“Looks Great”
Turns a First Guess Into Your Plan.

Someone on your team asks for five minutes. They turn their laptop around, or they hold up their phone. And they show you a fix for something that's been going wrong. It's already built. Maybe they made it with AI over a few days. You can click through it. It works, and it looks great.

So you say, "Looks great." And when it goes up the line, maybe your own boss says it too. And the only other thing anyone asks is, when can we roll it out?

So what just happened? They saw something going wrong, and they made a first guess at why. Then they built a fix for that guess. Nobody checked the guess. And now it's your plan.

That's what I want to talk about today, how "looks great" turns a first guess into your plan.

This is The Autistic Leader. I'm Uday Kiran Bolusani. Each week, I take one piece of advice every leader already gives, and I give you one skill that makes it work for every kind of mind.

Now, the advice behind that moment. "Bring me solutions, not problems." You've probably said it, or had it said to you, right?

It's an old management saying, and it stands for a good habit. The idea is, whoever spots the problem works out the details, then brings you one finished answer for a yes or a no. There are two good reasons behind it. One is to keep half-finished ideas away from you, the one who decides. The other is to teach people to think for themselves.

The advice comes from good intentions. But in the rush to a quick answer, a team can skip a skill by accident. It's the skill of working out what's actually going wrong. And what you'd need to see to know it's solved.

Look at what the advice leaves out. When you say it, your team can't tell what you think the problem is. And they can't tell what kind of fix you'd say yes to. So your team guesses. They might guess from your tone, or from the plans you've said yes to before. Anyone on your team can end up guessing, however their mind works.

So now they're doing two jobs at once. One is working out what's going wrong. The other is predicting what you'll approve. That second job takes real mental effort. That's what's called cognitive load. And it pulls their attention away from the problem itself.

And over time, people can learn something from all that guessing. They may start describing each problem so it matches a fix they already have. Because that's what gets approved.

So what does that cost your organization? You already pay your people to think problems through. When you ask only for the fix, what they noticed along the way stays with them.

And this advice can cost you the most when a problem is complicated. You may have heard leaders use the word VUCA. It stands for volatility, uncertainty, complexity and ambiguity. A lot of what leaders deal with looks like that.

Think about a problem with a lot of parts that affect each other. Or one where nobody's pinned down the causes yet. Your team's first guess is only one way to explain it. Ask for solutions first, and you can miss a critical opportunity to see it another way. You can miss the other things driving it, and why it's hurting the business.

And with AI, a fix can reach you already built and running. I think a version that already works is harder to set aside than a plan on paper. It feels real and finished, so you're less likely to ask what else could be going on.

And it's good that people on your team can build that fast. You want people who can do that. But when a fix is this easy to get, the critical skill is working out what's going wrong, and what outcome you need.

I'm autistic. I notice patterns, and I come at a problem from a different angle than the rest of the room. That's how my mind works.

Earlier in my career, with the leaders who saw the value in that, I thrived. Other leaders found the same way of thinking a burden. They saw it as a sign I wasn't executing, or as a challenge to views they'd held for a long time.

Looking back, I read that difference as a design question. It's about what a leader asks for, and whether that kind of thinking ever reaches them.

And that's true on your team, too. Someone on your team may see what's going wrong from an angle nobody else has tried. Whether you ever hear it depends on the question you ask.

So here's the skill. It's three questions, and the order matters.

When someone brings you a fix, even one that already runs, you put it to one side. And before you look at it, you ask three questions.

The first question is, "What is happening, and what outcome do we need?" You're asking for the facts, and for what you'd need to see to know it's solved.

The second question is, "What is driving it?" You want every cause they can name. The cause their fix was built for is just one of them.

The third question is, "What else could it be?" This is where someone on your team gets to say, maybe we're looking at a different problem.

Then you ask what they'd do. And you make the call.

And you're like, okay, but it already works. What's the problem? Fair enough. What they built will still be there after the three questions. It may well end up in what they recommend.

Let's try it on a clinic first. Say you oversee a clinic, and since spring, more and more patients have been missing their appointments. Your clinic manager spends a weekend with AI and builds a reminder app. It texts each patient the day before, with a button to confirm. And they want to show you a sample text on their phone, with a made-up patient. So before you look, you ask the three questions.

"What is happening, and what outcome do we need?" Your clinic manager says, "Missed appointments are up by about a third since spring. Each one is a visit another patient could've had. We need them back where they were."

"What is driving it?" They say, "Patients forgetting. The 7 a.m. slots we added in spring. And the new patients who joined this year."

"What else could it be?" So your clinic manager goes and asks the nurses. And guess what? Nearly all of the extra missed appointments are at seven in the morning. The first bus gets there after seven. So the bus times may be behind a lot of those missed appointments. Some patients can't get there by seven, even when they remember.

One of your nurses had already noticed who was missing at seven. They'd never raised it, because nobody had asked what else could be going on.

Then you ask your clinic manager what they'd do. They say, move the patients who come by bus to later slots. Offer 7 a.m. to the patients who drive. And send the reminder texts to everyone.

So the app still goes out. Reminders do help more patients keep their appointments, so the app's a reasonable fix. Your clinic manager is still accountable for fixing the missed appointments, and the recommendation is still theirs. They just don't have to guess what you had in mind anymore.

Now a second team. Say you run operations, and your deliveries have started running late. One of your dispatchers spends a few evenings with AI and builds a route planner. It reorders each driver's stops to cut the drive time. Your dispatcher wants to show it to you with a day's worth of made-up orders. So again, before you look, you ask the three questions.

"What is happening, and what outcome do we need?" Your dispatcher says, "For about six weeks, more of our deliveries have been arriving after the window we promise customers. We need them back inside that window."

"What is driving it?" They say, "More stops on each route since we took on new customers. Traffic. And two new drivers who are still learning the routes."

"What else could it be?" Your dispatcher goes and checks when the trucks leave the warehouse. About six weeks ago, your warehouse crew moved loading from the night before to the same morning. So now the trucks leave about an hour later. That means the late deliveries may start at the warehouse, before any driver's on the road.

Then you ask what they'd do. Your dispatcher says, go back to loading the trucks the night before, so they leave on time. And use the new route planner on the days with the most stops.

And one more, from a product team. Say you lead the team behind your company's app, and more people are quitting partway through sign-up. A designer on your team uses AI to build a shorter sign-up form in two days. It's three screens down to one. Your designer wants to show it to you on a test account. And before you look, you ask the three questions.

"What is happening, and what outcome do we need?" Your designer says, "Since our last release, more people start signing up and never finish. We need more of them to finish, and to use the app in their first week."

"What is driving it?" They say, "A form that's too long. The new step where people confirm their email. And more people signing up on their phones."

"What else could it be?" Your designer looks at where people stop. Most of them stop at the email step, and that confirmation email can take up to an hour to arrive. So people may be giving up while they wait for an email.

Then you ask what they'd do. Your designer says, let people start using the app right away, and confirm their email later. And ship the shorter form, too.

Product work has its own version of this lesson. Product people have a name for a team that's measured only by the features it ships. They call it a feature factory. And product leaders who work the other way give a team a problem to solve, with the reasons it matters. Then they measure that team by the outcome.

You already ask what outcome you need in your first question. That part is critical. So whatever fix someone brings you, start there. Then you can hear how each person on your team sees the problem, and that can make your call better.

Now try it on one of yours. Think of the last fix someone showed you. It might already be running. Ask yourself the first question right now. "What is happening, and what outcome do we need?" Then the second. "What is driving it?" And then the third. "What else could it be?"

Facts first, then their recommendation, then your call. That order comes from my Decision Charter, my framework for who does which part of a decision. In it, whoever's closest to the work lays out what's true first, without yet making a recommendation. On those teams, that's your clinic manager, your dispatcher and your designer.

And this is where AI helps. Your dispatcher can ask AI to list what might be making the deliveries late, and to draft other explanations for the delays. AI can do that part quickly. Then your dispatcher checks each one against what they've seen at the warehouse, and decides which ones matter. That split comes from my AI Cognitive Strategy Matrix. It's my framework for which parts of the work AI can take on, and which your people keep for judgment.

So with AI in the work, your dispatcher has more possible causes to weigh than before, and still has to judge them well. I think that makes your people's judgment worth more to your organization.

So what did those three teams have in common? The app, the route planner and the shorter form were all built and working before anyone asked what was going wrong. When you ask your team for solutions first, they can end up solving only their first guess at the problem.

When you ask about what's going wrong first, a few things change. Whoever sees it from another angle gets heard, like your nurse at the clinic. You learn what your people noticed before anything gets rolled out. And your team still brings you a plan. Your clinic manager's plan starts with the patients who can't get there by seven.

There's an essay that goes with this episode, and it puts all of this on one card. I call it the Operating Brief, and it has four parts.

The moment. Someone shows you a fix that already runs. You approve it before anyone's asked what's going wrong.

The inherited design. That's the advice you were handed, "Bring me solutions, not problems." It was built to keep half-finished ideas away from you. And it leaves your team guessing what you think is wrong, and what you'll approve.

The redesign. Before you look at what they built, you ask your three questions. Then you hear what they'd do, and you make the call.

And the Monday test, the one thing to try this week. The next time someone on your team wants to show you a fix, ask your three questions before you look at it.

The essay also prints a prompt. You paste it into Claude, or whatever AI your organization has approved for work. Then you add a few lines on what's going wrong, and what the fix does. Take out names first, and anything else that could identify a person. The AI restates the problem without the fix, and lists what could be driving it. Then it drafts two other explanations, with questions to ask whoever built the fix. And the prompt tells it to leave the call to you.

The essay's on my site, under the title "'Bring Me Solutions' Puts the Fix Before the Problem." The three questions are printed there with the clinic manager's answers, so you can copy them for your own team. The Operating Brief is there as a card you can share. And the prompt's printed in full, so you can copy it straight into your AI.

Here's my ask. I built this show to help leaders who think differently reach senior roles, by changing how organizations are designed. Every person who subscribes is one more leader this work reaches, every week. So subscribe to the Substack, and take each week's skill to your team. And if you'd like, follow the show wherever you're listening.

So this week, pick the next fix someone wants to show you, and ask your three questions first.