Joe OviedoAt a certain point in designing a product, thinking about exceptions is more fun than testing the main path. Someone arrives with experience, so they should skip the introduction. Someone is short on time, so they should get an abbreviated version. Someone already has a team, so maybe they need another example. In an afternoon I can draw six paths and feel pretty sure I understand six kinds of people.
What I still don’t know is whether any of them could get through the first one.
I don’t mind personalization. I like it when a product recognizes something about my situation and saves me a step. What’s hard to justify is asking someone to tell me more about themselves when I don’t yet know what I’ll do with the information.
The question that seemed to solve everything
Imagine we’re making a resource to help someone organize their first week with a new client. We want them to leave with a simple plan: what they need from the client, what they’ll deliver, and when they’ll review the first round of work together.
We think of opening with a question: “Do you work on your own or with a team?” It sounds useful. Solo workers could get a personal checklist. Teams could get a template with a column for who’s responsible. It sounds like we’re adapting the resource to a real difference.
Now imagine someone picks “with a team,” opens the template, and stares at the date field. The client hasn’t sent the materials yet, so they don’t know what to enter. Someone else picks “on my own” and gets stuck in exactly the same place. They received different versions of the resource. Neither helps them decide what date they can commit to without what they need to start.
I could add another question: “Do you have all the materials?” But first I’d want to know what we’d tell people who say no. Maybe an example of agreeing on a date that depends on receiving the materials is enough. Or maybe the exercise is out of order: first list what’s missing, then who’ll send it, and only then talk about the first review.
That’s what interests me most about this example. The difference we’d planned for wasn’t what explained where people got stuck. And the first fix wasn’t to give them two more paths. It was to repair the step they shared.
What I’d watch in a first version
I’d try a shared path with a few people. I’d ask them to use it with a real client, not tell me whether the template looks clear. I want to see where they start, what they leave blank, and when they step away to open an email or talk to someone on their team. Leaving the template isn’t always a problem. Sometimes it’s exactly what they need to do. What matters is whether the resource helped them get there or they had to abandon it to figure out what to do.
I’d also resist doing the exercise for them. It’s easy to say, “Oh, put the date you agreed on with the client there,” and move on. If I had to say it, I want to write it down. Maybe the question is badly worded. Maybe I’ve assumed they’ve already agreed on a date. Maybe we don’t even need that field yet. Each explanation would lead me to change something different.
Then I’d look at which differences are actually worth addressing. Suppose that once we’ve sorted out the materials, solo workers can finish the plan in one sitting. People with teams, on the other hand, need someone else to confirm who’ll do each task before they send anything to the client. Now it makes sense to offer something different: a version they can share and fill in together. Not because they chose “I have a team,” but because we saw what they needed to do together and the first version didn’t let them.
Sometimes it’s enough to let people choose
Even then, I wouldn’t necessarily add another question up front. Maybe at the moment they assign the work, two options are enough: “I’ll organize this myself” and “I need to review it with someone else.” I’d rather ask when the person knows what they’re deciding. And I like letting them change their mind without starting over. Someone who works alone may still need to check a date with a collaborator. They don’t have to fit the category they picked on arrival.
I wouldn’t postpone differences that keep someone from using the product either. If the resource doesn’t work on a phone and people need it when they’re out of the office, that’s not personalization that can wait. It’s part of making the basic path work where people will use it.
Every extra path is another thing to maintain. If I fix an example tomorrow or change how people get the plan, I want to know how many versions I need to check so nobody gets outdated instructions. That’s worth doing when it removes a real obstacle. Less so when people get different greetings before reaching the same point where everyone gets stuck.
Before drawing another branch, I’d go back to that first week with the client. Can someone use what we made to write a clear message, ask for what’s missing, and agree on when they’ll talk again? If they can’t yet, I know where to work. If they can, and I see someone needs to do it differently, then I have a good reason to personalize it.
← Back to the Journal