← All posts
Operating Cadence   Aug 6, 2026 · 6 min read · by Peter Vin

The Compression Chain: How a Comment Thread Becomes a Colour

A status report loses information at every hop. The blocker in a comment thread becomes a sentence in an update, the sentence becomes a percentage, the percentage becomes a colour on a slide. Each step is a reasonable summary of the one before it, and the end result carries almost none of what made the original signal worth having.

This is not a competence problem. Every person in the chain is doing their job correctly. Compression is what summarising means.

The four hops

Hop one: the work to the update. An engineer writes in a ticket comment that the approach will not scale past a certain volume and they need a decision on the data model. The team lead writes "integration work continuing, minor technical questions open."

Hop two: the update to the status. Several team updates get merged into a project status. The technical question is one of eleven items. It becomes part of "on track."

Hop three: the status to the number. The project is reported at 70 percent complete. The open decision is inside the 30 percent, but the number does not say so, because percent complete is not progress.

Hop four: the number to the colour. 70 percent with no flagged risk renders green on the portfolio view. The board pack shows a green square.

The decision maker sees the green square. The thing that would change their decision was lost at hop one and had four chances to disappear after that.

Why more reporting makes it worse

The standard response to a missed surprise is to add a reporting layer. More fields, a weekly written update, an extra review. Each addition adds another hop, which adds another compression, which loses more of the specifics while consuming more of the week. The cost of the weekly status meeting is usually calculated in hours, and the hours are the smaller loss.

There is a second failure that arrives with the compressed artefact. When ten people read the same green 70 percent, they read ten different meanings into it. One sees permission to hold course. Another sees a project close enough to done that resources can move off it. A third sees something worth pushing extra effort at to land it this quarter. The number is identical and the plans diverge, which is interpretation debt accumulating in plain sight.

What survives compression and what does not

Some information is compression-resistant. A date. A number with a target next to it. A binary shipped or not shipped.

What never survives is the reason. Reasons are specific, contingent, and long. They live in threads. "We can do this by the 30th, but only if the pricing decision lands this week, and if it slips we should probably stop rather than half-build it." Compress that and you get "on track."

What also never survives is history. A KPI sitting at 82 percent of target after climbing steadily for a month and a KPI sitting at 82 percent after falling from 95 look identical in the cell. The trajectory carries the signal and the snapshot deletes it.

And what never survives is the link back to the goal. By hop four, the slide shows project health, not whether the project still serves an objective the company is pursuing. That is the strategy execution gap forming inside a reporting process rather than despite one.

The alternative is not a better template

Rewriting the status template shortens hop two. It does nothing about hops one, three, and four. The structural fix is to stop reconstructing the picture out of summaries and read the underlying record instead.

That means the review reads the actual tasks connected to the goal, with their comments and their history attached, rather than a slide assembled the night before. When a KPI is behind, the question "why" gets answered by opening the work beneath it and finding the thread where the problem first came up, which is roughly what a QBR built on traceable work looks like in practice.

Two things change when the review runs that way. Preparation stops being a production exercise, because there is nothing to assemble. And the specifics reach the decision maker intact, because nobody had to compress them into a colour to get them into the room.

A test for your own chain

Pick one green item from the last board pack. Trace it backwards: which project, which tasks, which person, which thread. Count the hops and note what was true at hop one that is no longer visible at hop four.

Most teams find at least one item where the original signal contradicted the colour. Not because anyone misreported, but because the chain worked exactly as designed.