How to Get Users for Your SaaS When Nobody Knows You Exist

You shipped. The landing page is up, billing works, onboarding is smoother than it has any right to be. You posted the link somewhere, got a small spike of visitors, and then the graph went flat. A week later you are refreshing analytics at midnight wondering whether the product is the problem.

Usually it isn't. The product is fine and almost nobody has seen it. Getting users for a SaaS with no audience is a separate discipline from building one, it runs on a much slower clock, and almost none of it feels like engineering — which is exactly why technical founders keep putting it off in favour of one more feature.

The first ten, the first hundred and the first thousand are three different jobs

Most advice about how to get users for a SaaS is written for one of these stages and then applied to all three. That is why it contradicts itself. What gets a product from zero to ten users is close to the opposite of what takes it from a hundred to a thousand, and doing them in the wrong order wastes the scarcest thing a founder has.

There is a second reason the order matters. The first ten users are not really an acquisition exercise, they are a research one. They tell you which sentence on your homepage is doing the work, which feature they thought they were buying, and what they were using before. Every later channel is cheaper and faster once you know those three things, and none of them can tell you.

Where your first SaaS users actually are

The people who will pay you first are the ones already complaining about the problem in public. They have the pain badly enough to have typed it somewhere. Finding them is a search problem rather than a marketing one, and it is entirely doable in an afternoon.

1. 1Name the room, not the channel“Reddit” is not a plan. A specific subreddit where your buyer posts weekly is. Do the same for Discord and Slack communities, niche forums, Stack Overflow tags, GitHub issue threads on adjacent tools, and the replies under accounts your buyer follows. Write down eight to twelve rooms by name. That list is your distribution strategy for the next month.
2. 2Read for a week before you post anythingYou are looking for the words people use for the problem, which are almost never the words on your landing page. Collect the phrasings. They become your headline, your outreach opener, and your search keywords later.
3. 3Answer the question, mention the product secondFind someone describing the problem you solve and give them a genuinely complete answer — one that would help even if your product did not exist. Then, in one line at the end, say you built something for it. A helpful answer with a link converts; a link with no answer gets you banned.
4. 4Take it to a callWhen someone signs up, email them personally and ask for fifteen minutes. A decent share say yes while you are still small and unknown, and those conversations are worth more than the next month of analytics. Ask what they were doing before, and what nearly stopped them signing up.
5. 5Do it again tomorrowTwo of these a day for a month is sixty conversations. That is enough to get a small product to its first paying handful, and it is the only reliable way to do it with no audience and no budget.

How to find early adopters for a SaaS when you have no list

If the rooms are not obvious, work backwards from the alternatives. People shopping for a replacement leave traces in predictable places:

• Review sites — the one- and two-star reviews of your closest competitor are a list of people actively unhappy with a tool in your category, annotated with the exact reason.
• “Alternative to X” searches — whichever competitor people are trying to leave, someone is running a thread about it. Those threads are rooms.
• Job posts — a company hiring for the role your product replaces has both budget and an admitted problem.
• GitHub — issues and discussions on adjacent open-source projects are full of people describing your problem in technical detail, for free.
• Your own support inbox, if you have any users at all — ask each one where they would have looked for something like this. They will name rooms you have never heard of.

Launch surfaces, and what each one is actually worth

Launches feel like distribution because they produce a graph with a shape. They are better understood as one-off injections of attention: valuable because they hand you people to talk to, not because the traffic stays.

One thing worth saying plainly: launching to nobody wastes the launch. If you have no list and no following, the spike is small and the badge is the only durable asset. Spend the first month building the list — even fifty people who asked to be told — and launch into it.

Cold outreach that does not read as spam

Cold email still works for B2B SaaS at small volume, and stops working almost exactly when you automate it. The version that works looks like a message from a person who read something you wrote, because that is what it is.

1. 1Twenty a day, hand-written, maximumNot two hundred from a sequencer. At twenty you can reference something specific about each recipient, and that specific line is the entire reason anyone replies.
2. 2Lead with the observation, not the product“I noticed you are still doing this by hand” beats any description of your features. You are demonstrating that you understand their situation before asking for anything.
3. 3Ask for a reply, not a demo“Is this actually a problem for you?” gets answered far more often than a calendar link does, and a reply is what starts the conversation. Book the call in message two.
4. 4One follow-up, then stopA single short follow-up after four or five days recovers a meaningful share of the people who meant to reply and forgot. A third and a fourth make you the thing they are trying to clear out of their inbox.

The two channels that compound

Everything above trades your hours for users at a fixed rate. It works, it does not scale, and at some point you need something that keeps returning after you stop pushing. For a small SaaS there are realistically two, and both are slow enough that the right time to start is before you need them.

Search, where only the long tail is winnable at first

A new domain is not going to rank for the head term in your category. It will rank for questions so specific that the large sites never bothered to answer them — the exact error message, the workflow between two named tools, the question a buyer asks in week three of evaluating. Those queries have low volume and startling conversion, because someone typing them has already diagnosed their own problem.

• Write for one specific question per page, and answer it completely enough that the reader does not need another tab.
• Take the phrasings from step two above — they are the queries, in your buyer's own words.
• One genuinely substantial article a week beats five thin ones, which is the trap programmatic pages set.
• Expect three to six months before anything moves. That delay is exactly why the long tail stays winnable.

Short-form video, and the problem specific to software

Short-form video is where the unpaid reach still is, and it is exactly where software products struggle. A cafe films a croissant. A B2B tool has an interface, and filming an interface well means screen recordings, timing, and captions that do not collide with the platform's own UI. Most founders conclude it is not for them, and leave the last free-distribution channel on the table.

It is worth doing imperfectly rather than not at all. A clip of the product doing the thing, with a hook in the first second and the payoff before the third, outperforms a polished brand film for a product nobody has heard of — and unlike search, it can work in week one.

Free tools, open source, and the thing you give away

The strongest acquisition asset a small SaaS can build is usually a smaller free thing that solves one slice of the problem completely, with no account required. A calculator, a linter, a template, a CLI, a package. It earns links, it ranks for its own queries, it gets shared where an ad never would be, and everyone who uses it has already self-identified as having your problem.

The rule that makes it work: the free thing has to be genuinely useful on its own to someone who never becomes a customer. Anything that is a demo wearing a tool's clothing gets recognised immediately, and shared by nobody.

What to measure in the first ninety days

A week that fits around actually shipping

None of this requires a marketing hire. It requires that some fixed part of the week is not available for coding. The version that survives contact with a real build schedule looks roughly like this:

1. 1Monday — write the one thingTwo hours on the week's article or teardown, aimed at one specific question a buyer types. This is the compounding half and the first thing that gets dropped, so give it the best hours of the week rather than the leftovers.
2. 2Tuesday to Thursday — twenty minutes in the roomsAnswer two questions properly in the communities on your list. No pitching, no link unless it belongs. This is the part that produces the next five users.
3. 3Any day — ten outreach messagesHand-written, specific, asking a question rather than booking a demo. Thirty minutes.
4. 4Friday — three numbers and one userActivated users, week-four retention, conversations had. Then one fifteen-minute call. Half an hour of analysis and the most valuable quarter-hour of your week.
5. 5Every day — something goes outOne post on the platforms where your buyer scrolls. This is the piece that is genuinely automatable, and the piece most founders abandon by week three because it is small, daily and never urgent.

That is roughly four hours a week, and it will not look like it is working for two months. Then the article from week three starts bringing people in while you sleep, someone links the free tool, a clip does better than it should, and the compounding half quietly takes over from the manual half. Nearly everyone who gets there describes the same shape — and nearly everyone who quits, quits in month two.