There is a fantasy where you build the right thing and people find it. It happens to roughly nobody.
Selling early is not a growth channel or a funnel. It is you, personally, talking to people who have the problem, one at a time, until enough of them say yes that a pattern appears. This chapter is about doing that deliberately instead of hoping.
Decide what you actually stand for
Before you write a word of marketing, decide what you are claiming. Not a slogan. A position.
A position is the answer to a question your customer is silently asking: why this one, and not the obvious one? If you cannot answer that in a sentence, no amount of design will save the page.
Three things make up a usable position.
Who it is for. Naming the customer excludes people, and excluding people is what makes the right person feel seen. "For independent clinics" is stronger than "for healthcare teams", even though it addresses fewer people, because the right reader recognises themselves.
What you are against. Every good position has an enemy: a way of working, an assumption, a category of tool that everyone tolerates. Being against something specific makes you legible. "Built for the clinics that big practice-management software forgot" tells me more than any list of features.
What you promise. One outcome, stated plainly, that a customer could later hold you to. Promises you can be held to are the only ones worth making.
Then there is the surface of it, the name, the colours, the tone. That matters less than people think at this stage, but consistency matters a lot. Pick one look and one voice and repeat them until you are bored of them. You will be bored long before your audience has noticed.
Write a page that sells one thing
Most first landing pages fail the same way: they try to be a brochure for everything the product could ever be, and end up saying nothing.
A page that works, in order:
The first screen. What this is, who it is for, and what changes. A visitor decides in a couple of seconds whether to keep reading, and they decide based on the headline. Write the headline as the outcome, not the mechanism.
Proof it is real. A screenshot, a short video, a number, a name. Anything that shows this exists. Abstract benefit copy over a stock illustration is the fastest way to look like nothing.
What it actually does. Three or four things, in the customer's language. Not every feature. The ones that map to the reason they came.
Why you. Your position, in a sentence or two. This is where the enemy from the last section earns its keep.
Handle the objection. There is one thing everybody worries about. Price, or lock-in, or whether it works with what they already use. Answer it on the page, in plain words. Pretending it does not exist does not make it go away, it just makes people leave without telling you.
One thing to do next. One. Same action repeated down the page. Two competing calls to action cut your conversion, they do not double it.
Two habits that will improve any page you write. First, use the words your customers used in the conversations from chapter two, literally. Second, read the whole page out loud. Anything you would be embarrassed to say to someone's face should not be on the page.
Get your first ten users
Your first ten customers do not come from marketing. They come from you sending messages.
This feels like cheating, or like it does not count. It counts. Almost every business you admire started with a founder doing something that did not scale.
Where to look, roughly in order of how well it works:
People you already know who have the problem. Not asking for a favour. Asking for fifteen minutes and honest feedback, and mentioning what it costs if they want to use it.
Communities where these people already gather. Forums, subreddits, Slack and Discord groups, local associations, industry newsletters. The rule here is simple: be a member first. Answer three questions before you mention what you built, and mention it once, in context, and let it go.
Direct outreach to people who fit. Short messages, specific to them, referencing why you are contacting them and not five hundred others. Two sentences and a question beats a template every time. If your message could have been sent to anyone, it will be treated like it was.
People who complained in public. Someone posted last month about exactly this problem. Replying to them with "I built something for this, want to try it?" is not spam, it is service, as long as it is true.
Ten is the target for a reason. Ten is small enough to do by hand and large enough that patterns start to show. Below ten, everything is anecdote. Around ten, you start noticing that four of them said the same odd thing, and that odd thing is usually your next move.
Talk to every one of them yourself. Do not automate this. The founder who has personally onboarded ten customers knows things that cannot be written down.
Write down your ideal customer
After those first ten, you will know something you did not know before: which of them were a real fit.
Some customers are wonderful. They get value quickly, they ask good questions, they stay. Others are hard work, they want something adjacent to what you built, and they churn. That difference is not random, and finding the pattern in it is one of the highest-leverage things you can do.
Write down the profile. Be specific and be willing to be wrong:
- What kind of organisation, what size, what industry.
- What has to be true for them to need this. A trigger, a scale, a rule they have to follow.
- Who inside the organisation feels the pain, and who signs off on spending.
- What they use today, and what they hate about it.
- Where you can reliably find more of them.
Then use it as a filter, not a wish list. When someone who does not fit the profile asks for a feature, that is not a roadmap item, that is a signal you are being pulled off course. The discipline of saying "we are not for you, yet" is worth real money.
You will revise this every few months. That is the point. It is a working hypothesis, not a marketing persona with a stock photo and a name.
Understand who you are up against
Competitors are worth understanding and not worth fearing. Their existence proves the problem is real and someone is already paying to solve it.
What to actually look at:
What they promise, in their own words. Their headline is their best guess at what the market wants. It is free research.
Their pricing, and its shape. Whether they charge per seat, per usage, or per outcome tells you how they think about value and where their model is uncomfortable.
Their reviews, especially the middling ones. Five stars tells you nothing. One star is often an angry outlier. Three stars is where people write the honest paragraph about what it does badly, and those paragraphs are your positioning.
Who they are clearly built for. Most established tools drift upmarket over time, toward bigger customers with bigger budgets, and abandon the smaller end. That abandoned end is usually available.
What not to do: build a feature comparison table as your strategy. Matching a competitor feature for feature means competing on their terms, and they have more people. Win by being obviously better for a specific person, not marginally better for everyone.
And be honest in public about where a competitor is genuinely better. It costs you almost nothing and buys a surprising amount of trust, because everybody can tell when a comparison page is rigged.
Choose how you are going to sell
There are a few different ways to sell, and picking one deliberately beats doing all of them badly.
Self-serve or conversation? If your price is low and the value is obvious in minutes, people should be able to sign up and pay without talking to you. If the price is meaningful or the setup needs thought, a conversation will convert better than any page. Do not put a demo call in front of a fifteen dollar product, and do not expect a three thousand dollar product to sell itself from a pricing page.
Inbound or outbound? Inbound means people find you, through content, search, communities, word of mouth. It compounds and it is slow. Outbound means you go to them. It works immediately and it stops the moment you stop. Early on you almost certainly need outbound, because you do not have time to wait for compounding. Start inbound anyway, in the background, so that in six months it has begun.
Who do you talk to? The person with the problem and the person with the budget are often different people. Selling to the sufferer who cannot approve spending feels great and closes nothing. Find out early who has to say yes.
Pick one motion. Run it for a month, properly, before judging it. Half-running three channels is the most common way to conclude that nothing works.
Show up every week
Consistency beats brilliance here by an embarrassing margin. One decent post a week for six months will do more for you than one excellent campaign.
What to publish, when you have no audience and no time:
What you are learning. You are doing something most people only talk about. The specifics of what surprised you this week are more interesting than any advice you could give.
What you built and why. Show the thing. A short clip of it working, and the reason it exists.
What your customers told you. With permission, the real problems from real conversations. This is the most useful content on the internet and almost nobody writes it.
Answers to questions you keep getting. Every question a prospect asks twice is a post, and eventually a page.
Two rules that keep this sustainable. Publish where your customers actually are, not where you enjoy being. And keep a rhythm you can hold on a bad week, because the value is entirely in not stopping.
Next up: Chapter IV is about what to do once there are enough customers to see patterns.