Choose backward. Look at the clients who went well, name what they had in common, and turn that into a short list of markers you can recognize on a first call. Skip the demographics. Then add the one requirement I never let a client drop: my ideal client can pay my fee.
List the clients you'd take again tomorrow, and write down what those clients had in common.
Your own past work holds better evidence about fit than any amount of market research does.
Write two lists this week: the markers of a great client, and the markers of a wrong one.
You would know an ideal client if you saw one.
Leah Neaderthal, Smart Gets Paid
Start with the clients you already had, because a mixed history is evidence you can use. Go through the last two or three years and sort every engagement into work you'd take again tomorrow and work you wouldn't. The pattern in the first pile is your starting point.
That's the route I took myself, and it took three passes:
Nobody handed me that on a worksheet. Each version came from working with the wrong people for a while and paying attention (I walk through the whole sequence in my Smart Gets Paid episode "How I figured out my niche"). Your mixed client history is the same raw material, and you've already done the expensive part.
Look at where your client is now and where they want to be, which tells you more than any demographic profile does. Demographics describe a person. Situations tell you what they'll say yes to.
Matt Lerner, formerly of PayPal, made this point well. He spent a million dollars on a segmentation study that produced personas scored for risk tolerance and price sensitivity, then realized none of it would help his marketing.
His contrast is the useful part. Knowing that "Tammy is a 39 y/o divorced mother of 2, from Atlanta, she works in compliance, earns $58K/yr" tells you nothing you can act on. Knowing that "Tammy is on the sweltering hot beach with two kids, they're getting hungry, and one of them has to pee" tells you to sell air conditioning and a kids' menu.
So I have my clients answer four questions about the person they want to work with:
| The question | When it lives |
|---|---|
| What are their frustrations? | Now. The problems happening today |
| What are their fears? | Future. What they're afraid of if this stays unsolved |
| What are their wants? | Now. What they want in place |
| What are their aspirations? | Future. What they hope happens once they get it |
Answer those four and you can write messaging, build offers, and price the work, because you know what your client worries about at eleven at night.
You'll know because you'll have written down the markers in advance, which is the whole point of the exercise. A marker is a condition that has to be true for the work to succeed, and it's usually something concrete you could point at.
Notice that neither marker is an industry or a headcount. Both describe what has to be in place for the work to land, which is also the answer when a client asks for industry experience you don't have.
Write two lists. The markers of a client you can do great work for, and the markers that tell you this one is wrong for you. My own second list includes people selling digital products or group programs, anyone too early to know they need a strategy, and anyone looking for a silver bullet who isn't willing to do the work. The second list will save you more time than the first one.
You will choose wrong, at least partly, and the choice is built to be revised. I tell my clients we're putting a stake in the ground so the right people can find us, and a stake is something you can pull up and move.
Every consultant with a clear focus arrived there by being wrong in public for a while. One of my clients did great communications work for nonprofits and eventually noticed she loved working with Jewish nonprofits most, because that work was close to her heart. She said so, changed her profile, and became the go-to marketing person for them.
Another worked with companies of every size and realized her best work happened inside large, complex firms, so that's who she speaks to now.
Neither of them knew that on day one. They found out by choosing, working, and watching what happened. This is why I call it choosing your future instead of figuring out your niche: you're deciding who you want to help most, and then you go and get seen by them.
One thing you should not do is leave the choice open while you think about it. An unmade choice keeps you taking what happens to come your way, which is the situation you're trying to leave.
Most of the women I work with come from marketing, communications, or another field that runs on data, so their instinct is to research this. They want a segmentation. They want to be sure before they commit.
I understand it, and it's the wrong tool. Your ideal client isn't waiting in a dataset to be discovered. You choose her, the same way you choose which conference to speak at or which service to stop offering. Up to now you've mostly been responding to whoever came your way. This is the moment you get to say how you want things to be.
Two things make the choice easier than it looks. It's version one, so being wrong costs you a revision rather than a year. And one criterion settles a surprising number of hard cases, which is that your ideal client can pay your fee. Not that they need you badly. That they can pay.
Then write it down as a picture rather than a category. They look like this. They'll say things like this. They'll be in this specific moment. And I'll know they're wrong for me if I hear these other things instead. That description is what makes everything downstream possible, including what you charge.
In my method they're the same move, so don't spend time separating them. A niche is usually described as a market and an ideal client as a person, and both come out of the same decision about who you want to help most. Choose the client, in detail, and you have described the niche at the same time.
Specific enough that someone reading it knows immediately whether it means them. "Nonprofits" fails that test, because a two-person foundation and a national organization share almost nothing. "Nonprofits with a development team of three or more who have just lost a major funder" passes it. Aim for the level where a wrong-fit client rules herself out.
Both, and they overlap more than people expect. Start from the clients who lit you up, because you do better work for them and better work sells itself. Then apply the fee criterion, which removes the candidates you couldn't build a business on. What's left is the honest intersection, and it's rarely as small as women fear.
No, and waiting for interviews is a common way to stall for six months. You've had conversations for years and you remember what your clients complained about. Take an educated guess, write it down, and let real conversations correct it. If you truly have no history to draw on, go where those people talk to each other and read how they describe the problem.