Why "People Chips" is a taxonomy nightmare

If you just pulled a face, you're in good company.

The name has prompted considerable eye-rolling online, and understandably so. But mocking the label misses the more interesting question: how does a company with Google's resources, design teams, and testing infrastructure end up with a label that makes people uncomfortable?

The answer is a familiar one, and it happens to teams of all sizes.

This is the trap.

When you design a label, you are not just naming one thing. You are adding a word to a system that already has words in it. That system has logic, patterns, and relationships. Every new label either reinforces that logic or introduces friction.

The problem with People Chips is almost certainly not that someone thought "chip" was a great metaphor for a human being and ran with it. The problem is more likely that someone focused hard on naming this one feature, in isolation, without fully accounting for how it would sit alongside everything else.

That kind of debt accumulates quietly. And occasionally, one badly placed label is enough to crack the whole taxonomy open.

Let's look at what People Chips actually does.

It displays a card containing contact information. When you hover over a person's name in a spreadsheet, you get a summary: their photo, job title, email address, the usual. The information is pulled from metadata sources attached to the contact.

This is a contact card.

That is not a controversial statement. Contact cards have existed for decades. The concept is well established, the mental model is familiar, and the label is already in common use.

Google built a contact card and called it something else.

There are legitimate reasons why teams end up here. Trademark conflicts, internal naming conventions, or an existing feature that already holds the term you need: all of these can force you away from the natural label.

When that happens, you are not just solving a naming problem. You are asking your users to learn a new word for a thing they already have a word for. That is a significant ask, and it carries a real cost.

The cost is confusion at the point of use, increased support queries, and a quiet erosion of trust in your interface. Users do not know your internal constraints. They do not care about them. They just know that what they are looking at does not match what they expected to find.

If you are ever in this position, the only responsible path is to test the label extensively, not just with your usual participants, but with people who will push back and tell you when something feels wrong.

There is a broader point worth making here.

Good taxonomy is invisible. Users navigate it without noticing it because the words they encounter match the words already in their heads. The moment a label draws attention to itself, something has gone wrong.

People Chips draws attention to itself. It makes users pause, frown, and wonder what it means, which is exactly what a label must never do.

The fix in this case is simple: call it a contact card. It is a sub-type of card, a category that already exists and carries useful meaning. The label would slot in without friction.

Whether Google had a reason not to use it, we do not know. But the silence on that point is frustrating, because the taxonomy community would genuinely benefit from understanding their reasoning.

When large organisations make choices like this and explain them, everyone learns something. When they do not, we are all left guessing, and picking apart labels that should never have needed picking apart.

You do not need to be Google for this to matter.

If you are building or maintaining a taxonomy, the discipline is the same regardless of scale. Every label you add is a decision about the whole, not just the part. The question to ask is not "does this word describe the thing?" but "does this word belong in the system?"

Those are different questions. The first is about accuracy. The second is about fit.

Both matter. And getting both right is harder than it looks.

Want to know whether your taxonomy is pulling in the same direction as your navigation and search? That's exactly what a Murmuration diagnostic is for.