Choosing the Right Customer & Differentiation: Field Guide

A practical reference for picking who you're building for and why they should care

How to Use This Guide

This is organized around the moments where student projects actually stall out on the customer question. Open it when you're stuck picking a segment, when interviews are giving you mush, when your idea looks like five other ideas, or when you've been "researching" for two weeks and still haven't named a single real person.

Each section has a few principles, a short explanation, and a source. None of them are theoretical. Most of them come from my own Bus Buddy / Route Tracker train wreck and the classroom conversations we had on lecture day. If something here saves you from repeating a mistake I already made, this guide did its job.

When You Might Open This Guide

• You have four possible customer segments and every one of them sounds reasonable, so you keep researching instead of picking.
• You're about to do customer interviews and you're worried you're going to lead the witness and get back whatever you want to hear.
• A friend said they "love" your idea and you're trying to figure out if that actually means anything.
• You're looking at your product next to the competition and they all look the same on paper.
• You're the smallest team in your market and you're trying to figure out where you can actually win.
• You've been researching customers for two weeks, you haven't talked to a real human yet, and your midterm is in a week.

When You Can't Stop Researching and Just Pick a Customer

There is no right customer. There is only a first customer. Students freeze on this question because they think one of the segments is secretly correct and the others are secretly wrong, so picking feels like guessing on a test. It isn't. You're not trying to find the ideal market. You're trying to find a specific human with a painful problem who will talk to you this week. That's the whole bar. Lower the psychological bar and the paralysis goes away.

Source: Jason, Customer Selection Anxiety working notes

A demographic isn't a customer. "Small business owners" is not a customer. "Students who ride the bus" is not a customer. If you can't name a specific human in a specific moment, you don't have a customer yet. You have the idea of a customer, and that's not the same thing. The idea of a customer can be talked to by AI. A real customer can be talked to by you, and only one of those actually tells you anything.

Source: Classroom recording, Getting the Right Customer and Differentiation

Tuesday at 2:17 PM. When you sit down to write out your customer, pick a day of the week and a time. Your customer is frustrated. What just happened? Three sentences, no vague language. If you can't do that exercise, you don't know your customer well enough yet, and no amount of additional research is going to fix it. The exercise fixes it, because the exercise is the research.

Source: Classroom recording, Getting the Right Customer and Differentiation

Pick, commit, iterate. Don't pick, stall, pick, stall. The goal of the first customer pick is not to be right. It's to get you into contact with a real person fast enough that the real person can correct you. Initial alignment isn't real alignment. It's directional alignment, and it's only valuable because you can iterate on it once the world starts talking back.

Source: Classroom recording, Getting the Right Customer and Differentiation

When You're Running Customer Interviews

Your pain is not the market's pain. I built Bus Buddy because I hated being late for the bus. It turns out most bus riders don't have my problem, because most bus riders actually show up on time. I had three downloads on that first release and one of them was my parents. Your own pain is a great starting point for a product idea. It's a terrible finishing point. You are a sample size of one, and a sample size of one will ship you a product nobody wants.

Source: Classroom recording, Getting the Right Customer and Differentiation (Bus Buddy story)

Don't ask what they want. Ask what just went wrong. When I went back and actually did customer research, I didn't stand around asking people what features they wanted in a bus app. I asked them what the most painful thing about riding the bus was today. People are bad at predicting what they'll use. They're pretty good at describing what they just suffered through. Anchor your questions in concrete recent moments, not hypothetical preferences.

Source: Classroom recording (UVU bus stop interviews, Route Tracker phase)

AI market research gives you a coherent story that breaks. Humans don't optimize for accuracy, they optimize for coherence, and AI is the best coherence engine that has ever been invented. If you ask an LLM to describe your customer segment, you will get a beautiful, plausible story that falls apart the second a real human opens their mouth. Use AI to prep, use AI to draft, use AI to sanity check. Do not use AI as a substitute for a 20 minute conversation with one actual person.

Source: Classroom recording, Getting the Right Customer and Differentiation

Proxy users count. If your real customer is hard to reach in the next seven days, find someone who used to be that customer, or who sits one seat away from them. A recent ex-user of the problem is almost as valuable as a current one, and they're usually easier to get on a call. Warm network, cold outreach with one simple ask, and proxy users. Those are your three lanes. Draft the outreach message before you close this guide.

Source: Jason, Customer Selection Anxiety working notes

When You're the Small Player in a Crowded Market

Pick your terrain. Don't pick your enemy. The interesting thing about Sun Tzu is that he spends most of The Art of War not fighting. He's picking where to go, picking the terrain, figuring out where the enemy is weakest, then engaging there. Sometimes he wins without fighting at all. That's differentiation. You are not going to out-feature the 800 pound gorilla in your space. You are going to find the one spot where their strength is actually a weakness, and you're going to stand there.

Source: Classroom recording, Getting the Right Customer and Differentiation

Underdogs win by attacking where the giant is weak, not where the giant is strong. The same qualities that make a big competitor strong are usually the qualities that make them slow, expensive, or committed to serving a different customer. Gladwell's whole point in David & Goliath is that underdogs who play the giant's game lose, and underdogs who refuse to play the giant's game often win. Your job isn't to be a smaller version of the incumbent. Your job is to be a different shape.

Source: Malcolm Gladwell, David & Goliath

The things that give the giant strength also give them weakness. A big competitor has distribution, brand, and a customer base. Those same things lock them into serving that customer base the way they already serve them. That's your opening. Find the customer they can't afford to care about, or the workflow they can't afford to break for, and you'll own that ground before they even notice you're there.

Source: Malcolm Gladwell, David & Goliath

Identify the 800 pound gorilla, then go somewhere else. Before you differentiate, name the biggest single competitor in your space. Not the list of ten. The one. Understand what they do, why it's good enough, and where it fails. That failure mode is where you live. Everything else is noise. And the failure isn't usually a missing feature; it's a customer they don't really care about or a friction they're too big to remove.

Source: Classroom recording, Getting the Right Customer and Differentiation

When Your Differentiation Looks Like Everyone Else's

Differentiation is removing a friction others ignore. Not branding. Not positioning fluff. Not "we're the AI-powered one." Write down the current solution, what sucks about it, why people tolerate it anyway. Now ask: what friction are we eliminating that everyone else has decided is just part of the job? If you can't answer that in one sentence, you don't have differentiation yet. You have a product.

Source: Jason, Customer Selection Anxiety working notes

Pick two dimensions and find the empty corner. Draw a 2x2 with your two best differentiators on the axes. Put yourself on it. Put every competitor you can think of on it. You want to be the only dot in the top right. If a competitor is already up there, that dimension isn't your differentiator, no matter how much you wish it was. Move on and try two different dimensions. The L-shape on the bottom left is where the losers live.

Source: Classroom recording (2x2 differentiation exercise)

Your differentiation is defined relative to a specific customer's pain. If the customer changes, the differentiation changes. "We're building an AI writing assistant" is nothing. "We're building an AI writing assistant for first-year consulting analysts who have to turn messy client notes into clean executive summaries under 90 minutes" is a differentiator, because now the friction is obvious. The customer clarifies the friction. The friction defines the differentiation.

Source: Jason, Customer Selection Anxiety working notes

Differentiation stays constant. The product changes around it. Your differentiation is the one thing that should not move as you go from design to build to launch to growth. Features get cut, scope changes, the UI gets rewritten four times. The reason customers pick you over the incumbent is supposed to stay the same the whole way through. If it's drifting, you don't have a differentiator; you have a guess.

Source: Classroom recording, Getting the Right Customer and Differentiation

When You've Already Shipped and Nobody Cares

Advertising and visibility do not fix weak positioning. I blanketed UTA parking lots with thousands of flyers for Bus Buddy. Three downloads, one of them my parents. People didn't fail to download it because they didn't know about it. They failed to download it because they didn't care. More marketing on top of a product that doesn't match a real customer's pain just means more people who see it and ignore it.

Source: Classroom recording (Bus Buddy flyer campaign)

Knowing who your real customer is changes the battlefield. When I figured out the real customer wasn't BYU bus commuters (there basically weren't any) but UVU and U of U tracks riders, everything downstream changed. The features changed. The marketing channels changed. The partnerships changed. One KSL article later, I went from 3 downloads to the #2 navigation app in the state for a couple of weeks. Same developer, same skills, same effort. Different customer. That's the whole difference.

Source: Classroom recording (Route Tracker pivot)

Embrace the discomfort of starting over. Looking at my first app and admitting I had the wrong product, the wrong market, and the wrong customer was the single worst professional moment I'd had up to that point. It was also the moment the whole thing turned around. If you're comfortable, you're doing it wrong. You can't be ready for anything if you haven't trained for everything, and that includes the training of being wrong in public and fixing it anyway.

Source: Adam Grant, Hidden Potential

Read the mean comments. When KSL ran the Route Tracker article, the comments section was vicious. Some of it was noise. A lot of it was right. One commenter pointed out there was already a free Android app doing the same thing, so I tracked the guy down on LinkedIn, partnered with him, and we went after U of U together. Harsh public feedback is brutal to read and almost always more useful than the polite feedback from your friends.

Source: Classroom recording (KSL comments story)

The Customer Clarity Cascade

This is the sequence from the lecture. Each one depends on the one above it, so if you skip a step, everything below it gets shakier.

If step 1 is a demographic instead of a person, every step after it is also a demographic instead of a person, and the whole cascade turns to fog. Classroom recording

Go Deeper

These are the books I drew from for this guide, plus the ones I'd point you at if you want to go further on the customer and differentiation questions.

A Final Note on the Adobe Interview Story

The lecture opened with me blowing an Adobe interview and then saving it in the last 30 seconds. The reason it turned around is that I had already failed with Bus Buddy, already pivoted to Route Tracker, already gotten my face pressed into the "your customer is not who you think it is" lesson the hard way. So when the interviewer asked me one last question, I had a real story with a real customer and a real pivot to give them. That's what changed their mind, and that's what got me the job.

Your midterm project does not need to be Route Tracker. It needs to be the thing that, a year from now, you have a story about. Pick a customer. Pick them this week. Get something wrong. Fix it. That's the assignment, and honestly, that's the whole course.