I sent 150 cold emails over 6 weeks. Conversion rate: 0. Here's what I changed.
Six weeks in, I had sent 150 cold emails to potential users for Genie 007.
Opens: decent. Click rate on the link: okay. Sales: zero.
Not a single trial signup from cold email.
I sat there staring at the spreadsheet thinking the product was broken or the price was wrong or the market didn't exist. Classic post-failure spiral.
But then I actually read my emails back.
They were technically fine. Good subject lines. Clear CTA. Not spammy. But every single one started with me talking about me. "Hi, I built a voice AI tool for..."
Nobody cares.
The person reading it has their own problems. Their own to-do list. Their inbox is already full of people telling them what they built.
What I changed when I rewrote the sequence:
First, I stopped leading with the product and started leading with the pain. Instead of "I built Genie 007 to help you type less" it became "If you're spending more than 3 hours a day on email and chat, here's something worth 2 minutes."
Second, I got more specific about who I was writing to. My original list was basically anyone who might benefit from a productivity tool. That's too wide. I narrowed it to people who had publicly talked about RSI, hand pain, typing volume, or accessibility challenges.
Third, I removed the link from the first email entirely. Counter-intuitive. But the click-through wasn't the problem. The problem was they were clicking out of curiosity and leaving because there was no relationship yet. No link in email 1. Just a question.
Reply rate went from 1% to 11% in 3 weeks.
Still not a great conversion rate overall. But it gave me real conversations. And real conversations gave me product feedback I couldn't have gotten any other way.
The thing I'd do differently if I started again: validate the copy before sending to 150 people. Test on 20. Rewrite. Then scale. I was so eager to move fast that I burned through a third of my target list before I had a message worth sending.
Lesson: in outreach, speed is the enemy until you've found the message. Then speed is your best friend.
For anyone who's done B2B outreach from a cold start, how many iterations did it take before you found something that actually landed?
That shift from 'me' to 'the user's pain' is the exact moment a project becomes a business. 150 failed emails isn't a loss; it’s the tuition you paid to find a message that actually works.
Since you've finally found your 'pain-first' hook, you should put it to the test in the Validation Arena (tokyolore.com).
It’s a 30-day $19 entry sprint where the only goal is to see who can get the most real traction.
The prize pool is at $0 right now, so your new 11% reply rate could easily put you at the top of the leaderboard.
The grand prize is a trip to Tokyo! 🏆
Same pattern applies to paid ads. The "test before scaling" rule is one of the most consistently violated things I see in Meta ad accounts — founders spend $500/day on a creative they've never validated at $5/day.
What you described with email — pain-first opener, narrow list, no CTA in the first touch — maps almost exactly to what works in cold traffic Meta ads. The creative that stops the scroll is almost never "I built a tool for..." It's "Are you still doing [painful manual thing] by hand?"
The narrowing piece is underrated too. Most founders build lookalike audiences from their entire email list or all-time purchasers. That's too wide — you're including trial churns and one-off buyers alongside your best customers. A lookalike seeded from your top 20% LTV customers (180-day window, purchasers only) produces fundamentally different traffic.
The test-20-then-scale lesson took me seeing ad accounts burn $10K+ on the wrong angle to really believe. Speed is the enemy until the message fits. Then it's everything.
If you ever want to pressure-test your paid acquisition angle before committing budget, DM me — happy to take a look.
Test before scaling sounds obvious until you realize how expensive obvious is to ignore. I ran the same 150-email sequence to 3 different segments before I changed anything — same message, same offer, different ICP assumptions. Two segments got zero replies. One got 12%. The message wasn't the problem. The list was. Would have scaled the wrong thing indefinitely if I hadn't forced the comparison first.
That shift from product → pain is huge.
But what stood out to me is what happened after people replied.
It feels like there are actually two separate problems:
And the second one is where things still break.
I’ve been noticing that even when someone is interested, the moment you ask them to “try something” or “sign up”, they drop — not because they don’t care, but because it still requires a decision.
Curious — where did most people drop off for you after replying?
It really is. Most cold email fails are actually positioning fails in disguise. The product language assumes the reader already believes they have the problem. Pain language meets them before they've made that mental leap. Once we rewrote our subject lines around outcomes they were already losing sleep over — not features we thought were clever — reply rate went from 1% to 8% in one round. The product didn't change at all.
This is exactly the part that keeps coming up.
Getting someone to respond is almost a solved problem at this point.
But the moment you ask them to do something — try, sign up, change something — they pause.
What’s interesting is that it’s usually not lack of interest.
It’s that they still have to translate:
“this sounds relevant”
→ “this applies to me right now”
That tiny gap is where most drop-offs happen.
I’ve been seeing that if you remove that step entirely — and show them their own situation directly — the need to “decide” almost disappears.
Curious if you saw a specific moment where that hesitation kicked in?
The hesitation usually kicks in right after the first call. They liked the conversation, they see the value conceptually, but then they go back to their desk and the urgency disappears. That is the "this applies to me right now" gap you are describing.
What changed it for us was showing them their own data in the pitch. Not a generic demo. Their numbers, their funnel, their problem. When someone sees their own situation on screen, the translation step you mentioned just vanishes. They do not need to imagine how it applies because they are looking at it.
That’s really interesting — especially the part where the “translation step” disappears.
Feels like that might be the core issue in a lot of these cases:
people don’t act because they still have to mentally map
“this is useful”
→ “this applies to me right now”
And that mapping step is where hesitation creeps in.
When you show their own data, that step just… isn’t there anymore.
Curious — did you see that also affect how fast they made a decision,
or mostly whether they acted at all?
That “technically fine but no one cares” point is the killer.
We’ve seen the same thing where the problem isn’t delivery, it’s that the message doesn’t immediately connect to something the person already feels.
Narrowing the audience usually does more than tweaking the copy. Once it’s specific enough, the message almost writes itself.
That last line is the bit most people miss. When I was writing to "anyone who might use a productivity tool" the emails took ages to write and still felt generic. The moment I narrowed to people who had publicly mentioned wrist pain or typing fatigue, I barely had to think about the copy. Their own words became the email. Went from spending 20 minutes per email to maybe 5. The specificity does the selling for you.
Not yet — but you're pointing at the missing piece. The Loom shift reduced friction on understanding, but what you're describing is getting them to feel it on their own data rather than watching mine. I've been thinking about a 'paste your own email and see it rewritten' demo. Maybe 10 minutes of build time and probably doubles conversion. Going to test it this week.
The 40 cold emails to 0 replies experience is what pushed me away from cold outreach entirely — same pattern you described, leading with the product instead of the problem. The 'no link in email 1' approach makes sense for the same reason the first DM shouldn't pitch: relationship before request. The narrowing of the list is probably the biggest lever though. 'Anyone who might benefit' is not a list, it's a hope.
You nailed it. "Anyone who might benefit is not a list, it's a hope" is exactly the trap I fell into. The moment I stopped writing to a category and started writing to one specific person with one specific problem, everything changed. Cold outreach isn't dead. Lazy targeting is.
Lazy targeting' — that's the right diagnosis. The list isn't the problem, the assumption behind the list is.
Right. The list is just a reflection of whoever you assumed your customer was. That assumption bakes itself into every part of the message from the subject line to the CTA. Fixing targeting doesn't just improve reply rates, it changes the whole emotional frame of the outreach. When you stop trying to convince people they might have a problem and start finding the ones who already know they do, the conversion logic completely changes.
This is a really solid breakdown-especially the shift from product--pain.
One thing I’ve noticed working with this kind of outreach:
even when reply rates go up, it doesn’t always translate into conversion.
Because there are actually 2 separate problems:
-getting attention (you fixed this)
-making the value click instantly once they engage
A lot of founders fix the first one and think they’re done, but the second one is where most revenue is lost.
The interesting part is:
if someone replies, they’re already “in”.
But if the product doesn’t match the expectation created by the message, the conversion still dies.
Curious, once people replied, where did the drop-off happen?
The drop-off for me was almost always at the trial stage. The email created an expectation about solving one specific pain, but the trial threw them into a generic onboarding flow that had nothing to do with that pain. The message and the experience were mismatched. What moved the needle was skipping the product tour entirely and showing them the exact thing the email mentioned within the first 60 seconds of signing up. Tiny change, meaningful difference in trial-to-paid.
Good question. Most of the drop-off happened between the first reply and actually signing up for the trial. The conversation was warm but the jump to "try it" was too big a step. What worked better was sending a 2-minute Loom showing the specific use case they mentioned in their reply, before asking for anything. Shorter path, less commitment required. The product-message match you're describing is exactly right. The message got them interested in solving a specific problem, but the trial started with a generic walkthrough instead of showing that same problem solved in 90 seconds. Once I fixed that the trial-to-paid rate moved.
That Loom shift makes a lot of sense-you basically collapsed the gap between “interest” and “seeing it work”.
One thing I’ve seen at that exact stage though:
-Loom fixes understanding -but not necessarily commitment
the user sees “this solves my problem”- but still stays in observer mode
What tends to push conversion further is: letting them experience their own case instead of watching it
so instead of: “here’s how it works for you”
it becomes: “try it on your exact case right now (no setup)”
That shift usually moves people from: “this looks good” to “I’m already using this”
which shortens the jump to trial a lot.
Curious-did you try giving them a zero-setup way to run their exact use case themselves?
That's exactly what I'm working towards now. The Loom got them nodding but you're right, watching isn't doing. I'm testing a flow where the first thing they see after replying is their own use case running live, no account needed, no onboarding wall. Early results are better. The "try it on your exact case" framing is the right one. Appreciate the push on this, it's sharpening how I think about the funnel.
This is such a good breakdown.
What’s interesting is that both steps (Loom-live use case) are solving the same underlying problem: people don’t trust their own interpretation fast enough
Loom says: “here’s proof it works”
Live demo says: “it works for you”
But the real unlock I’ve been seeing is one layer deeper: it’s not just about showing the solution-it’s about removing the need to decide
The highest converting flows I’ve seen feel like:
– no signup
– no setup
– no “start trial” moment
You’re already in it. That’s where the shift happens: from evaluating-to using
Curious if you noticed the same: once people interact with their own case, the “trial” step almost becomes irrelevant?
Mostly content strategy for SEO right now. Which keywords to go after, whether a page needs a rewrite or a new piece entirely, how to structure for intent vs. just volume. The pattern I keep seeing is that the people getting real value aren't asking 'what should I write' but 'why isn't this ranking.' That question is what actually exposes the gap between what they think searchers want and what Google is rewarding. Your point about semi-manual being intentional resonates. The personalization layer is where the product earns trust before automation can safely take over.
That’s a strong use case.
“What should I write” vs “why isn’t this ranking” is exactly that shift from surface-level to actual leverage.
At that point it’s no longer content. It’s a diagnosis problem.
And once someone sees the gap on their own page, they don’t really need convincing anymore-the action becomes obvious.
We’re seeing a similar pattern with stores. Generic advice doesn’t move anything.
But when you show someone exactly where they’re losing revenue on their own site, the “should I fix this?” question disappears.
That’s where the trust gets built before any product or trial step.
Out of curiosity-are you showing them this inside the product already, or more through manual walkthroughs right now?
100% seeing this. The moment someone uses it on their own words, the whole "should I try this?" question disappears. It just becomes "oh, this works." The biggest drop-off I was getting was between "sounds interesting" and "let me sign up to find out." Removing that gap changed everything. Now I am testing a flow where someone replies to a message and the next thing they see is their own text, already processed. No account. No decision point. Just the result. Early numbers are way better than anything the old funnel produced.
That “removing the need to decide” framing is sharp.
What I’m starting to see is that most funnels are built around forcing a decision too early: sign up, start trial, explore
But the user is still trying to answer: “does this even apply to me?”
So the friction isn’t really in the product. It’s in that gap between understanding and belief. Once they see their own case processed, there’s nothing left to evaluate.
At that point, the “trial” step almost feels artificial.
I’m curious how far this can go - whether the best funnel is actually no funnel at all,
just progressive exposure to value.
The understanding-to-belief gap is the real problem. I started noticing the same pattern - people who saw what the tool did on their specific workflow converted at 3x the rate of people who got a generic demo. The decision had already been made before they ever hit 'start trial'. Your framing maps to this exactly: once they see their own case, the funnel step becomes a formality. Curious what you're using to build those personalized flows - is it automated or still manual at this stage?
Exactly-that’s the shift. Once someone sees it applied to their own context, the decision is already made. Everything after that is just friction.
Right now it’s mostly semi-manual.
Not because it has to be, but because that layer is where the actual value is. The biggest impact doesn’t come from the analysis itself, it comes from how specifically it connects to the user’s situation.
That’s what creates the belief moment.
If you automate too early, you usually end up with outputs that are technically correct… but don’t feel relevant enough to trigger a decision.
Automation can scale it later.
But first you need to understand what actually makes someone go from “this is interesting” to “this applies to me”.
Curious: what kind of decisions are people actually running through your engine right now?
This comment was deleted 5 months ago