it was supposed to create.

One thing worth noting: some teams do this well without abandoning scores entirely. They use the score as a filter and the journey as the brief. Score gets you to the short list; the sequence of actual behavior tells you what to say and when. It's a small workflow change, but it treats the number as an index, not an answer. That distinction matters more than most scoring model improvements.

A number that tells you someone is engaged without telling you how is a spoiler for a story you never read.

More from Beltmar