Speaker A
Are you designing a BlackBerry in the iPhone era? Probably, because most teams are still shipping outdated software without even realizing it. But then you have players like Intercom and the browser company who have seen the next curve. They're not just adding AI to interfaces, they're replacing interfaces with outcomes. And in this video, we'll learn what that means in the AI era, why these companies are doing it, and the framework we can use to design outcome-based interfaces that will thrive. But before we get into any of that, we need to understand what actually happened to poor old BlackBerry. Because most people forget BlackBerry was winning. They owned almost half the smartphone market in the US. Presidents used it. CEOs swore by it. And then something happened. An Apple keynote with Steve Jobs announcing the iPhone. It didn't take long before texting with keyboards and navigating with these odd scroll balls was a thing of the past. And at this point, it was clear the phone game had changed forever. BlackBerry didn't fail because they were bad at building phones. They failed because the rules of the game had changed and they didn't realize it until it was too late. The iPhone introduced a new interaction layer which was touch first and app-centric. Well, BlackBerry was still building for a world people were walking away from. Today, the same shift is happening. Interfaces like keyboards back then are now becoming less important because with AI, we're forced to think much more about outcomes and much less about screens, something most startups are still missing. We'll still see today's app builders pad AI features onto legacy interfaces. And while that's understandable since everyone else is doing it, it's the wrong approach entirely. We're at a stage in tech where we need to rethink what an app even is. And this is something the browser company understood when they, despite a bunch of backlash, decided to sunset their innovative and hugely popular browser called Arc. Before we dig into that though, if you're looking for help in exploring how to design your product for the future, we offer free monthly design strategy sessions at ZipSap. Just click the link below and remember, we only have a few spots per month, so don't miss out. Now, Arc was ambitious and beautiful. While it was clearly very innovative at its core, it was still based on an old model of how interfaces should behave. And this was something leadership understood. So instead of incrementally iterating and padding it with AI features, they announced to the world that Arc would be no more. In its place, they would build a re-imagined browser called Dia. What makes Dia feel like a real leap, not just a facelift, is this: it's not trying to deliver a better browsing experience. It's trying to deliver browsing outcomes without making you browse. And that's the key. This isn't outcome-first design as we knew it. It's outcome-first design for the AI era where the journey itself often disappears entirely. You don't navigate, you just ask and it answers. Dia collapses the flow and gives you what you came for faster than you even knew was possible. And this is why it's clear that Dia isn't trying to be a better BlackBerry. It's aiming to become the iPhone moment for browsing. However, they're not the only ones aiming for this. But before we get into the lessons and the frameworks we can take with us from this, let's look at an arguably more interesting case of recent outcome-first design in the AI space. Billion-dollar customer service behemoth Intercom. They saw the same exact wave coming. But the crazy thing about them is how they were already a category leader in customer support. They have a super clear product-market fit, huge brand equity, and thousands of customers, which makes it kind of insane to think that instead of protecting this old playbook that has been working for them, they just threw it out because, as their chief product officer Paul Adams said, within a week of ChatGPT launching, they paused the roadmap and built Finn, an AI-native product focused on one thing: resolving questions instantly. Finn, much like Intercom, still has a dashboard for now, but the product is outcome-first. Customers get a direct AI-powered resolution before they ever touch a help article or a human rep. And this is even reflected in their pricing model, which is based on resolved tickets, aka successful outcomes. The interface didn't disappear, but its role changed. The outcome comes first and everything else gets out of the way. And as a result of this, Finn is now Intercom's fastest growing product ever and has become the company's main focus, which is really cool. But how do we learn from this? And how do you apply it to a smaller-sized startup, for example? Well, I'll go into that in just a moment. But first, if you're looking for someone to talk to about these strategic product and design questions, check out the free strategy calls in the link down below. Now, designing for outcomes is nothing new. But with AI, it's just different. We can now design for outcomes by skipping steps entirely in some cases. Which begs the question, how do we evaluate what those cases are? The cases where AI will disrupt our product or where it will unlock shorter paths to outcomes. Well, it starts with a simple but powerful matrix. First, what type of value does your product deliver? Is it utility or experience? Is the user's core task repetitive or nuanced? Place your product or even individual flows in your product inside a matrix and you'll get a rough idea of where AI can take over the flow completely versus where the experience still matters. When something is highly repetitive, mechanical, rule-based, and utility-focused, that's a red flag. You're in an AI dangerous zone at this point. These are the kinds of interfaces that will be automated, maybe even removed or rebuilt from scratch. Take something like Adobe Premiere, for example. A huge part of professional editing is rule-based, like trimming gaps, fine-tuning audio, syncing tracks, removing silences. And all of this is now automatable by AI in some way or form. So, if you're building the next editing app, the question is no longer how do we make the timeline easier. It's rather, does the user even need to touch the timeline at all? Now, on the other side of the spectrum, we have areas with low AI risk and these are typically apps that are emotion-oriented, judgment-heavy, or based on interaction itself as the value. For example, a game. While you're trying to get to an outcome, winning the game, the gameplay itself is the goal. In this case, AI can amplify the experience like adjusting difficulty or generating narrative, but it won't replace the UI or remove the human from the loop. In fact, designing for delight becomes even more important in a case like this. So, can AI replace this interface? Isn't the question. The question is, is the interface just a means to an end or is it the experience itself? Now, there are also some interesting edge cases where full outcome-first automation or invisible UI might not make sense. For example, say you're checking your bank balance in a co-working space. You probably don't want to say it out loud and you definitely don't want an assistant reading it back to you. In this case, the invisible interface like voice might take a step back. But of course, AI can still play a support role in the background like flagging unusual transactions, predicting upcoming bills, or even proactively nudging you when your spending spikes. And the same goes for information-heavy interfaces like investment dashboards. Users often need to see multiple data points side by side, scan visual patterns, and make human decisions fast. Therefore, interfaces here need to stay visible and flexible with outcome-oriented automation for summarizing trends, recommending actions, or highlighting anomalies. So yes, there are edge cases where full removal of UI or re
