Apple had lost the plot. Did they rediscover it?

Last year I sold almost all of an Apple position I’d spent fifteen years building. The decision had nothing to do with the stock price. I assumed it would keep climbing for a while, the way Intel kept climbing long after it stopped being the company that defined its industry. Great companies run on a surplus: being the default choice, market position, momentum, the accumulated trust of millions of people. That surplus can carry a company for years after the thing that earned it has gone rotten.

I sold because I had stopped believing Apple. And that meant I stopped believing in Apple.

The moment that broke it for me was the advertising. In 2024, Apple built a campaign around Apple Intelligence and presented a slate of features as the reason to buy that year’s iPhone. The features did not exist. Apple had told its customers to buy a phone for capabilities it could not deliver, under pressure from a market that badly wanted it to have an AI story. John Gruber’s Something Is Rotten in the State of Cupertino was widely shared at the time because it perfectly captures the sense of betrayal. This wasn’t a slipped schedule; it was an indication that Apple allowed itself to be dishonest internally which led to them lying to customers.

What Apple used to be

That is the kind of thing plenty of companies do. It was not the kind of thing Apple did.

The Apple I had invested in, the one that took shape in the years after Steve Jobs’ return, had a discipline about exactly this. It showed you things it could actually deliver. It shipped things you hadn’t asked for and didn’t know you wanted, then made them feel obvious the second you used them. The focal point was always the experience of the person holding the product.

For a long time, Apple was the proof that sweating the details paid off. I could hold them up and credibly claim that delighting customers in a way that doesn’t appear on a spreadsheet had an effect on the profitability and sustainability of the company. And that the actual details did matter.

There used to be a simple way to settle a disagreement about how an interface should behave. You opened one of Apple’s own apps and copied what they did, because they had thought about it harder than you had.

How far it drifted

That assumption slowly started slipping. As Apple became larger, the number of apps that were shipped and never updated got longer. The APIs followed a very similar path of impressive expansion, but missing, forgotten or broken pieces. I remember years ago blowing a fellow engineer’s mind when I told him I had originally taught myself Objective C based off of how good Apple’s documentation was at the iPhone launch. He had come to the platform later and just assumed the docs were always thin.

The Mac in particular lost its footing. When the iPhone first launched, Apple’s own developer sessions were adamant that you could not shrink a Mac app onto a phone; you had to rethink it as a real iPhone app. Years later Apple seemed to argue the reverse, that you could inflate an iPhone app, call it a Mac app, and ship it. Liquid Glass underscored the same drift. I never minded it on the iPhone and I never installed it on my Mac, but it clarified something I’d been feeling for a while: Apple kept pushing its platforms toward a strange parity instead of asking what each one’s users actually needed.

And finally, after Gruber posted the aforementioned criticism, Apple stopped sending its executives to his Talk Show Live. You can read that however you want. I read it as a company that had grown allergic to hearing the truth about itself.

Then they talked about corner radii

So I went into this year’s WWDC expecting very little. I had heard the rumors of 2024’s promises, but this time it works. I knew they were going to be using Google’s Gemini as their base model, which sounded promising. But you know, fool me once…

Early on, about 7 minutes in, Shubham Kedia came on to talk about Design. First, he said that listening and incorporating feedback and developers was an important part of a new design. And then he talked about the rounded corners of windows in macOS and I sat up in my chair. For the past year those corners had been a quiet disaster. Depending on how an app was built you got possibly three different radii, and in the common case the rounding pushed the resize handle out past the edge of the window, so you had to click into empty space to grab it. Experienced Mac users were taking two or three tries to resize a window. Nitpicky details like this are what defined the Mac. PC people would stare at you blankly and wonder what’s wrong with you if you tried to explain what makes a Mac “nicer”, but it was precisely the lack of these little paper cuts.

We hadn’t gotten to anything even vaguely AI related, but already the keynote said they heard the complaining and they understood why it mattered. That made me take notice.

The second big little detail I noticed was search. Stacey Ford even said “We’ve all had that moment where you search for something you know is there, but it just won’t show up.” Yes! Sometimes as a user you wonder “doesn’t Apple use these products? Don’t they get hit by this bug too?” The search in Mail had become comically bad. My cofounder Joe brings it up to me every few weeks: he knows an email exists, he can picture it, but he cannot find the incantation that surfaces it again. That is no small thing when the message you can’t find is the one you need to do your job. And now Apple says they’ve rebuilt the search index from the ground up. There’s a theory I find persuasive that part of why the 2024 Siri effort collapsed is that Siri was sitting on top of that same broken search. You cannot build intelligent retrieval on an index that doesn’t work. Before you can make search smart, you have to make search work.

The unglamorous work

None of this is glamorous, and that is what makes it meaningful. Inside a company the size of Apple, nobody builds a career by staying on Mail for three years and quietly making it better. The rewards go to the people shipping the new thing, running the team that gets keynote time, announcing the big initiative. Hunting down the obscure bug that drives a daily user up the wall is a thankless assignment, which is why most large organizations simply let it rot.

So when a company chooses to do it anyway, that’s a leadership decision. Someone had to stand up and say that this cycle, finding and killing the boring, infuriating bugs is the work that gets rewarded. And that requires telling your own teams an uncomfortable truth: what you shipped isn’t good enough yet, and we are going to spend this year fixing it instead of building the next exciting thing.

There’s an old story about Jony Ive and Steve Jobs. Ive once hesitated to tell the design team that something wasn’t good enough, because they’d worked hard and he didn’t want to discourage them. Jobs’s response was that if Ive really cared about the product, he’d tell them the truth. People sometimes take that as proof Jobs was cruel. Maybe he was. There’s certainly evidence in both directions.

But the reality is that while you can certainly deliver that truth as a jerk, or you can deliver it in a way that inspires your team toward something better, you simply cannot lead a team to do great work if you can’t tell them the truth. Shielding people from honest feedback isn’t kindness. It is caring more about their comfort than about the thing you are all supposedly there to build.

The main thing

The best move is to keep the main thing the main thing in the first place. Apple’s main thing was always the same: care about the experience of the person using the product, down to details most companies would consider nitpicky and esoteric. That care was never a marketing posture. It was the actual source of the loyalty, the differentiation, and the surplus that let the company coast even while it drifted.

The second-best move, once you’ve lost it, is to stop, admit it, and do the slow work of earning it back. You don’t get there with a more confident keynote. You get there by fixing the corners, rebuilding the index, and rewarding the people doing the invisible work of making things right.

I sold most of my Apple stock because I thought they had forgotten all of this. This year, for the first time in a while, I’m cautiously optimistic. But the lesson underneath applies to far more companies than Apple. The companies that win the long game are the ones that treat the small, boring, invisible details as the actual product, because that is what they are. Lose sight of that and even an enormous surplus runs out. Remember it, even late, and you can (hopefully) still find your way back.