What a brand owes the people who tell it how they feel

The Feeling Is the Data Now

In 2023, Apple moved mood logging into the default Health app on every iPhone running iOS 17. No separate download, no onboarding, no explanation of what the data was for. Just a quiet prompt: how are you feeling right now? That single design decision — normalising emotional self-reporting at the scale of a billion devices — changed the ground beneath every product that had been built around the same ask.

How We Feel was built at the height of the pandemic to do something that public health infrastructure visibly couldn't: capture what people were actually experiencing day to day, in real time, at a granular emotional level. The premise was genuinely strange at the time. Asking someone to report a feeling — not a symptom, not a purchase, not a click — as a meaningful unit of community health data was a real departure from how data collection worked. The strangeness was the point. It was a way of seeing something that conventional instruments were missing.

That strangeness has since been absorbed into the default layer of consumer technology. Mood logging is now a standard feature of the world's most widely used consumer health platform. Which raises a question that James Hurst's framework — the one at the centre of Weird the Normal / Normal the Weird — seems almost designed to complicate: what happens when someone else normalises your weird?

The book's core argument is that neither move works alone. Estranging the familiar exposes what's hiding underneath it; giving strange ideas enough structure to survive brings them into contact with the world. The discipline, James argues, is the switching — reading the moment and knowing which direction to move. But the framework assumes you're the one doing the switching. When Apple normalised mood logging at the scale of a billion devices, the person who built How We Feel was no longer in control of that switch.

Woebot ran into an adjacent version of this. Alison Darcy launched it out of Stanford in 2017 as a mental-health chatbot with genuine clinical framing — cognitive behavioural techniques, research backing, a therapeutic register. It scaled fast during the pandemic, which is when the tension between its clinical credibility and its consumer product reality became harder to hold together. The more people used it, the more the question pressed: is this therapy, or is this a product that feels like therapy? The clinical framing was the trust mechanism. But trust built on a clinical frame is fragile in a way that trust built on experience isn't — because the frame can be examined, questioned, and found to be approximate.

How We Feel sits in a different position, but the underlying question is the same one. When a product is built around emotional data — when the thing being collected is not behaviour but feeling — what is the implicit contract with the person reporting? And what happens to that contract when the data becomes, in some form, the product itself?

This is where James' working belief about brands gets genuinely tested. The argument running through Use Design To Design Change is that the product builds the brand — that marketing can introduce but only experience can convince. Applied to emotional data products, that argument becomes almost uncomfortably precise. The experience of reporting how you feel is also the experience of being known, or of being processed, or of being studied. The brand relationship lives in which of those it turns out to be. That's not something a positioning statement resolves. It's something the product either earns or doesn't, over time, through the texture of how it handles what it's been given.

The interesting design question — one about product architecture, not just policy — is whether there's a way to make the handling of emotional data legible as an act of care rather than an act of extraction. Some products have tried showing people their own data back, giving them tools to interpret it, making the collection feel reciprocal. Others have leaned into clinical partnerships as a form of trust by association. When Woebot discovered that institutional credibility was not the same as institutional trust, the gap between those two approaches became visible.

What Apple's move into mood logging actually did — beyond normalising the gesture — was shift the competitive question. Before iOS 17, the question for a product like How We Feel was roughly: can we make this strange ask feel natural enough that people will do it? After iOS 17, that question is largely settled. People will do it. The new question is harder: now that everyone is collecting this data, what does it mean to be the product that people actually trust with it?

James' framework would probably say this is a switch moment — that the weird has been normalised by someone else, which means the interesting move now is to find what's strange again inside the newly familiar. But what's strange inside emotional data collection now isn't the act of reporting a feeling. It might be something closer to the question of what a brand owes someone who has told it, repeatedly and honestly, how they feel.

What would it look like to design around that obligation rather than around the data it produces?