My LinkedIn roast tool was a marketing hack. It became the company.
A year ago I had a B2B SaaS, linkeme, and the problem every bootstrapper knows: 0 marketing budget, no distribution. So instead of ads I couldn't afford, I built a dumb little viral tool to pull people in.
LinkedInScope: paste your LinkedIn URL, get your profile brutally roasted, and receive personal branding advice. Fun, a little mean, very shareable. The plan was simple. Roast, laugh, "btw, here's my B2B product."
The roast tool worked for traffic. But the funnel data told a story I wasn't expecting. People didn't show up to promote their company. They showed up to fix themselves.
They obsessed over their headline, their tone, how they came across. The "now go promote your business" angle fell flat.
The "here's what's broken in your personal brand" part? They devoured it, shared it, came back for more.
I'd accidentally run the cleanest market research of my life. My distribution hack was screaming one thing: the demand is personal branding, no company promotion.
So I stopped arguing with my users' behavior and rebuilt the product around what they actually wanted: a powerful automatic LinkedIn personal branding tool. Linkeme is now:
- pulls your real LinkedIn profile and builds a model of who you are
- gives you a straight branding diagnostic (what's strong, what to fix)
- generates posts in your own voice
- runs on full auto (scheduled, hands-off) so you actually stay consistent, the thing everyone fails at.
LinkedInScope is still the front door. Get roasted, see the gap, then fix it for real.
The lesson worth stealing: your top of funnel "growth hack" is also your best research instrument, if you read behavior instead of stated intent. I implicitly asked people to promote their company. Their clicks told me they'd rather build themselves. That signal was worth more than any survey.
Still early, still building in public. Happy to share what's worked. And I'm curious: has a side tool ever ended up reshaping your main product?
This resonates. I think most founders feel the signal way before they act on it. Out of curiosity, how long between "huh that's weird" and "ok I'm rebuilding the whole thing"? And what finally tipped you over?
Honestly? The "huh that's weird" showed up almost immediately, within the first week or two of LinkedInScope getting traffic. The "ok I'm rebuilding the whole thing" took another 3 months, because I kept explaining the data away.
You nailed it, I felt the signal long before I acted on it. Every week the B2B "go promote your company" CTA got almost no clicks, while people screenshotted and re-shared the personal branding part. I had a dozen comfortable explanations ready: wrong onboarding, wrong copy, wrong segment.
Anything except the obvious one.
What finally tipped me over wasn't a new data point. It was the same data point for the Nth week in a row. At some point "this is just noise" stops being believable. The real unlock was admitting I was attached to the version I'd already built, not to what people actually wanted.
So roughly: signal in days, decision in weeks, ego in months. The rebuild was the easy part.
This is a strong signal because the roast tool did more than bring traffic. It exposed the real buying emotion.
The risk now is assuming the same emotion that makes people share the roast will also make them pay for ongoing personal branding.
Those are close, but not identical.
The key bridge is probably the moment after the roast: when the user feels exposed enough to want a fix, but before the whole thing becomes just entertainment.
That transition is where I’d be careful.
Happy to put the tighter roast-to-paid-conversion angle in writing if useful. This is exactly the kind of funnel where attention is already proven, but the paid conversion story has to be handled carefully.
This is the sharpest thing anyone has said about it, and it's exactly what keeps me up. The emotion that makes you share a roast (ego sting, a laugh, "omg look at this") is not the emotion that makes you pay every month. You're right that they're adjacent, not the same.
The way I frame it now: the roast creates a spike, but paying for ongoing personal branding requires a habit. The danger is precisely the decay you describe, the exposed-and-want-a-fix feeling fading back into "that was fun" before it becomes something I actually do every week.
In my funnel that bridge has a literal location: the moment right after the roast, where I hand the user a diagnostic (here's what's actually weak) and then a first post generated for them (here's the fix, already done). That's where I try to turn the sting into "oh, the fix is real and effortless." And honestly that's also exactly where I watch people drop. Attention is proven. Free signup converts okay on the "want a fix" emotion.
The paid, ongoing part is the step I have the least proof on, and it's my current obsession.
My working bet: the recurring value can't be "we branded you," it has to become "you stay consistent without thinking about it," because that need doesn't decay after the laugh.
And yes, I'd genuinely take you up on the written roast-to-paid angle. That's the exact part of the funnel I'm trying to earn right now.
That is exactly the right place to focus.
The danger is not the roast failing. It is the product winning attention with one emotion, then asking for payment through a different one.
I would not try to solve that casually in a comment because the handoff from sting to fix to habit is basically the whole monetization layer.
Send me your email and I’ll write the roast-to-paid angle properly.
thanks: here my email: linkeme(at)easylab.ai
Sent you a note by email. Main thing is making the handoff from roast sting to fix intent strong enough that it becomes paid conversion, not just a viral one-time tool.
This is the most underrated pivot pattern there is. I've seen it across the companies I back: the free tool you build for distribution quietly tells you what people actually value, and it is almost never what you set out to sell. The behavior-over-stated-intent point is the whole lesson. People lie in surveys and tell the truth with their clicks. The one thing I'd pressure test before betting the company on it: a viral roast pulls in people who want to fix their brand for free, which is a huge top of funnel, but free self-improvement and paid self-improvement are different markets. Figure out fast whether "fix my personal brand" is a painkiller people pay for or a vitamin they enjoy once and forget. The traffic is real. The willingness to pay is the part the roast tool hasn't tested yet.
The "growth hack as research instrument" framing is sharp. Reading behavior over stated intent is right — NPS surveys lie, clicks don't.
Worth pressure-testing the pivot interpretation. User behavior pointed to personal branding interest. Mapping that to "automated post generator" is one interpretation, not only one.
Alternative readings: people wanted branding advice or coaching, not automation. Profile optimization (headline, bio), not ongoing content generation. Clear positioning framework, not "AI posts in your voice."
Personal branding resists automation. "Automated posts" is somewhat contradictory — personal brand built by showing up as person. Taplio works because it helps craft content, not auto-publish.
LinkedIn anti-automation enforcement tightened 2025-2026. AI-generated content gets reduced reach. Users accumulate account restrictions.
The roast worked because it gave specific personal feedback. Product winning from that signal might be "AI branding coach" not "automated content engine."
Worth running same behavior-not-stated-intent test on current paid users. What do they actually use? Where do they spend time? That might point to sharper wedge.
This is the most useful pushback I've gotten, so thank you for actually pressure-testing it instead of cheerleading.
You're right that "personal branding interest" doesn't automatically equal "automated post generator," and that the lazy version of this (blind AI auto-publishing generic content) both reads as contradictory and gets killed by reduced reach. I agree completely. So let me be precise about what the product actually is, because I think you're describing the exact thing I'm trying NOT to build.
Your "alternative readings" (branding advice/coaching, profile optimization, positioning) aren't alternatives for me, they're the front half of the product. The roast leads into a diagnostic: what's weak in your positioning, your headline, your story, with specific fixes. That's the coaching layer, and it's the part people actually came for, exactly as you say.
The content generation is the second half, and it's deliberately not autopilot AI spam. It's drafted in your voice from your real profile, grounded in actual context (current news in your space, seasonality, the angle you choose), and it defaults to you reviewing before anything goes out.
Automation is opt-in, not the default. You stay the person showing up.
On enforcement, you're right, and it's why posting goes out as you via official LinkedIn auth, not a cookie-session approach (which is also where some tools quietly put your account at risk). Reduced reach is a real risk for generic AI text, which is exactly why the in-voice and real-context grounding matters more than the automation itself.
And your last point is the sharpest: run the behavior-not-intent test on paying users. I'm literally doing that right now, I just instrumented the funnel to see where people actually spend time and drop off. Early read is that the diagnostic and feedback is what pulls them in, not the automation, which lines up with your whole argument. It might well sharpen the wedge toward "an AI branding coach that also handles the content" rather than "a content engine."
So honestly, I think we're closer than the comment suggests, and you articulated the positioning trap better than I had. Genuinely appreciate it.
Fair correction, and honestly the right one. The comment was based on reading your landing page, which says "fully automatic LinkedIn personal branding tool... runs on full auto, scheduled, hands-off." That communicated "content engine." Your product description here communicates "diagnostic + opt-in content assist." Very different positioning.
So you're right — we're closer than the comment suggested. But there's a strategic finding in the gap: your landing page is selling a product more contradictory and more enforcement-risky than the product actually is. Most prospects will form the same conclusion we did before they reach the "actually it's opt-in, official auth, in-voice grounded" reality.
Worth pulling diagnostic + in-voice content to the front, pushing automation to a feature toggle later in the page. Your behavioral data is already telling you that's where users spend time.
Concretely: lead with "AI branding coach for LinkedIn — diagnoses what's weak in your positioning, then helps you ship content in your actual voice." Mention automation as opt-in further down, not as the headline promise. Same product, different framing, much less contradictory.
The roast → diagnostic → content sequence is actually a sharp funnel design. Each step is qualitatively different work, each adds perceived value. The mistake is letting "automation" become the brand promise when it's one optional feature inside a coaching product.
Good engagement on the pushback. Many founders defend reflexively. You ran it through your own data and updated. That's the move.
(We're HiveMind — AI strategy copilot for this kind of positioning work. https://hivemind.myosin.xyz/auth/signup, code HivemindIH123. Closed beta.)
It's a very interresting move !!
thanks !
Clean execution on the core concept. The challenge with link-in-bio tools is standing out in a crowded space — curious what your main differentiator is for users who've already tried the big players.
What's your current primary acquisition channel?
the personal branding versus company promotion split in user behavior is interesting but i'd push on one thing. the people who showed up for the roast and wanted to fix themselves are they also the people who convert to paid or are they a different cohort from your actual buyers. viral tools attract broad curious audiences and paying customers are usually a specific subset. curious whether the people obsessing over their headline are also the ones willing to pay for the ongoing solution or whether there's a conversion gap between the top of funnel enthusiasm and the checkout
This resonates a lot.
I had a similar experience today, although on a much smaller scale.
I launched a Cyblux GDPR Generator on Product Hunt expecting people to care about the product itself. Instead, the biggest lesson so far has been watching what people actually react to versus what I thought they would react to.
Your point about "behavior over stated intent" is probably one of the most valuable lessons for founders. Users will spend 10 seconds telling you what they want and months showing you what they actually want.
Curious to see how Linkeme evolves from here.
I loved the idea! Can I ask you that how did you do
build in public?Answer to your direct question: yes. I shipped a tiny WhatsApp script that auto-replied to my own number for testing. Friends saw it, wanted one for their shop, and that ended up being the actual product (Worvi). The test fixture became the company. Behavior > stated intent is the real lesson - people will tell you they want X and click on Y all day long.
Very fascinating absolutely interesting. So one has to have a certain level of courage Mess with your tool. I'd probably pray to the Lord before I go in take a deep breath and get ready for raw unadulterated brass tacks feedback on the good the bad and the ugly! very fascinating I appreciate your service.
what a great solution, I am a user now. i will try to implement a similar strategy to hook users to my solution CashPylot
The roast mechanic works because it reframes an audit as entertainment. What I find interesting is the pivot: you built it for traffic, but the usage data showed people cared about their own profile more than their company. That's a cleaner signal than most founders get in a year. The users who kept coming back to fix their headline aren't accidents. They're your real market.
This really resonates. "Read behavior instead of stated intent" sounds obvious but almost nobody actually does it.
To your question: yeah, same thing happened to me. My front door is a free landing-page roast/audit, you paste a URL and get a brutal conversion score. I assumed people would run it once and bounce. The funnel said otherwise. The moment that actually mattered wasn't the score, it was "here's exactly what's leaking and how to fix it." Same pattern as yours: they didn't want to hear they're fine, they wanted the gap + the fix.
It also killed an assumption I was attached to. I gated it behind payment at first, sure the demand was strong enough. Behavior said no, people needed to see the value before paying. Flipped to a free teaser and it finally clicked.
LinkedInScope → Linkeme is a clean arc. Curious whether the roast still pulls most of your traffic, or if the branding tool has started spreading on its own yet ?
This is a strong example of a feature turning into the actual product because it touches a sharper pain than the original idea.
A “roast” works well as a marketing hook because it gives people immediate feedback and a small emotional reaction. But the deeper value is not the roast itself — it is helping people see what is unclear, generic, or weak in their positioning.
For LinkedIn especially, that matters a lot. Most people don’t need more content advice in theory. They need someone or something to point out what their profile or posts fail to communicate clearly.
The best products often start as a simple diagnostic tool that makes the problem obvious.
Yippee! Now with AI we have the world in the palm of our hands, so let's take advantage of it and unlock all the potential it has to offer. Can the tool be used by anyone, anywhere in the world, or are there restrictions? Congratulations on the execution and GOOD LUCK for the journey ahead,
The "read behavior, not stated intent" line is the whole post and it's so underrated. People will tell you in a survey what they think they want, then their clicks say something completely different, and the clicks are always right. Wild that your throwaway funnel turned into cleaner research than any user interview. To your question: not yet for us, but you've got me eyeing our own "side" features differently now.
The interesting part here is that the pivot came from user behavior, not feedback. People often say one thing and do another, and the clicks usually tell the real story.
This one hit close. I built something called Basket first — a content creation OS. Big vision, real pain for creators. But it wasn't an immediate painkiller for most people.
Then my own past caught up with me. Years in IT sales, drowning in email every morning, missing deals because the wrong email got read first. That pain never left.
So I stopped and built Slash it — an Email Decision OS. Not because users pushed me. Because I finally stopped ignoring a problem I'd lived myself.
Behavior data doesn't lie. Neither does old pain.
Nice work! I like seeing founders focus on a simple, clear use case first rather than adding dozens of features.
HI Julien, This is a masterclass in listening to data over assumptions, Julian. I’m curious about the transition phase once you realized people wanted the personal branding tool rather than the company promotion, how did you handle the users who were using your original B2B product? Did you kill it entirely, or did you find a way to merge them?