Why the first screen of a product is so hard to get right, the sentence that settled a debate between 2 surfaces, and 4 books on deciding what deserves someone’s attention.
Homepages are so hard to get right. These meant-to-be springboard experiences have been designed thousands of times in the history of software, and a senior leader I work with said this week that he can’t think of one that’s truly great. I’ve been trying to come up with a counterexample since, and I’m not doing well.
A team I work with is rethinking the first screen of a product, and the conversation around it has been one of the better product debates I’ve been part of this year. Asked what the page is for, the first answer was the natural one: it shows you what needs your attention. The response from that same leader was sharp and fair. That’s the job of the entire product. If the home screen’s purpose is the product’s purpose, then it has no rule for what to leave out, and you end up rebuilding everything inside it, in tabs and modules, a little smaller.
His next question is the one I’d hand to any team designing a landing surface. If this is a hub and spoke model, what is the hub supposed to do? Does work get done there, or does it point you to where work gets done?
We’d already answered a version of that in a session between design, engineering, and product leads who had 2 surfaces that looked like they overlapped. One was a deep workspace for a specific kind of work. The other was the home screen, which had started showing some of the same tasks. The sentence that settled it was this: the workspace is the system of record, and home is the prioritization layer. One-off things that need a person today can surface on home. The full list, the bulk actions, and the history live in the workspace, and home links straight into it. One of our principal designers described the failure mode in a way I won’t forget. If home shows all your tasks except one category, and sends you elsewhere for those, that’s a Frankenstein experience, and it’s shipping the org chart.
If the home screen’s purpose is the product’s purpose, it has no rule for what to leave out.
He also offered a reframe in hindsight that I think is right. A home screen like this is a more capable inbox. Linear frames theirs that way, and it sets expectations well: things arrive, you deal with them, they go away. I’d add one difference. A good one tells you what’s happening and also what you should do about it, which is why another designer’s analogy fits: a really good executive assistant, who knows that not every outstanding item is equal and only brings you the ones that deserve you today. If everything gets in, you’ve recreated the notifications problem with nicer cards.
Two more lessons from the same conversation:
Experiment, don’t only iterate. The push from leadership was to stop refining one baseline and try radically different first screens with different audiences. Iterating teaches you how to improve the idea you have, and it takes an experiment to find out whether you have the right idea.
A beta label gives you room. Keeping a new home screen opt-in, clearly labeled as beta, means people who want it can have it, and the team can change course without breaking a promise. Everyone who uses software understands what beta means.
When the team documented what was unproven, including the page’s whole structure, the review got easier. Nobody had to pretend the first screen was solved, and given the track record across our industry, that seems like the right amount of humility.
Four books on deciding what deserves someone’s attention.
The shelf
