Chapter 7: Choosing Your Customers

Core Idea

Startups don’t starve, they drown in too many options. If your segment is vague, you’re really having one conversation each with twenty different types of customer, so the feedback contradicts itself. Slice the segment until you have a specific who-where pair you can go and find.

Frameworks Introduced

• Segmentation:
  • Big companies started narrow. Google helped PhD students find obscure bits of code. PayPal helped collectors trade Pez dispensers and Beanie Babies. Evernote helped moms save and share recipes.
  • A fuzzy segment causes three problems:
    1. You get overwhelmed by options and don’t know where to start.
    2. You aren’t moving forward but can’t prove yourself wrong.
    3. You receive mixed feedback and can’t make sense of it.
  • Remind yourself that you’ll reach the whole world eventually, but you have to start somewhere specific.
• Customer Slicing:
  • When to use: when your segment is generic or your feedback is inconsistent.
  • How: start with any broad segment and ask:
    1. Within this group, which type of person would want it most?
    2. Would everyone within this group buy/use it, or only some?
    3. Why does that sub-set want it? (e.g. what is their specific problem)
    4. Does everyone in the group have that motivation or only some?
    5. What additional motivations are there?
    6. Which other types of people have these motivations?
  • You’ll get two lists: demographic groups and motivations. Keep slicing any that are still generic.
  • Stop when you know a physical or digital place where you can find them. If you can’t name one, the segment is still too broad.
• Choosing which slice to start with: from your who-where pairs, pick the one that seems most:
  • Profitable or big
  • Easy to reach
  • Personally rewarding
  • Spend minutes on this, not weeks. The author stresses the third criterion: pick customers you admire and enjoy being around, because the work is a grind otherwise.
• Talking to the wrong people happens in three ways:
  1. Your segment is too broad and you’re talking to everyone.
  2. You have multiple customer segments and missed some of them.
  3. You’re selling to businesses with a complicated buying process and overlooked some of the stakeholders.
  • Hidden segments: a kids’ app must satisfy kids and parents. A product for public schools involves teachers, students, administrators, and possibly the parent-teacher association and taxpayers. Partners for manufacturing, distribution, or promotion count too. A multi-sided marketplace needs separate conversations with each side.
  • Talk to people who are representative of your customers, not the most senior or impressive ones. The author talked to advertising executives but never to the kids who were supposed to love the product.

Key Concepts

• Customer segment: a specific group of customers with shared problems and goals. It is useful only if it helps you make decisions and find people.
• Who-where pair: a segment defined together with the place you can find it.
• Industry taxonomy: an expert’s map of the types of customer in an industry. It’s a good starting point for slicing.
• Demographic trap: a segment such as “sales organisations with 25–250 salespeople” still hides very different needs, workflows, and goals.

Mental Models

• Before you can serve everyone, you have to serve someone.
• When feedback is all over the map, suspect the segment before you suspect the questions.
• If everything “sort of works” and you can’t cut any feature, you’re serving too many segments.
• Ask partners for a commitment that tests them, such as shelf space.

Anti-patterns

• “Our market is everyone who advertises”: a market that feels infinite gives you feedback you can’t use.
• Demographic-only segments: “finance professionals, 25–35, in a major city” doesn’t tell you what to build or where to go.
• Planning segments forever: pick a concrete starting slice in a few minutes and go talk to people.

Rules of Thumb (author’s wording)

• If you aren’t finding consistent problems and goals, you don’t have a specific enough customer segment.
• Good customer segments are a who-where pair. If you don’t know where to go to find your customers, keep slicing your segment into smaller pieces until you do.

Worked Example

Slicing a high-end fitness gadget for busy professionals.

1. First guess: it’s expensive, so customers need a high income (finance professionals). It’s digital, so they’re young (25–35). Finance professionals live in big cities. Result: “finance professionals, age 25–35, in a major city.” This is worthless, because it doesn’t help you decide anything or find anyone.
2. Slice by who wants it most: “finance professionals in London who are training for a marathon.” They clearly take fitness seriously, so they’re more likely early adopters.
3. Slice again: “…who go to the gym during their lunch hour.”
4. Who-where pair found: you can have all the conversations you want for the price of a gym membership in London’s financial district.

The powdered condiment. A founder made a sweet, nutritious, all-natural powder with many possible uses. Bodybuilders wanted one thing, restaurants another, and moms a third. Pleasing one group disappointed the others, and she couldn’t even choose a label colour. She focused on moms with young kids who already shop at health food stores, which also told her where to find them. Because the stores were a key partner, she asked small independent stores for a commitment: a little shelf space beside the breakfast foods. She planned to return a week later to ask how it went, and if it sold, to reach shoppers through in-store tastings.

“Students” and “sales organisations.” Five early users of a “student” app wanted formal citations, practice questions, iPad support, 80 collaborators on one shared computer, and offline use over a weak connection. They turned out to be a PhD student, a prep-school student, a homeschooling parent, a rural village sharing one computer, and a student on an unreliable mobile connection. A pair of founders selling to “sales organisations” had the same problem: twenty conversations gave twenty different must-have features.

The author’s “advertisers” segment (Habit). They talked to small shops, online retailers, big brands, agencies, music labels, and more. Some talked of paying $10,000 a month while others laughed at $10. Every feature was moderately popular and none could be removed. They finally noticed unusually strong signals from creative agencies that wanted to be edgy, ignored everyone else, cut features, and could at last tell what worked.

Key Takeaways

1. Start specific, and broaden later.
2. Keep slicing with “who wants it most, and why?” until you know where to find them.
3. Choose the first slice by size or profit, reachability, and whether you enjoy the people.
4. Map every party who must say yes: users, payers, partners, and stakeholders inside a business.
5. Talk to representative customers, not impressive ones.
6. Treat contradictory feedback as a sign that you’re mixing segments.

Connects To

• Ch 5: shelf space is a commitment that separates serious retailers from polite ones.
• Ch 6: more than 10 conversations with scattered results is the trigger to slice.
• Ch 8: the first step of the process is choosing a focused, findable segment.
• Brant Cooper: the author credits Cooper’s work on segmentation in the acknowledgments.