Meta's Muse Charm copies a trend Gen Z built for themselves. The question that tests whether you watch customers that closely.
On September 24, 2026, TechCrunch reported on Meta's newly announced Muse Charm, a small AI device built to dangle from a bag or keychain, and asked whether it would make AI approachable for everyday consumers or join the list of AI hardware that never took off. "Reactions, so far, have been mixed," the outlet wrote. "What the Charm is, however, is on-trend." TechCrunch traced that trend through the bag charm economy, including the Labubu craze that the outlet said grew Pop Mart's revenue to $1.8 billion in 2024, and then pointed to a more current reference than the Tamagotchi everyone had reached for: young people were already turning older Apple Watches into keychains and bag charms, "adopted by younger women as a fashion accessory first, rather than a piece of technology." The same piece argued that Meta "has abused consumer trust over the years to such an extent that it may never fully recover," and noted that Meta said it will monetize Muse "by taking a small fee from transactions."
Read the two halves of that report together. If TechCrunch's reading is right, nobody filed a feature request for an AI pendant; the product is a twist on a behaviour customers had invented for themselves with hardware they already owned. Someone saw that behaviour before it had a name and built toward it. That is the useful part of the phrase "customer obsession", and it is a behaviour rather than a sentiment. The second half is the harder lesson. Being on-trend and being trusted are different assets, and a company can hold the first while it is still paying down the second. Watching customers closely, acting before the ticket arrives, and knowing exactly what the customer is being asked to give up in return: those three things, in that order, are what an interviewer is listening for when they ask this.
Why they ask it
The question now appears in engineering and product loops well beyond the company that made the phrase famous. The interviewer is not checking whether you like customers. They are checking for two capabilities. The first is detection: can you find an unstated need from how people behave, rather than from what they submit as a request. The second is judgement: was "above and beyond" a decision about what the customer needed, with a cost you can name, or was it a display of effort. Senior candidates are expected to have a story where the observation came first and the work followed from it, not a story where the work was heroic and the customer was the backdrop.
The trap
The trap is the heroics story. The candidate stayed up all night, personally issued the refund, drove to the site, rewrote the integration over a weekend. It sounds like devotion. What the interviewer hears is a broken process rescued by adrenaline, and no evidence that the customer wanted any of it. A second version of the trap is quieter: telling customer obsession as agreement. The customer asked for something, you built exactly that, everyone was pleased. That is responsiveness, and it is fine, but it is not the question. The Charm is a useful contrast. On TechCrunch's reading it is on-trend because it follows a behaviour customers had not articulated, and it is at risk because, in the outlet's account, trust was spent elsewhere. A strong answer contains both halves: what you noticed that nobody asked for, and what you declined to do because it would have cost the customer something they would not have chosen.
Applying STAR-T
Situation. Set the scene with the signal, not the crisis. "Our analytics product had flat support volume and healthy renewals, but I noticed in session data that a cluster of accounts exported the same dashboard to a spreadsheet every Monday morning." That sentence already shows detection. The interviewer knows you were watching behaviour, not waiting for a ticket.
Task. Name the ownership you took, and say that it was not assigned. "Nobody had asked for anything. I decided that a weekly manual export was a need the product was failing to meet, and I owned finding out why before proposing a fix."
Action. Describe what you did to confirm the need, then what you built, in that order. "I shadowed a handful of those customers on a Monday. The export existed because an executive wanted a one-page summary in email, and our product could not send one. I proposed a scheduled summary report instead of the additional export formats those same customers had been asking for, and I argued that trade to the roadmap owner." The verification step is what separates obsession from guessing. Skip it and the story reverts to a feature you wanted to build.
Result. State the outcome in the customer's terms first, then the business's. "The Monday exports largely stopped in those accounts, the executives started forwarding the report internally, and renewals that had been flagged as at risk closed without a discount." Keep it concrete and keep it modest. Inflated results invite the follow-up you do not want.
Trade-off. This question rewards naming a cost, because customer obsession without a cost is just enthusiasm. "It pushed a planned admin feature by a quarter, and we told the customers who had asked for more export formats that they were not getting them. That was the right call, and it was not free." Notice the contrast with the Charm. TechCrunch frames the cost there as a privacy tradeoff in exchange for free AI, and is careful to say consumers may or may not accept it. Your answer should show that you knew what your customer was giving up, and that you chose something they would have chosen too.
The follow-up that breaks weak answers
"What did the customer actually ask for, and what happened to that request?" A heroics answer has no reply, because the story was never about a request. An agreement answer collapses in the other direction, because the request was the whole story and there was nothing observed underneath it. A strong answer has been waiting for this. You name the request, you explain why the observed behaviour pointed somewhere else, and you say plainly that you declined the request and told the customer why. The interviewer is listening for whether you can disappoint a customer in service of that same customer, which is the version of obsession that survives contact with a roadmap.
Score your answer against the director’s bar
Q: Tell me about a time that you went above and beyond to create customer obsession.
Bank this story now and rehearse the follow-up until the cost you named is the first thing you say. Try it free →