I Built a Tool to Find People Already Looking for What You Sell

You've spent weeks or months building a product. The landing page looks good. The features work. You've probably told a few friends about it.

Then comes the difficult question:

Where are your customers?

You could run ads, build an email list, cold-message hundreds of people, or spend your evenings posting on social media and hoping someone notices.

But there's another approach that I think deserves more attention: finding people who are already asking for something your product solves.

Someone on Reddit asks for an alternative to an expensive SaaS tool. A founder on LinkedIn asks for recommendations for a developer. Someone in a Facebook group needs a website designer. A person on Hacker News describes a problem that your side project was built to solve.

These aren't hypothetical customers on a spreadsheet. They're real people expressing a need.

The challenge is finding them consistently, figuring out which opportunities are worth pursuing, and responding without spending your entire day doing marketing.

That's the problem I built Larvabot to solve.

The problem with marketing a product you've built

Most early-stage founders have a familiar set of options.

1. Wait for organic traffic.

Publish content, improve your SEO, share your launch and hope the right people discover you. This can work, but it takes time, and publishing a product doesn't automatically create demand for it.

2. Build a cold prospect list.

Find businesses that resemble your ideal customers, collect contact details and send emails. The problem is that a company fitting your target audience isn't necessarily looking for your solution right now.

3. Search social media manually.

This is closer to what I wanted. Search Reddit, browse relevant communities, follow discussions on X, read LinkedIn posts and keep an eye on forums. When someone asks a question you can answer, join the conversation.

The trouble is that this approach doesn't scale particularly well when you're also writing code, fixing bugs, handling support and building the actual business.

You can easily spend more time looking for potential customers than talking to them.

I wanted a way to automate the discovery and research without turning outreach into another generic, automated messaging machine.

What if you searched for demand instead of searching for people?

Consider two approaches to finding your next customer.

The first is to identify 500 companies that might need your product and send them a pitch.

The second is to find a handful of people actively discussing the problem your product solves, understand what they need, and reach out with something relevant.

The second approach isn't guaranteed to convert better. But it gives you a useful starting point: an observable problem, a relevant conversation and a reason to make contact.

This is the principle behind intent-based lead discovery.

For example:

• A SaaS founder asks for an affordable alternative to an established product.
• A small business owner asks for recommendations for a web designer.
• A blogger publishes a comparison of tools in your category.
• A startup directory accepts submissions from products like yours.
• Someone describes a frustrating manual process that your software automates.

Each represents a different kind of opportunity. Some are conversations. Others are potential partnerships, directory listings, relevant publications or people who might benefit from a direct introduction.

The goal isn't to message everyone. It's to identify opportunities where your product genuinely belongs.

How I approached building Larvabot

I built Larvabot around a straightforward idea: give it your website, let it understand your business, and use that information to find relevant opportunities across the web.

Rather than starting with a blank prompt and asking an AI to generate a list of generic prospects, Larvabot begins with your actual product.

Here's how the process works.

1. It learns what your business does

When you create a project, Larvabot reads your homepage, sitemap and selected pages to build a profile of your product.

That profile includes your audience, value proposition, tone, things you offer and claims you should avoid.

You can review and correct the profile before approving the plan. This matters because every subsequent search and draft depends on whether the system understands your business correctly.

The result isn't simply a list of keywords. It's a working understanding of what you're promoting and where it might be relevant.

You can read more in the Larvabot guide to how the system works.

2. It looks for conversations and relevant opportunities

Once your project is running, Larvabot searches for relevant conversations and pages.

Depending on the configured sources, this can include Reddit, Hacker News, Stack Exchange, YouTube, GitHub, Bluesky, Lemmy, niche forums and other web sources.

For example, if you've built a tool for managing customer appointments, you might want to discover discussions about appointment scheduling, recommendations for booking software, or businesses publishing content related to the problem you solve.

If you're promoting a developer tool, the relevant opportunities will look different.

The point is to find places where your product is relevant, rather than treating every mention of a keyword as a potential customer.

Larvabot also has an email outreach workflow that discovers relevant pages, blogs and websites through search campaigns. These can become opportunities for introductions, relevant content suggestions or potential backlinks.

See the email outreach documentation for the details.

3. It researches the opportunity before drafting anything

Finding a matching phrase isn't enough.

A conversation might mention your category but have nothing to do with your product. A website might look relevant at first glance but target a completely different audience.

Larvabot researches the page and evaluates its relevance to your project. Its email workflow assigns a fit score, records why the opportunity is relevant, identifies a specific detail worth mentioning and looks for a suitable contact where appropriate.

Poor-fit opportunities can be filtered out rather than filling your dashboard with noise.

This research step is important because personalization should come from understanding the actual opportunity, not just inserting someone's name into a template.

4. It drafts a response you can review

When Larvabot finds a relevant conversation, it can prepare a reply based on the discussion and your product profile.

For email outreach, it researches the opportunity and prepares a personalized message, with follow-ups where appropriate.

You can review the reasoning and draft, edit the wording, approve it, schedule it or skip the opportunity entirely.

Larvabot doesn't automatically post community replies. Every reply remains a draft for you to publish yourself. Emails also wait for your approval before they enter the sending workflow.

That's deliberate.

The aim is to remove the repetitive work of discovery, research and writing, while keeping the human involved in deciding what gets communicated.

I don't think finding opportunities should mean giving up your judgment about how to approach people.

5. It helps you measure what actually works

Another frustrating part of early-stage marketing is not knowing which activities lead to results.

You might post in a community, email a blogger, submit your product to a directory and share a launch announcement. Later, you see a few visitors in your analytics, but connecting them to the original activity can be difficult.

Larvabot includes visitor attribution to help connect outreach activities to website visits. Its reports can also show email replies and backlinks, giving you more context about which opportunities are producing results.

This doesn't mean every visit becomes a customer, or that attribution can explain every conversion. It gives you better information to use when deciding where to focus next.

A quick demo of the workflow

The video above shows Larvabot in action. The general workflow is:

1. Add the website you're promoting.
2. Review the profile and proposed marketing plan.
3. Approve the plan.
4. Let Larvabot find and research relevant opportunities.
5. Review the generated replies and email drafts.
6. Approve, edit, schedule or skip the outreach.
7. Use the reports to see which activities brought visitors and generated responses.

You don't have to keep the dashboard open while the background tasks run.

If you'd like to try it yourself, the quickstart guide walks through creating your first project and getting to your first draft.

Why I didn't build another mass-email tool

There's no shortage of software for sending cold emails, collecting contact information and automating sales sequences.

Those tools have their place. But I wanted Larvabot to address an earlier question:

Where should I spend my time reaching out in the first place?

A contact database can tell you who works at a company. It can't necessarily tell you that the person is currently asking for a solution like yours.

A generic AI writer can generate a convincing email. It can't make an irrelevant opportunity relevant simply by improving the wording.

And a social listening tool can alert you to a keyword mention without necessarily helping you research the opportunity and prepare a useful response.

Larvabot brings discovery, research, drafting and attribution into one workflow.

It isn't designed to blast thousands of strangers with the same message. The emphasis is on finding relevant opportunities and preparing thoughtful outreach.

That philosophy also informs the product's sending defaults: small batches, personalized messages, opt-outs and deliberate sending limits.

I've documented the reasoning in Outreach without being a spammer.

The pricing model: one fee, no credits

I also wanted to avoid a pricing model where using the product more means constantly worrying about consuming credits.

Larvabot currently has two plans:

• Starter: $19/month for one website.
• Pro: $49/month for unlimited websites.

Both include unlimited leads, drafts, emails and conversations without per-lead charges, per-email charges or credit bundles.

There is one important detail: Larvabot's platform fee doesn't mean every third-party service is magically free. You configure the AI, search and email providers your projects need, using their own accounts and available allowances. Larvabot includes quota safeguards to avoid crossing configured provider limits, and work may pause if all available providers run out of quota.

The Pro plan also includes Larvabot's own search server and additional community-monitoring capabilities.

You can find the full breakdown in the pricing and billing documentation, along with the explanation of how the free-tier and quota safeguards work.

Who is Larvabot for?

I think Larvabot is particularly useful for people who have a product or service worth promoting but don't have a dedicated marketing team.

That includes:

• Indie hackers launching their first SaaS.
• Developers promoting open-source projects or developer tools.
• Freelancers looking for relevant hiring opportunities.
• Small agencies looking for businesses that need their services.
• Startups looking for product discussions, partnerships and relevant publications.
• Small businesses trying to find customers without relying exclusively on paid advertising.

It can also help with the less glamorous work of getting a new product noticed, such as finding relevant directories and preparing submissions.

It's probably not the right fit if your entire strategy depends on sending huge volumes of unsolicited emails or if you expect a tool to guarantee customers without any involvement from you.

Larvabot helps you find and work on opportunities. You still need a product people want, a relevant message and good judgment about when to reach out.

What's next?

I'm continuing to build Larvabot around a simple principle: marketing shouldn't require spending every day manually hunting for people who might need what you've built.

There are already people asking questions, comparing products, looking for recommendations and searching for solutions across the internet.

The challenge is finding the right conversations, understanding their context and responding in a way that's actually useful.

That's what I'm trying to make easier.

If you're working on a side project or building a SaaS and customer discovery is eating into your development time, you can explore Larvabot here:

Visit Larvabot

You can also start with the documentation if you'd rather understand how the individual workflows work before trying it.

I'd be interested to hear how other indie hackers approach this problem, too. Do you manually search for people discussing your product category, rely on content and SEO, or use a completely different approach?

I'm especially interested in what you've found works when you're a small team without a dedicated marketing budget.