The most underrated feature for an early-stage product is a button to message the founder directly.
When I launched my first ad campaign for Todoza, I burned part of my budget for absolutely nothing.
People were coming to the website, registering, trying to connect their Amazon Ads accounts... and leaving.
I was wondering what went wrong. Was the product useless? Was the onboarding confusing? Were these simply the wrong people?
Of course, I tried emailing them. Zero replies.
Only when I decided to recreate the entire user journey myself did I discover the problem.
The main connection feature was BROKEN.
I'd literally been paying to send people to a locked door.
An expensive but useful lesson.
Since then, I've started tracking the important steps of the user journey. Where people click, where they get stuck, which pages they leave, and whether they actually come back.
And, of course, I added a small button to message me directly on Telegram - not just for bug reports.
Sometimes a user needs a tiny feature you could implement in an afternoon, but you'd never even thought about it. Sometimes they simply don't understand what they're supposed to do next. And sometimes they just want to know there's a real human behind the product. Nobody wants to open their email, write a support ticket, wait for a reply, then return to your app to explain the problem all over again. A quick chat removes a lot of that friction.
Especially when the person answering is the same idiot who broke the connection feature in the first place. 🍌
If you're building something new, make it ridiculously easy for your first users to talk to you.
Their feedback might be worth more than your next 10 features.
1 post. 249 impressions. 4 spam emails.
Scary to imagine what happens when a post gets 100k views 🥲.
But seriously, does this kind of AI-powered cold outreach actually work?
Let's go through my inbox.
1) Pricing on the Todoza homepage
First sentence gives me an AI-generated summary of my X post.
Second one - apparently, my homepage has a problem. No pricing in the navigation.
Third paragraph - an offer to fix it? What does that mean "quick fixes for homepage"?
What exactly are you selling me? What's the benefit?
I run multiple landing pages for different A/B tests. I don't even send paid traffic to my homepage.
A personalised email that completely misses the context. Nice.
2–3-4) THREE emails selling essentially the same thing: a Reddit scrapper.
Find relevant discussions, generate replies, get customers.
Sounds useful. But we're in 2026. Apify exists. Claude exists.
I've already built my own SEO automation, integrated with the official Amazon Ads API, and built systems processing millions of advertising records.
Another Reddit scraper isn't exactly what I'm missing 😁
Now, if you could show me how your tool brings actual paying customers, that would be a different conversation.
And that's what bothers me about these emails.
You've automated the research, the personalisation and probably the sending.
But somehow skipped the most important question:
Why would THIS particular person want to buy your product?
Maybe I'm completely wrong.
Maybe sending 10,000 emails at almost zero cost and converting 0.1% is a perfectly good business.
So I'm genuinely curious:
If you're running AI-powered cold outreach at scale, what conversion rates are you seeing?
Not opens. Not replies.
Actual paying customers.
Just curious. Maybe I chose a wrong niche?:)
Nope, that's related just to organic traffic via search engines. Other channels have good sing up / target action ratio, but users using that two-three times and not returning back. I just started collecting the user's journey and it may helps understanding what could go wrongly as soon as collect enough data. Anyway, that's interesting
My first 10,000 organic impressions.
Many builders has a pretty clear goal in the beginning of their journey - get the first paying customer.
I’m no exception. But to get one, I first need someone to connect their Amazon Ads account. Before that, they need to register. Before that, visit my website. And before all of that, they need to discover that my service even exists.
So I started at the very top of the funnel - becoming discoverable.
Getting organic traffic to a brand-new website is quite a challenge, especially if you don't want spending a fortune. Since Todoza had already been under construction for over a year (mostly because I was building it for myself), I figured I'd start working on SEO before the product was even ready. And the earlier, the better - organic search is a long game..
So I built a small AI-powered SEO engine - you feed it product information and a list of topics, and it generates pages, groups them into clusters and interlinks everything. Fully automatically.
You play games - it works.
You sleep - it works.
You call your mummy to cry because your super-cool revolutionary SaaS isn't taking off - it still works.
You get the idea, darling?
Anyway, three months later - 3112 pages indexed. 11193 organic impressions. 59 clicks. 0 signups.
Is that sad? Not really. Because you remember - organic search is a long game.
And I think the most interesting part isn't even the traffic itself. It's what you can learn from it.
Search queries give you an early glimpse into what people are struggling with.
In my case, sellers are searching for ways to apply negative keywords in bulk, get useful insights from Amazon Ads reports, and solve plenty of other annoying problems.
And now I have real search data to investigate instead of just guessing what people need.
How many startups end up pivoting because their original idea didn't match actual demand?
Alright, time to call my mum again.
Next post - let's dig into what those first 11k impressions actually taught me. 🍌
Oh yes! Actually, registration → connection works pretty well: 62% of registered users connect their Amazon Ads accounts.
The real challenge starts after that - getting them to come back and actually use the insights.
So I've started tracking WAU/MAU, where users click, what they look at, and where they lose interest.
Funny how understanding human psychology can be harder than building the actual software 🥲
@pcshipp I have UKs driving licence, and in front of my name there is "MR". Google decided it's part of my name and not matching with details provided. And month after still waiting for their help, getting tonnes emails and calls from their sales managers...
Ola Amigos,
Anyone here running OpenAi Ads, like me?
Native Ads manger is a bit castrated, but happily it has API.
Built a wrapper for my own needs, but don`t mind to share for free.
In opposite of native manager mine one -
🍌FASTER. (we sync data in the background every 5 minutes and serve it directly from our database, rather than waiting for multiple API requests every time you open a page)
🍌Copy ads between campaigns and ad groups
🍌Set custom ad schedules by day and hour
🍌See which days and hours actually perform best
🍌Manage everything from one dashboard
🍍SHITTY / NOT SHITTY campaign badge is coming.
Happy to give free access to anyone who wants to try it, just DM to me