How to segment customers without guessing: build resources that solve one problem each, and let what people choose tell you what they need
The short answer: You don't need to guess which customers are ready or what they need. Offer useful resources that each solve one specific problem, and the one a person chooses tells you who they are and what to talk about. Then follow up on that problem as an advisor, and keep a memory of what they've told you.
The best follow-up call I know starts with one sentence: Hey, I noticed you downloaded our playbook. I'm curious, how long have you been thinking about buying a home?
There's no pitch in it, and no guessing. The person told me what they're working on when they chose that guide. My job is to pick up the conversation they started.
Most businesses work the other way around. They collect contacts, then try to guess who's who.
Why guessing doesn't work
Here's how I describe our role to agents: "in real estate we never have to convince somebody to make a move. We just have to convince them that we are the best consultant to help navigate that process for them."
Most service businesses are the same. Life changes create the need: a new job, a growing family, a business that's outgrown its setup. Your marketing can't create that moment. It can only make sure that when the moment comes, you're the obvious person to call.
That's why announcement-style marketing struggles. In one of our sessions, a member volunteered their social feed for a live review. It was better than most: educational content, a clear call to action. Then I asked what the number one problem is for buyers in their market. Affordability, they said. Across roughly ten to fifteen posts, how many addressed it? None.
That's the gap in most marketing. It shows what you do. It rarely answers what your customer is stuck on.
The Self-Selection Loop
The Self-Selection Loop is a way to market and follow up without guessing: offer resources that each solve one specific problem, let the one a person chooses tell you who they are, follow up on that problem as an advisor, and have AI remember what they tell you.
Each turn of the loop makes the next conversation better informed.
Build a library, not a feed
Start with the people you serve and the problem each is stuck on. In real estate I think in avatars, a simple profile for each type of client: a first-time buyer, a move-up seller trading into a bigger home, a long-time owner with a lot of equity who isn't sure what to do next, a relocation buyer, and the future seller who'd like to move in the next 16 to 18 months.
Then build one resource per problem, and put them together in a library on your website. For example: a first-time investor's playbook, a guide to selling in the 2027 market, a guide to buying a home with an ADU (an accessory dwelling unit, a second living space on the property you can rent out to help offset the cost). Visitors click on whichever one is interesting to them. They're self-selecting, and all you need is a way to capture their contact information.
The choice is the information. If someone downloads the investor guide, now I can send them investment material and invite them to an investment seminar. If someone downloads the first-time buyer's guide, they're in a different group with a different set of needs. Nobody had to fill out a qualifying questionnaire.
A client texted me an introduction. Someone he knew had just bought his first rental property. My client had already sent him our landlord guide, and now he wanted to learn how I help investors. I didn't have to qualify that lead. The guide had done it.
This isn't only for clients. We use the same approach to attract agents to our brokerage, with a free blueprint for building a six-figure agent business. Any business that hires, sells advice, or serves more than one kind of customer can run it.
Two practical notes. First, these resources last. Once you build them, you can use them for years with a light refresh. Second, one resource can feed everything else. From a single guide I can have AI write a blog post on one idea in it, then turn that into a set of social posts I release over months, plus an email to my database and a text. The guide is the hub. Everything else is a spoke.
Follow up on what they chose
That's the call I opened with.
"It's a warm conversation at that point, not a cold contact." It works because it's about a problem they told you they have, not one you're guessing at.
That's also where trust is earned or lost. If you follow up as an advisor, people notice. If you follow up as a salesperson who happens to have their email, they notice that too.
Give your AI a memory of your clients
Self-selection tells you what someone needs today. Memory tells you what they'll need next.
Here's a quick test. Ask your AI assistant: who are the top five clients I should reach out to right now? Then ask: why did you pick them?
If the answer is generic, your AI doesn't know your clients. A useful answer would sound more like this hypothetical: these clients told you seven months ago they were expecting a baby, so the baby is now a few months old, and growing families often start thinking about more space. Check in. That isn't a real client. It's the kind of reasoning you want your system to be capable of.
You get there by giving AI something to remember. I use a client-profile skill, a repeatable process I've taught AI: a one-page document per client that AI builds and keeps updated from our conversations, things like their family and their history with us. When I mention a client in a new chat, it pulls that context forward, so I don't have to.
One rule I'd hold to: keep that memory to what clients have told you and to your own records. The goal is to be a better advisor, not to watch people.
If you haven't taught AI a process like this before, I walk through how in How to Teach AI Your Business Processes.
Where self-selection breaks
A download is a signal, not a commitment. Some people grab a guide out of curiosity and won't act for years, so treat the first call as a question, not a pitch. A library also can't rescue a resource nobody wants: if a guide sits untouched for months, the problem it solves isn't the one your customers have. And the loop only learns if you write down what people tell you. Skip that step and you're back to guessing.
Where to start this week
Pick the one problem your customers ask about most and build one resource that answers it. Put it somewhere people can choose it, and make sure you capture who does.
Then call the first few people who download it. Ask about the problem they told you they have, and write down what you learn in a place your AI can read.
FAQ: segmenting customers and following up
How do I segment customers without guessing? Let them choose. Offer a few resources that each solve one specific problem. What a person downloads tells you their situation and what to talk about, without a qualifying form.
What resources should I build first? Start with the most common problem each type of customer has. In real estate that might be a first-time buyer's guide, an investor playbook, or a guide for long-time owners deciding what to do with their equity. One strong resource per problem is enough to begin.
How should I follow up when someone downloads a resource? Reference what they chose and ask about the problem behind it. It's a warm conversation, not a cold pitch, and it should sound like advice, not a sales script.
How can AI help me remember what clients need? Give it a structured place to store what clients tell you, such as a one-page profile per client that it updates from your conversations. Then ask it who you should contact and why. If the why is generic, it needs better memory.
Get the next one. In the newsletter I share the resources and follow-up systems I've built, and what's working. Join the newsletter
About me: I'm Shoney Ivens, CEO of a brokerage that's produced over $5.4B in sold real estate volume in the last 5 years.
