Your first customer is the narrowest group that has the problem often, knows it, has already tried to fix it, and can decide to buy alone. Describe them by situation, not age or gender. If you can't list 20 real people or businesses that match by name, the profile is still too vague.
This is for founders choosing who to build for first, before positioning, pricing or channels. The running example is CleanRoute, an imagined scheduling app for small cleaning companies.
In short:
- Situation predicts buying better than demographics do.
- Customers who already tried a fix are the easiest first buyers.
- Scoring three candidate groups shows the best first one quickly.
- A profile you can't name 20 people for is a guess.
Describe the situation, not the demographics
"Women aged 30 to 45" doesn't tell you who has the problem. "Owner of a cleaning company with 3 to 10 cleaners, scheduling them by WhatsApp every Sunday night" does.
The jobs-to-be-done idea makes the same point. Harvard Business Review's article on knowing your customers' jobs to be done argues that people "hire" products to make progress in a specific situation. Describe that situation and you've described your customer.
Use five filters for the first customer
Your first customers need to feel the problem strongly enough to try something new. Five filters capture that:
- Frequency. The problem happens weekly or more, not once a year.
- Awareness. They know it's a problem and can describe it.
- Past attempts. They tried a fix, even a spreadsheet.
- Budget. They already spend money or time on it.
- Authority. One person can decide to try and pay.
Steve Blank calls customers like this earlyvangelists, early adopters who are willing to take a risk on a startup's product. His post on the minimum feature set lists how to spot them, starting with "they have a problem".
Score three candidate groups
Write down three groups who could use the app, then rate each one on the filters. For CleanRoute, small firms beat both solo cleaners and large agencies.
| Group | Frequency and pain | Tried a fix | Can decide alone |
|---|---|---|---|
| Solo cleaners | Low, own diary | Rarely | Yes |
| Firms with 3 to 10 staff | High, weekly rota | Spreadsheets, WhatsApp | Yes, the owner |
| Agencies with 50 staff | High | Already use software | No, a manager team |
Solo cleaners don't have a scheduling problem. Large agencies have it, but already bought a tool and need several people to switch. Small firms have the pain, a homemade fix, and one decision-maker.
These scores come from the imagined example. Base yours on interviews, using the questions in our customer interview script.
Write the profile as one line
Combine the winning group's details into one sentence you can repeat:
Owners of cleaning companies with 3 to 10 cleaners, in one city, who plan the week's jobs by spreadsheet or WhatsApp and lose hours fixing clashes.
Each part is a filter you can check. "3 to 10 cleaners" rules out solo cleaners and agencies. "Plan by spreadsheet" means they've tried a fix. "Lose hours" is the pain in their words.
Test the profile by naming 20 people
A profile is only useful if you can find the people in it. Try to list 20 real businesses or people that match, with a way to contact each.
For CleanRoute, that's a search of Google Maps for cleaning companies in one city, then a look at each website for team size. Twenty names take about an hour.
If you get stuck at five, the profile is too narrow or the group is hard to reach. If you can't tell whether a name fits, the profile is too vague. Either way, rewrite it.
Find where the profile gathers
Once you have names, look for the places they gather. Those places become your first channels.
- Directories and maps list local businesses by type and area.
- Trade associations often publish member lists or run events.
- Online groups on Facebook, Reddit and forums where owners ask for advice.
- Audience tools such as SparkToro show what a group reads, watches and follows.
For professionals, Boolean search on LinkedIn combines job titles, industries and places in one query. Our guide to choosing your first marketing channel shows how to test these places.
Update the profile after real customers arrive
Your first profile is a hypothesis. After 10 to 20 users, look at who actually completed the core action and came back.
You may find that firms with 5 to 8 cleaners stick and firms with 3 don't. Narrow the profile to match. Real behaviour beats your first guess.
Give GrowApps the profile
When you start a new app, GrowApps asks about your relationship to the idea, how the problem is handled today, and what a win would be, then researches apps like yours before designing screens (how GrowApps works). Paste your one-line profile into the first message. The research and screens then aim at that customer, not at everyone.
Name your first five now
Write your one-line profile. Then open a map, a directory or your contacts and write five names that match it, with how you'd reach each one. If five is hard, change one filter and try again. Next, turn the profile into a message with our positioning guide for small apps.
Frequently asked questions
Can I have more than one target customer?
Later, yes. At the start, pick one. Two profiles split your messaging, your features and your channels, and make it harder to tell what's working.
What if my app is for consumers, not businesses?
The same filters apply. Replace "can decide alone" with "can install and pay without asking anyone". Then describe the situation, such as "parents of a child with a food allergy who plan meals every Sunday".
Sources and further reading
- Know your customers' jobs to be done, Harvard Business Review
- Perfection by subtraction: the minimum feature set, Steve Blank
- Customer development articles, Steve Blank
- SparkToro, audience research
- Use Boolean search on LinkedIn, LinkedIn Help
- How to talk to users, Y Combinator
- The Mom Test, Rob Fitzpatrick