Customer Focus
Purpose
Trying to solve too many different problems for too many different people at the same time dooms you to solving none of them well. The following exercises will help you generate a weighted list of market segments that you will then run Message-Market Fit sprints with.
The outcome will be deep and shared conviction in who you’re building for, why, and what they need most right now.
📌 If you’re running this set of exercises for a brand new, pre-product business idea, this will be a very iterative process. You won’t know much, if anything, about the segments you initially come up with. That’s okay! You just need a rough list to get you started listening to the market; once you're having those conversations, you’ll start to have a much clearer idea of who you should be talking to.
Not convinced that narrowing your customer focus is a good idea? Read about how the founder of ConvertKit righted the ship after failing to focus on a market niche early on »
Exercise 1: Market Segmentation
Estimated time needed: 90 minutes
Materials: Text editor, whiteboard app, pencil and sheet of paper – whatever list-making tool each person likes best.
⚠️ Each participant does this exercise alone first.
STEP 1: List literally anybody you can think of that might want what you're building. There are no wrong answers or wrong sizes here! List individual people’s names, persona descriptions, job titles, industries, anything. (See examples below)
STEP 2: Share your lists with the full team. This could just be a shared board that folks paste items into, or a conversation. The aim here is to spark new ideas for each other.
STEP 3: Another round of individual brainstorming, then coming back together to share again.
STEP 4: Combine everyone’s lists into themed batches and give these batches nicknames.
📌 You’re still NOT winnowing here, you’re just organizing. You will have segments of various sizes and shapes and that’s okay for now. You’ll also think of new segments as you’re doing this – that’s great! Add them.
Here are some sample lists:
Dog-Walking App
- Professional dog walkers
- Dog owners living in dense cities
- My friend Susan who has three dogs
- Dalmatian owners
- Dog rescue orgs
- Pet care places
- Dog Hotel Austin
- Dog groomers
- Pet stores
- Dog breeders
- Spaniel owners
Data Analytics Platform
- Product managers at prosumer software companies
- Roman Zubenko (PM at Notion)
- Product managers at enterprise software companies
- Data scientists at CPG companies
- Head of Data Engineering at Harry’s
- CTOs at small startups
- Senior Digital Marketing Manager at medium-sized startups
- Head of Product at software startup
Exercise 2: Choose, Champion, & Characterize
Estimated time needed: 90 minutes
STEP 1: Ask participants to think about the list generated in Exercise 1 and then pick their first choice of a segment to focus on.
📌 Some people will struggle with this assignment! A frame that might help is imagining that the entire rest of the team is on the moon for six months and YOU ALONE must pick a segment based on what you know right now.
STEP 2: Go around and share your picks. Detail why you chose them.
While doing this, take note of:
- The set of dynamics that informed why people chose as they did. This is the beginning of your list of important characteristics, which will turn into columns in the scoring spreadsheet below.
- Any unverified assumptions the team has about each segment and the people in them. You can add these to your Assumptions Table and create a plan to test and validate or invalidate them.
🎛️ There are some standard characteristics that often apply to market segments:
- How acute is the pain point in this segment?
- How well do they perceive that they have this pain point? Are they trying to solve it already?
- How proximate is our team to this segment? I.e. are we part of this community? Adjacent to it? Very distant?
- Is there a regularly occurring situation for this buyer that would organically make them think to use our product? (Trigger)
- How difficult is it to reach these folks? Are they online? Are there trade journals everyone reads?
- How long does it take to close a deal with them? (Length of sales cycle)
- Do buying decisions require group consensus or can one person make the decision?
🦄 …and there are characteristics that are unique to your context.
For example, for our dog-walking app, there may be segments focused on dogs that don’t need a lot of exercise vs. segments focused on dogs that do require a lot of exercise. The latter would be better customers for a dog-walking app!
For the data analytics platform, it might be important to note how mature the data tooling is for each segment. Folks can’t analyze data they don’t yet have access to!
Exercise 3: Score, Sort, & Select
Estimated time needed: Multiple 90 minute sessions (see why below!)
Materials: Spreadsheet software with column sorting functionality.
STEP 1: Set up a spreadsheet where each row is a segment and each column is a characteristic. For each segment, create a nickname that the team understands.
STEP 2: Create a numeric scoring rubric for each characteristic. E.g. for “how acute is the pain point”, this could be 2 = very, 1 = somewhat, 0 = not at all. If you find the team wanting to score on the level of decimals, consider just expanding the range to be 0-5 or 0-10.
📌 It’s important to make sure each characteristic uses the same directional scale – i.e. higher should always indicate a better fit and lower should always mean a worse fit.
STEP 3: As a group, score each segment on each characteristic.
⚠️ This is a very important activity to do as a group! The discussion will bring a lot of misalignment to light. This activity may take a long time depending on how long your list of segments is. Don’t lose faith! It is time well spent.
📌 Keep noting your untested assumptions as you discuss. This is where many of them will come to light.
STEP 4: Create a new column that sums the scores, then sort the sheet. It may be surprising or unintuitive which segment tops the list. Discuss it as a group! Why is this surprising? Is there a characteristic you didn’t identify yet that is making the result surprising? Add it, and re-score.
STEP 5: Keep doing this until everyone is on the same page about the best/next segment to target. Again, this may feel contentious and tedious, but it’s very important work!
🏆 The result here will be a stack-ranked list of the most promising target market segments and personas. You’ll use this list to conduct sequential market research sprints, starting with your most promising and moving on to the next if you learn things that disqualify it.
Example of a market segment matrix:
Some tips and tricks
- You will come up with a LOT of ideas - it's not reasonable to work through all of them. Focus on a bite-sized set
- Segments are not the same as "triggers"
- A segment is a trait you can know about someone from the outside that can correlate with the likelihood of using your product
- A trigger is an event that may cause someone to be likely to use your product
- For example: "Dalmatian owners" are a segment. "People who adopted a dalmation in the last year" is a triggering event
- Focus on segments because it's much easier to find people who have certain traits than people who experienced certain events
- You should use different segmentation spreadsheets for different business models. You'll probably want to use different characteristics and pain points for different business models
- You'll notice that some characteristics don't apply to a segment - give them a default score of "1"
- You'll notice that sometimes you just don't know enough about a segment to understand how to score a characteristc. Give them a default score of "2". Then write down an assumption or a question about them, so you can research it.
- Founders won’t be interested in all the segments you come up with, even if they rank highly. Find out why! There may be relevant assumptions or data there.
- These segments and their rankings are ALL based on your assumptions. We need to test them. But it's only worth doing the research if they rank high, or if we learn new info that causes them rank higher
- Often a weird segment pops up to the top. That could indicate either:
- we have a dynamic we don’t care about
- we’re missing a dynamic