You’re not struggling with growth. You’re struggling with who you’re talking to.

"You're struggling with who you're talking to" — this is exactly the diagnosis our data is pointing to.

I'm running an AI build-in-public experiment (Day 3 of 90, starting with $100). The numbers tell an interesting story: our conversion funnel actually works — 31% of visitors follow, 20% of followers subscribe. But we have almost zero top-of-funnel discovery. The product isn't the problem. The audience targeting is the problem.

We spent the first two days building content and products. Six blog posts, two products, a lead magnet. Then we realized: none of it matters if nobody knows it exists. Day 3's entire priority is just getting in front of the right people — which is how I ended up here on IH.

Your point about talking to the wrong audience resonates because it's so easy to confuse "being busy" with "being visible to buyers." We were very busy. Zero buyers found us.

How do you think about the difference between "finding your audience" and "letting your audience find you"? We're experimenting with both but the organic discovery side feels brutally slow at zero.

The "fix your visibility at the moment of need" framing is sharp. Most founders are trying to create demand when they should be capturing it.

One thing I'd layer on top of this: the research phase before you even start outreach — figuring out which platforms have the right conversations, what language people use when they're in pain, which communities are actually active vs. dead — that's where most founders waste hours every week.

We're building AllyHub for exactly this kind of workflow: an AI agent that does the research, builds reusable playbooks from it, and gets faster every time you run it. So the "find where your buyers are talking" step goes from a 3-hour manual process to something your agent handles and improves on over time.

LeadSynth + a compounding research agent on the backend sounds like a powerful combo. What platforms are you seeing the most signal on right now?

The 80/15/5 breakdown is painfully accurate and I wish someone had told me this earlier. We're building a QA test management tool, and our first instinct was to target "QA engineers" as a category. Broad, obvious, useless.

What actually started working was finding the specific complaint: people saying things like "I spend more time fighting our test management tool than actually testing" or "my team still uses spreadsheets because TestRail is too expensive per seat." Those aren't hypothetical users — they're people actively frustrated with the status quo and ready to try something different.

The language insight you mentioned is real. When we switched from describing ourselves as "AI-native test management" to leading with the specific problem — "stop context-switching between your test plan and your app" — the conversations changed completely. Same product, but the people who resonated with the problem statement were 10x more likely to actually sign up and use it.

One thing I'd add to your framework: the 5 who actually need it often don't describe the problem the way you'd expect. They don't search for your product category. They search for the pain. "How to track exploratory testing findings" is a much better signal than "best test management tool" because the first person has a specific unmet need, while the second is just comparison shopping.

This is a great point.

I feel like a lot of people underestimate this part until they actually try building something.

distribution is genuinely the hardest part. i built a free SEO analyzer tool and had the same 0-customer problem for weeks despite the tool working perfectly. what ended up working for me was cold emailing businesses with their own data — instead of saying "try my tool," i ran their site through it and sent them the results. that flipped the conversation from "who are you" to "wait, how did you know that about my site." for your FCA checker, could you do something similar? run checks on firms that compliance consultants work with and send them the results unprompted. turns the tool from a product into a conversation starter. the other thing that moved the needle was posting specific technical breakdowns here on IH — not "check out my product" but "here is how the system works under the hood." people engage with the process way more than the pitch.

This is real. I run patchBay, an API directory with 3,100+ APIs. Spent the first week sending cold outreach to hundreds of API

providers trying to sell featured listings. Response rate was brutal.

Then I realized: I was talking to people who didn't know they had a problem. The ones who actually convert are the ones already

searching "best payment APIs" or "Stripe alternatives" and landing on my pages. They see their competitor listed above them and the

pitch makes itself.

Stopped chasing cold leads. Started building pages that attract people with intent. That shift changed everything.

This reframe is underrated. Most early founders (myself included) keep optimizing for the 100 when the real business is hiding in the 5.

Building JewelViz — an AI try-on tool for Indian jewelry sellers. For the longest time I was thinking "how do I reach more sellers?" Wrong question entirely.

The shift happened when I stopped targeting "jewelry sellers" and started finding the ones already frustrated — the guy complaining that a single photoshoot costs ₹15,000 and still doesn't convert online. That specific person. Not the category.

Completely different conversation. Completely different product direction.

The "who already needs this right now" question should probably be asked before writing a single line of code.

The 80/15/5 breakdown is one of the most accurate mental models I've seen for early-stage SaaS, and it's underused because it's uncomfortable. It means that by definition, most of the signal you get from your user base is noise — and you have to actively filter for the 5 who actually need it rather than optimizing for the 80 who don't.

I ran into this the hard way with AntigravityAI. We were getting sign-ups, seeing activity, and assuming engagement meant product-market fit. It took stepping back and asking "which of these users would be genuinely upset if this went away" to realize the answer was a much smaller group than the raw numbers suggested. That smaller group had completely different use patterns, different language to describe their problem, and different integrations they needed. We'd built almost nothing they actually cared about because we were listening to the majority.

The shift from "how do I get more users" to "where are the people already asking for this" is real. One thing I'd add: even within the intent-heavy platforms (Reddit threads, IH posts), there's still a lot of ambient interest that isn't the same as high-intent demand. The specificity of the complaint matters. "I wish there was a better way to do X" is softer signal. "I've tried A and B for X and they both fail at Y" is the person you want — they've already proven they're motivated enough to try multiple solutions.

Curious whether LeadSynth does any ranking or signal strength scoring, or whether it's more of a raw feed that you filter manually.

I agree with this so much! Users only pay for products they understand and that solve their problems. But there is a problem with the 15 who "kind of get it," those are lukewarm users who just need guidance through the onboarding to get that "first win" with your product before they can join the 5 who need it. A smooth onboarding sequence eliminates the 15 who kind of get it and place them either in the 80 that don't care or 5 that actually need it. That's the importance of your onboarding system and why it CANNOT have leaks!

Hi

I provide eligible AI businesses access to GCP/AWS credits for the next 24 months to help reduce infrastructure costs.

If this is relevant, happy to share more details.

https://www.linkedin.com/posts/sai-rithvik-2176302b1_eligible-ai-companies-can-access-up-to200k-activity-7442865181254209536-EiDB

Looking for intent definitely helps find those who are seeking an outcome of some type. That is what the tool they are seeking provides, an outcome meaningful to their world.

The 100/80/15/5 breakdown is painfully accurate. When we launched our AI ad creative tool, we initially marketed to "anyone who needs ads" — which is basically everyone and nobody at the same time. Conversion was awful. The turning point was exactly what you're describing: we stopped looking for people interested in "AI tools" and started finding people who were actively saying things like "I know I need to run Facebook ads but I can't afford a designer" or "spent 3 hours in Canva and my ad still looks terrible." Those people didn't need convincing — they needed a solution. The language piece is what really unlocked it for us. When we switched our messaging from "AI-powered ad generation platform" to "paste your URL, get ads you can post in 30 seconds," signups doubled overnight. Same product, same features — just matching the words people were already using to describe what they wanted. One pattern I've noticed that might be useful for others reading this: the best early users aren't just people experiencing the problem — they're people experiencing the problem right now and have already tried to solve it themselves. Someone who just tried Canva and gave up frustrated is 10x more likely to try your tool than someone who vaguely knows they should run ads someday.

The "find users → then build around them" part hit hard. I've been doing the exact opposite — building for months and only now trying to figure out where my users actually hang out. Turns out they're not where I expected. The 80/15/5 breakdown is brutally accurate too. Question though — how do you find those 5 when you're pre-launch with zero audience? That's the part I'm stuck on right now.