Houston, we have a problem

How often do you hear this sentence in your project?

The good news is that for most of us, it’s probably not as life-threatening as it was for the Apollo 13 crew. The bad news? It seems to happen every other day, instead of once in an entire movie.

So, you have a problem. Now what?

Sometimes, a cup of (virtual) coffee and thinking out loud can solve it.

But other times, you’ll need more than that.

That means it’s time for one of my favorite games in business. The game that has drawn me to project management and consulting in the first place.

I love cracking a good problem. And it’s never been more relevant either. In the age of rapidly rising AI, our ability to dissect an issue and leverage technology to solve the right challenge is a vital one to develop.

You don’t need to have all the answers - there’s an app for that. But you do need to assess the situation, break down the complexity, and ask the right questions.

There are a lot of frameworks for solving complex problems, but my favorite is one I learned on a stint with McKinsey. They’re paid millions to do this for clients, so it’ll probably be fine for your project too.

Warning: we’re going deep today. In some situations, this may seem like an overkill. But it’s a fascinating look into McKinsey’s kitchen, and I’m sure some of it will come in handy one day.

Ready?

Let’s break down their 7-step framework:

Step 1: define the problem

This seems obvious. Most skip it because of that. But if you don’t get this right, the rest of this framework is useless.

The goal of this step is to get 100% clear on “What do we need to know?”.

Zoom out and take the whole context into account. Try as best as you can to clearly describe the following:

• Describe the problem as specifically as possible

• Know what successfully solving your problem looks like

• Is 80/20 enough? How good does the answer need to be?

• Define any boundaries you have to solve it. Think scope, time, people, budget, etc.

To get to the root of the problem, keep asking “Why?” until you can’t get any deeper.

Step 2: Structure the problem

With your problem defined, it’s time to break it apart. Big goals, small steps - remember?

Now we need to identify paths to possible solutions.

You know the starting point (the problem), and you know what it would look like if it was solved.

The next step is to brainstorm solutions. Go crazy, and focus on volume. Nothing is off the table here. We’ll prioritize them later.

To get from A to B, you probably know more than you think. But there will also be gaps in your knowledge.

The goal here is to break a complex problem down into multiple, manageable steps.

You do that by:

1. Listing possible causes of the problem

2. Brainstorm solutions to (parts of) the problem

3. Find all the assumptions in your solutions (gaps)

You have now identified what you need to find out to solve the problem.

That already feels way more manageable, right?

Next up, structure your approach to finding that out with a hypothesis tree.

• Take the gaps you identified

• Form hypotheses of how to close the gaps

• Break those hypotheses down into questions to answer

Make sure nothing overlaps, while not missing anything.

Your hypothesis tree will look something like this:

And yes, if you have multiple solutions, you’ll have multiple (rough) hypothesis trees.

Step 3: Prioritize!

Next up: speed! You’ll never have time to research everything, so you’ll need to prioritize your efforts.

This is where one of our first questions comes in: how accurate does the answer have to be? A software experiment has lower consequences than landing a broken rocket, so don’t treat them the same.

Plot your possible solutions in a matrix to find out where to focus your energy.

People often ask why you do this after you’ve put effort into breaking down possible solutions. Wouldn’t it be easier to do this as you’re brainstorming and save you a lot of time?

Yes, but…

The better you understand the implications of possible solutions, the easier it is to put it in the “right” spot on the matrix. Trust me, it’s worth the effort.

With all your solutions on the matrix:

• Prioritize high ability to influence & high impact

• Consider something valuable & hard (top left)

• Try some quick wins (bottom right)

• Don’t waste time on bottom left

Know what to focus on? Ok, let’s move on!

Step 4: Plan

Now that you know which solutions to focus on, it’s time to make a plan.

Review the hypothesis tree for the prioritized solutions and add detail if needed. This is your starting point.

Next up, Identify research methods & sources for answering your questions one by one. Where can you find an answer? Experts, academia, internal data, and yes, AI - nothing is off the table here. What’s relevant and reasonable depends on your situation.

The goal here is to map how you’ll find the answers, and uncover how you and your team should spend your time during the analysis.

Sketch out a timeline and set clear expectations with everyone involved.

Before you move on, everyone involved should know exactly what to do, by when, how, and have access to the sources they need to answer their piece of the puzzle.

Step 5: Analyze!

Finally.

It’s time to roll up your sleeves and get busy!

It’s now time to gather data, check facts, and test ideas.

The hardest part in this phase is the balancing act between focusing and eliminating distracting new ideas, but also not being blind to new information that may disprove your initial direction.

The best way to do this? Diversity in the team, and a lot of conversations about what you’re finding. What supports the initial idea, and what disproves it?

What did you find that lured you in a different direction? And does the team think it’s worth pursuing or not?

The best process here is progressively going deeper. Start with summarized information (ChatGPT is GOLD for this!), and peel the onion layer by layer. Go deeper where required, but don’t pursue fun facts.

Focus on finding evidence, not on forming opinions.

Keep validating your data sources, get feedback on your ideas, and have conversations to cover blind spots.

Whenever you’re stuck here, remember this: you’re not trying to solve the problem. You’re looking for one fact at a time, that can help answer one question.

Many answers form a hypothesis, and hypotheses close a knowledge gap.

Brick by brick. That’s how you solve a complex problem!

Step 6: Synthesize

Now it’s time to gather what you’ve found and build up your solution to the problem. You do this by re-building the hypothesis tree from the ground up:

Gather your facts, answers, and data. Use them to build up arguments, and use arguments to make a recommendation for a solution.

Don’t fall for the trap of summarizing everything you’ve found here. You’re distilling, synthesizing, and building an answer based on many small bits.

For everything you add to your recommendation, ask yourself “So what?”.

What does this add? What are the implications of this piece of data? Does it support or disprove my hypothesis?

If you’ve decided earlier on to pursue more than one solution, this is also the time to prioritize those. You may have closed all the gaps from problem to solution for 3 different paths.

But you now know one costs $10K, while the other two cost millions.

Needless to say, this is the moment to kill some darlings and be realistic.

Got a clear answer?

Great, we’re almost there!

Step 7: Communicate

You’re almost home free. As a consultant, this step is often the most vital one.

No matter how good your solution is, if you can’t make your case it’s worthless.

But we’re not consulting here. We’re getting our project unstuck. So why communicate, and what?

As always: make your case by persuading with emotion, and convincing with facts.

To tell a (brief) story that links your conclusion & solution back to the initial problem. You back your solution up with arguments, and support that with your sources, facts, and data.

You can use the headlines of this to explain to your stakeholders why you made a certain decision, and the rest of the information goes (in summarized form) into your lessons learned.

If your organization ever faces a similar challenge, your documented path to a solution will be worth 10x the solution itself.

Wrapping up

Whew. That was a big one!

Overwhelmed? Three quick things to lighten the mood:

1. The full process is overkill for most situations. Take what’s useful.

2. You’re probably already doing much of this, but not structured like we did here.

3. Teams who are good at this can go through this entire process in hours, instead of “just trying to fix it” for weeks.

I asked last week what you’d like to read next. This topic won, and a close second was “Salvaging a derailed project”.

Guess what the topic for next week will be? 😉

If you want to learn even more, I highly recommend this book called “Bulletproof problem solving” by two former McKinsey partners.

And for now, that’ll do for this week.

While technically not a strict project management topic, I truly believe that this is a skill worth mastering. If you enjoyed this - let me know by replying! I love reading what you think of the newsletter and always reply personally.

Cheers,
Jasper