We all mess up. Projects fail. Campaigns flop. Things that were supposed to work just...don't.

That's fine. Seriously. What's not fine is when we mess up the same way twice because we never figured out what actually went wrong.

That's where the 5 Whys comes in. It's one of the most underrated tools in our toolkit, and we need to get better at using it.

Here's the concept



Most failures aren't random. They're the end of a chain of cause-and-effect relationships that started somewhere upstream, probably in a place we weren't looking. The 5 Whys is dead simple: we take a problem, ask "why did this happen," then ask "why" again. And again. Five times total.

Toyota invented this. Not a consulting firm. Not an MBA program. A car manufacturer that needed to figure out why machines kept breaking. They asked why five times and kept finding the answer hiding three or four layers deeper than anyone expected.

We're going to do the same thing.

When to use it

This works best for problems that are within our control and moderate in complexity. Think: the campaign that underperformed, the launch that got delayed, the feature that nobody used.

Use it when:

• Something went wrong and we own the outcome
• The problem is specific enough that our team can reason about it
• We need to stop treating symptoms and start treating causes

Don't use it when:

• The fix is already obvious (just fix it and move on)
• The problem is massive and systemic, touching dozens of teams — the 5 Whys can be a starting point here, but we'll need a much broader investigation
• The root cause sits entirely outside our team's knowledge or control —no amount of brainstorming will uncover something we fundamentally can't see

How we run it

This isn't complicated. If you've ever run a retro or a brainstorm, you can run a 5 Whys. All we need is a whiteboard (or a shared doc if we're remote) and about 45 minutes.

Step 1: State the problem

Write down one clear problem statement. Not three. Not a vague feeling. One statement.

Example: "Our fundraising campaign missed its goal by 30%."

Step 2: Set up the space

Divide a whiteboard (or doc) into five columns. The first column gets the problem statement, phrased as: **"Why did [problem] happen?"**

Step 3: Set the tone

This matters more than you think. Before we start, say this out loud: *We're here to find the cause, not to find someone to blame.* Everyone needs to understand this is collaborative. There are no wrong answers. We're investigating, not prosecuting.

Step 4: Brainstorm the first "why"

Give the team five minutes. Everyone throws out contributing factors. Don't filter. Don't debate. Just collect.

Example answers for why the campaign missed its goal:

• Click-through rate was too low
• Digital leads didn't convert
• The campaign wasn't engaging enough for our audience

Step 5: Pick one thread and pull

Vote on which answer is most worth investigating. Then rephrase it as the new "why" question.

Example: "Why wasn't the campaign engaging for our audience?"

Step 6: Repeat three more times

Same process. Ask why. Brainstorm. Vote. Rephrase. We do this until we've asked "why" a total of five times. By the fifth round, we should be staring at something that looks a lot like a root cause.

Example of how the chain might go:

1. Why did the campaign miss its goal? → The campaign wasn't engaging.
2. Why wasn't it engaging? → The copy didn't connect emotionally.
3. Why didn't the copy connect? → We wrote it based on assumptions, not research.
4. Why were we relying on assumptions? → We skipped the audience research phase.
5. Why did we skip audience research? → We committed to a launch date before scoping the work.

That fifth answer is the real problem. We over-committed on timeline before understanding what the work actually required. That's fixable. That's systemic. That's the kind of insight that changes how we operate going forward.

Step 7: Decide what we're going to do about it

Don't leave the room without assigning action. Pick one or two solutions that address the root cause. Assign each to a specific person with a specific check-in date.

Example: "From now on, we scope audience research into every campaign timeline before committing to a launch date. Sarah owns updating our campaign planning template. Check-in: two weeks."

Tips that actually matter

• Run it backwards to stress-test. Start from your root cause and use "therefore" to walk back up the chain. If the logic breaks anywhere, you haven't gone deep enough.
• Never land on "someone made an error." Human error is almost never the root cause — it's a symptom. If someone made a mistake, ask why the mistake was possible. Unclear process, unrealistic workload, no review step. Keep pulling.
• Go past five if you need to. Five is a guideline, not a law. If your fifth answer still feels like a symptom, keep asking. Stop when you've hit something structural you can actually change.
• Use breakout groups for large teams. More than eight people in the room? Split into smaller groups after the first round, let each follow a different thread, then reconvene and compare root causes. One problem often has several contributing roots.productive and makes sure nobody walks out feeling blamed.

The real point

The real value of the 5 Whys isn't any single root cause you find. It's building a team that asks "what happened in the system" instead of "whose fault was this." That shift — from blame to curiosity — is what separates teams that repeat mistakes from teams that compound their learning. Every problem we dissect properly makes us harder to break next time.

If you work anywhere in financial services pricing data, you know how much complexity hides beneath the surface — across fixed income, derivatives, structured products, treasuries, whatever your coverage universe looks like. Every feed, every source mapping, every transformation step is a place where upstream assumptions can quietly go wrong. The 5 Whys gives you a way to stop chasing the same categories of errors and start eliminating them.

Of course, the even better version of this is starting with reference and pricing data that's already been through that level of scrutiny — where someone else has already traced the cause-and-effect chains so you don't have to. That's what we build at SQX. But whether you're using our data or maintaining your own, the principle is the same: stop fixing symptoms. Find the root cause. Fix it once.