I actually like how Dan Brown describes discovery in his book Practical Design Discovery:

Discovery is a mindset, not a phase. Oh, it would be nice to compartmentalize learning in a single phase. Time-boxing your efforts to acquire knowledge makes them predictable and cost effective. But learning doesn’t work on a timetable… the truth is we just don’t know when the creative breakthrough will come, or when we’ll have to answer more questions.

Here’s what I don't like: an agency acting like they have no clue how to solve a problem. Then getting the client to pay them to learn so they can then charge more to actually start.

I saw it again this week. I was coaching an agency owner about an audit—often just another name for a discovery project—they were doing for a client. I asked what the audit was for.

They said, “To lower the risk on the project.”

I asked, “Whose risk?”

They said, “Mine.”

This isn’t always the case. But I’ve seen it happen way more than the reverse. Agencies often ask clients to cover their own risk. Then they’re surprised when clients are hesitant to pay for their “help.”

I’m not saying we always overplay our hands. Honestly, I think the bigger problem is that we underplay it.

Saying you have no idea what will work without a discovery phase isn’t true. If you’re an expert, you have guesses. And guesses are enough to start. A project isn’t one big burst of work with a sure outcome. A good project is a process. You build something, measure it, learn from it, and make it better over time.

Teresa Torres says it well in her book Continuous Discovery Habits:

Discovery isn’t a one-time activity. A digital product is never done. As we learn more about our market, as our customers’ needs change, as new technology becomes available, good products adapt.

If you’re an expert, a discovery project shouldn’t be step one. A guess should be step one. Then the project follows, and the project is really just discovery that never stops. My wife’s doctor started with smart guesses, based on years of experience. Then came the tests to validate or invalidate. Then the prescription.

What if you’re not an expert?

If you’re not an expert, your guesses aren’t worth as much. That’s why a discovery project feels like the right first step for a beginner. Their guesses are random, not informed.

I teach agency owners how to make more money by pricing their work based on value. To figure out what’s valuable to a client, you also have to figure out what isn’t. And there’s almost no value in paying a beginner to go learn something for you. I wouldn’t pay my neighbor who’s a carpenter to run a discovery project on how to do heart surgery on me. No thanks.

So what should a beginner do instead?

Don’t ask a client to cover your risk by paying for a discovery project.

Yes, I do believe discovery should be paid. But not by clients. By you.

Find someone who’s exactly the kind of client you want. Offer to do the work for free. Or go further: pay them a small fee to let you do it.

Learn everything you can about solving that one problem for that one kind of client. Then find another client with the same problem and do it again. Then one more.

By the third project, your guesses are actually worth something. When the fourth client comes to you, you might spot the problem after just a few questions; no need to ask them to pay you or to spend weeks of time. That builds their trust fast. That makes them more likely to hire you. Which makes the fifth client easier to land. And the sixth. And the seventh. And the twenty-first. Every project sharpens your guesses, because every project gives you more to learn from.

I believe this is the true spirit of discovery. You’re not learning just enough to do a project. You’re learning the ins, outs, and edges of your niche and your market. That kind of learning never stops, and it makes your work incredibly valuable to the people you serve.

If you’re already an expert, don’t be afraid to be confident in your experience. Evidence builds confidence. I’m not suggesting that you treat your hunches as validated test results, but don’t disregard them as worthless either. I see too many people who do the latter, and I’m sad for them and for their clients.

My agency specialized in design systems, which means we talked regularly and worked with thousands of people in that space over many years. With just a few questions over a few minutes in early conversations, we had reasonable guesses about where our clients’ problems lay that turned out to be right the majority of the time. We never assumed or acted like our guesses were enough that we didn’t have to test them, but there were enough to start. We never did upfront discovery projects. We made guesses and started projects, because all projects are discovery projects, from start to finish and beyond.

Our clients would often say that just the questions we asked already helped them, that they understood their own problem differently by the end of the call. They hadn’t even paid us yet. I’m very proud of the fact that our sales process was just as valuable as other agencies’ paid discovery projects, if not more.

I think it’s something that anyone who’s spent enough time in a given space can do this too. But only if they’re confident enough to assign value to their guesses.

There’s a career’s worth of discovering in solving one problem for one kind of person.

Long live discovery.