2
11 Comments

I could build products—but struggled to turn them into something that converts

After getting past the idea stage, I ran into a different problem.

I had something built—but no clear way to present it.

Writing hooks felt like guesswork.
Landing pages took too long.
And I wasn’t sure how to structure a simple funnel without overcomplicating it.

I tried piecing things together from templates and prompts, but it never felt cohesive.

So I ended up building a system to make this part more structured.

It became the Conversion Engine Kit—a set of prompts that take me from:

“I have an idea” → to a hook, conversion copy, and a simple funnel… usually in an afternoon.

It’s built around three steps:

  • Generating hooks and offers that actually fit the product and audience

  • Writing the core landing page copy (headline, bullets, CTA) in one pass

  • Mapping a simple funnel—entry, nurture, offer, and follow-up—without overthinking it

The main goal was to reduce friction between “something built” and “something that sells.”

I’ve been pairing this with Blink.new to quickly turn ideas into working tools, then using this system to shape how they’re presented.

Still refining it, but it’s made that middle stage much clearer.

Curious how others here approach this—
Do you write your copy first, or figure it out after building?

posted toAvatar for product Conversion Engine Kit
Conversion Engine Kit
  1. 1

    This hits exactly where I am right now.

    I'm a doctor who built a free AI travel app.

    257 users, 26 countries, zero conversions

    on affiliate links.

    The product works. People use it. But I clearly

    failed at the "something that sells" part —

    I never thought about hooks or funnels because

    it's free.

    Your question about copy first vs build first

    is interesting. I did neither — I built,

    launched, and only now am I thinking about

    how to communicate value.

    Do you think a conversion framework still

    applies to free tools with affiliate revenue?

    Or is it a completely different game?

    1. 1

      Thanks for sharing your story — really appreciate it.

      257 users from 26 countries on a free AI travel app built by a doctor is honestly impressive. The fact that people are actually using it shows the core idea works.

      I’m in a similar boat myself, so I totally get that moment when you realize “it works, but how do I actually make it useful for monetization?”

      To your question: I believe a conversion framework can still help even for free tools with affiliate links, but maybe in a lighter way.

      Instead of aggressive selling, it’s more about making the value clearer in the first few seconds and gently guiding people toward the affiliate links without being pushy.

      The Hook + Offer Generator has helped me (and a few others) find better ways to explain “why this is worth using” even when the tool is free.

      Would be curious to hear — what’s the biggest challenge you’re facing right now with communicating the value or getting people to click the affiliate links?

      Happy to brainstorm together if you want. No pressure.

  2. 1

    That middle stage is where most products quietly die.

    Not because the product is bad — but because the story around it never gets sharp enough.

    The interesting part is what you said about “it never felt cohesive.”
    In my experience, that usually comes from building → then trying to retrofit positioning after.

    Curious — have you found it works better when the hook/positioning is defined before building, or does your system work equally well post-build?

    1. 1

      That's a great observation — and you're spot on.

      In my experience, the best results come when you define the hook and positioning before building.

      The Conversion Engine Kit is actually designed with that in mind. The Hook + Offer Generator prompt forces you to clarify:

      • Who it's for

      • What problem it solves

      • Why it's different / better

      • What the transformation or outcome is

      before you even generate the tool itself.

      When I build with it, I usually start with that prompt first. It gives a much sharper hero section, clearer value proposition, and stronger overall story. The tool then feels cohesive from day one instead of needing heavy retrofitting later.

      That said, the prompts also work surprisingly well post-build. A lot of people use the Hook + Offer Generator as a "positioning rescue" tool — they already have a working tool, run it through the prompt, and it often surfaces much better angles they missed initially.

      So it’s flexible, but I’ve personally seen the highest success rate when positioning is done early.

      Have you run into situations where retrofitting positioning worked well, or do you also prefer doing it upfront?

      1. 1

        Yeah — I’ve seen retrofitting work, but usually only when the product already has some traction signal.

        If it’s early, the difference is huge. When positioning is clear upfront, even small things like naming, domain, and first impression start compounding.

        I’ve noticed a lot of solid tools struggle not because they don’t work — but because they feel generic at first glance, so people don’t even give them a chance.

        Curious — when you’re generating hooks, do you think naming plays into conversion early, or do you treat it as something secondary?

        1. 1

          Great point — I completely agree. When there’s no traction yet, upfront positioning makes a massive difference. Retrofitting later usually feels like damage control.

          On naming: I treat it as very early and important, not secondary.

          Because Blink locks in the URL the moment the project is created, I always recommend choosing (or at least deciding on) a tool name first. A strong, benefit-driven name helps shape the hook, headline, and overall offer much more effectively. When the name is weak or generic, even good copy tends to feel flat.

          So yes — I strongly encourage doing naming early, before running the prompt.

          Have you seen cases where changing the name late completely turned around a tool’s reception?

          1. 1

            Yeah — I’ve seen it happen a few times, especially with early tools.

            Usually the product didn’t change much, but the name shifted from something generic/descriptive to something that felt more distinct or outcome-driven.

            The interesting part is it’s not just memorability — it changes how people interpret the product before they even read the copy.

            Same landing page, same features, but the first impression feels sharper so people actually give it more attention.

            It’s subtle, but it compounds a lot in that early stage when every click matters.

            Do you usually lean more towards descriptive names, or something more brandable for your tools?

            1. 1

              Yeah, that “subtle but compounds a lot” effect is exactly what I’ve observed too. Even with the exact same landing page and features, a sharper name makes people slow down and actually read instead of scrolling past.

              A strong brandable name tends to make the whole tool feel more premium and memorable from the very first impression. Purely descriptive names (like “AI Hook Generator v2”) often make the tool feel more like a utility/demo.

              That said, I always recommend starting with a descriptive working title during ideation, then evolving it into something brandable once the core value is clear.

              1. 1

                Yeah — that transition from descriptive → brandable is where things usually click.

                One pattern I’ve noticed:

                It’s not really a choice between descriptive or brandable — it’s more about what the name is doing in the first 3 seconds.

                Early stage, the job is:

                → reduce confusion

                → signal outcome fast

                Later stage, the job shifts to:

                → build recall

                → feel like a “real product”

                The tricky part is when a name tries to do neither well:

                too generic to stand out

                but not clear enough to explain itself

                That’s usually where good tools get ignored.

                The middle ground that seems to work best is:

                a name that hints at the outcome, even if it’s slightly abstract

                So it feels like a product…

                but still guides interpretation without needing heavy explanation.

                That’s also where I’ve seen naming start to impact conversion directly — not just memorability.

                Curious — when you renamed those tools you mentioned, was it more about making the outcome clearer, or making it feel more distinct/premium?

                Happy to share a few patterns I’ve seen work (and fail) if you're experimenting with this more deeply — didn’t want to overload the thread.

                1. 1

                  Thanks for the detailed breakdown — really insightful.

                  For me, the rename was mostly about making the outcome clearer and helping the tool feel less like a generic AI demo from the first glance.

                  Appreciate you sharing your patterns.

                  1. 1

                    Yeah — that makes sense. Clearer outcome usually sharpens everything else.

                    One thing I’ve noticed though:

                    A strong name doesn’t just clarify — it forces a decision.

                    Early tools usually have multiple possible directions.

                    A generic name keeps them all open.

                    But a sharper name collapses that into one clear interpretation.

                    That’s where conversion improves — but also where positioning gets more opinionated.

                    So naming isn’t just labeling… it’s locking in direction.

                    Curious — when you renamed, did it feel like you were narrowing the use case too, not just improving clarity?

                    Happy to swap a few patterns I’ve seen work here — feels like we’re thinking along similar lines.