← Insights·Startups

You're Still the Solver: Justify the Takeover First

A successful rescue doesn't prove you needed to take over. Check the standard before replacing someone's answer.

By Ken Gavranovic · October 29, 2026

The hard problem lands on your desk, and you can see an answer before the person finishes explaining it. You take it, fix it, and get back to the rest of your day.

Except the rest of your day is more of the same.

If this is your standing role across the company, the problem may be bigger than a crowded calendar. The solving habit never retired. It got rewarded. Your team gets relief; you get the satisfaction of being useful. Meanwhile, setting direction and developing an executive team get whatever time is left.

I wouldn't start by telling you to solve less. I'd make you justify the takeover: what does this situation need from you that its owner can't provide?

Why being useful can keep you stuck

I've written before about how routinely supplying answers can teach people to bring you the next problem. The harder part is what that arrangement teaches you.

A visible save gives you a result now, often with appreciation attached. Helping someone develop judgment takes longer, and the benefit may show up in a decision you never hear about. If your sense of contribution rests on having the answer, that can feel like a demotion.

That's why a successful intervention is a poor test of whether it was necessary. You solved the problem. You haven't established that you needed to be the solver.

A better calendar can protect time. It won't, by itself, change what you do when someone brings a problem you enjoy solving.

Some problems do need you

Early in a business, hands-on solving may be essential. Even in a large company, a new venture or an important technical problem may deserve your close attention. A well-led, established area may need much less.

Some decisions stay yours by design. When someone needs a clear standard or a decision only you can make, give it. Coaching isn't a way to avoid either.

An escalation can also be sensible when the decision affects another function, exceeds someone's authority or requires information they can't access. Supply what's missing rather than automatically taking over the whole problem.

And if nobody has time to own the work, willingness to let go won't create capacity. Reduce commitments, change staffing or keep an explicit temporary role yourself.

Replace the percentage with a standard

The 80 percent rule says to let someone proceed when their solution gets most of the way to the goal and won't cause a catastrophe. I like the permission to accept a different approach. I wouldn't use the percentage as the decision threshold.

Eighty percent of what? Your preferred method isn't the business requirement. And avoiding catastrophe is too low a bar.

Instead, agree on the outcome, deadline and quality the work must meet. Check that the owner has the skill, authority and time to make the call. Then decide what downside is acceptable and how you'll detect trouble while there's still time to act.

Suppose your head of support wants to clear a backlog by grouping cases differently than you would. If the plan fits the available staffing and meets the required customer response times, your preference alone isn't a reason to replace it. If it puts a service commitment at risk, name that gap and work through it.

Don't relax safety or legal requirements for a learning opportunity. Narrow the decision or add review before an irreversible step.

Look at the moments you take over

Try a short Solver Log. It's a practical exercise, not a validated diagnostic.

Reconstruct one recent week from your calendar and messages. For each problem you personally fixed, record what it was, who owned it, what they needed from you and what you actually did. Include the work you took over without being asked.

Then add the part worth examining: why did you intervene? A missing approval? A real risk? A gap in their knowledge? Or an answer that wasn't the one you would have given?

Check a few entries with the people involved. Your memory of a rescue may leave out the authority or information they had been asking you for.

An empty owner column doesn't prove understaffing. It may reflect an unassigned responsibility, unclear authority or a genuine capacity shortage. The log gives you questions to investigate, not a diagnosis of anyone's character.

Pick one recurring category where a capable owner has room to act. Agree on the decisions they'll make, the standard they'll meet and when you'll review the work. Be explicit about what should bring you in sooner, and tell the affected team who owns the call.

On the next few cases, ask for the person's recommendation and the help they need before offering your answer. When you feel the urge to take over, write down the specific requirement or risk that justifies it. If you intervene, record what you changed.

Keep bad news welcome. Asking for help shouldn't automatically cost someone ownership.

Give the time you're trying to recover a job, too. Name one upcoming demand the company needs you to prepare for, such as a financing or launch, and check whether your week actually makes room for it.

At the agreed review, look at the work and your interventions. Did the owner meet the standard? Did your involvement address a material problem, or mostly change the method? If results fell short, examine the support, workload and capability before deciding what to change. If the owner keeps missing an explicit standard despite adequate support, address that performance problem directly.

Under pressure, you may want to take everything back. That urge isn't proof that delegation is working or failing. Use the work to judge.

If the work meets the agreed standard and your main complaint is that you would have done it differently, leave it alone.

Frequently asked questions

How do I stop solving my team's problems for them?

Before you take anything over, make yourself justify it by asking what the situation needs from you that its owner can't provide, because a successful intervention is a poor test of whether it was necessary. The habit never retired because it kept getting rewarded: your team gets relief, you get the satisfaction of being useful, and setting direction and developing an executive team get whatever time is left. A better calendar can protect time but won't by itself change what you do when someone brings a problem you enjoy solving, so give the time you're trying to recover a job too: name one upcoming demand the company needs you to prepare for, such as a financing or a launch, and check whether your week actually makes room for it.

When does a problem actually need me to step in and solve it?

Step in when someone needs a clear standard or a decision only you can make, and give it, because coaching isn't a way to avoid either. An escalation also makes sense when the decision affects another function, exceeds someone's authority, or requires information they can't access, and then your job is to supply what's missing rather than automatically take over the whole problem. The baseline shifts with context too: hands-on solving may be essential early in a business, and even in a large company a new venture or an important technical problem may deserve your close attention, while a well-led, established area may need much less of you.

What if nobody on my team has time to own the work?

Willingness to let go won't create capacity, so delegation by itself can't fix a real shortage. Reduce commitments, change staffing, or keep an explicit temporary role yourself. And if problems keep showing up with no owner attached, that doesn't prove understaffing; it may reflect an unassigned responsibility, unclear authority, or a genuine capacity shortage.

Is the 80 percent rule a good way to decide when to let someone proceed with their own solution?

Keep the permission to accept a different approach, but don't use the percentage as the decision threshold. The rule says to let someone proceed when their solution gets most of the way to the goal and won't cause a catastrophe, yet eighty percent of what is never defined, your preferred method isn't the business requirement, and avoiding catastrophe is too low a bar. Instead, agree on the outcome, deadline and quality the work must meet, check that the owner has the skill, authority and time to make the call, and decide what downside is acceptable and how you'll detect trouble while there's still time to act.

How do I figure out why I keep taking over problems?

Keep a short Solver Log: reconstruct one recent week from your calendar and messages and, for each problem you personally fixed, record what it was, who owned it, what they needed from you, what you actually did, and why you intervened, including work you took over without being asked. Sort the why into a missing approval, a real risk, a gap in their knowledge, or an answer that wasn't the one you would have given, then check a few entries with the people involved, because your memory of a rescue may leave out the authority or information they had been asking you for. Treat the log as a practical exercise, not a validated diagnostic: it gives you questions to investigate, not a diagnosis of anyone's character.

How do I hand off a recurring problem to someone on my team?

Pick one recurring category where a capable owner has room to act, and agree on the decisions they'll make, the standard they'll meet, and when you'll review the work. Be explicit about what should bring you in sooner, tell the affected team who owns the call, and on the next few cases ask for the person's recommendation and the help they need before offering your own answer. When you feel the urge to take over, write down the specific requirement or risk that justifies it, record what you changed if you do intervene, and keep bad news welcome, because asking for help shouldn't automatically cost someone ownership.

Should I relax requirements to give someone a learning opportunity?

No, don't relax safety or legal requirements for a learning opportunity. Narrow the decision or add review before an irreversible step instead. Preferences are a different matter: if your head of support wants to clear a backlog by grouping cases differently than you would, and the plan fits the available staffing and meets the required customer response times, your preference alone isn't a reason to replace it, but if it puts a service commitment at risk, name that gap and work through it.

How do I know if the delegation is actually working?

Use the work to judge, because the urge to take everything back under pressure isn't proof that delegation is working or failing. At the agreed review, look at the work and your interventions, meaning whether the owner met the standard and whether your involvement addressed a material problem or mostly changed the method. If results fell short, examine the support, workload and capability before deciding what to change; if the owner keeps missing an explicit standard despite adequate support, address that performance problem directly; and if the work meets the standard and your main complaint is that you would have done it differently, leave it alone.

Keep reading

Work with KenMore insights