CRO & Analytics | Data → Growth | 130 projects, 0 fails. Daily AI & tech insights that actually matter.

Joined July 2022
31 Photos and videos
Pinned Tweet
Should governments allow AI companies to train on citizens' personal data?
9% Yes, with consent only
3% Yes, it's public anyway
81% No, never
6% Depends on the data type
1,241 votes • Final results
4
15
24,102
Rebuilt ConvRadar's landing page. Marked June 18 to check the funnel. Then ran the numbers: ~25 visitors a day, ~175 a week. Only a huge swing would register. I built a tool to diagnose funnels. Mine has no traffic to diagnose. #buildinpublic #indiehacker #solofounder
1
8
Every site I audit, the team wants to test the headline or the button color. Then we check mobile load time. Half of phone visitors leave before a slow page renders. You can't optimize a page most of your traffic never waits to see. #buildinpublic #indiehacker #solofounder
16
GA4 shows you where a funnel leaks. It never tells you why. Spent today buried in customer reviews, pulling the exact words people use for the problem. Analytics finds the leak. Their language tells you what to write to fix it. #buildinpublic #indiehacker #solofounder
3
28
The default move when sales slow down: buy more traffic. But if your funnel converts at 2%, more visitors won't help. You just pay to push a bigger crowd through the same holes. It's not a traffic problem. It's a plumbing problem. #buildinpublic #indiehacker #solofounder
1
1
17
Ran ConvRadar's checkup on ConvRadar today. Landing → signup: 1.7%. SaaS median is 2–5%. The conversion tool sits below the conversion benchmark. Also: every signup so far is desktop. Mobile: zero. Physician, heal thyself. #buildinpublic #indiehacker #solofounder
1
26
I don't do product recommendations often, but this one genuinely sped me up. If you build anything — code, docs, prompts — a dictation app is worth trying. Typing is the bottleneck you stop noticing until it's gone. 48K words in 25 days. I'm talky. And I speak English and Russian, it is rare when a dictation app can freely work with both of them. #willow #solofounder
9
The coupon code box is the quietest leak in checkout. Empty field, new tab, hunt for a code, land on a competitor, gone. The field meant to save the sale is the one that loses it. #buildinpublic #indiehacker #solofounder
1
2
34
They tell you to start with an unfair advantage. I don't have one. Except ten years raising conversion rates. I know how to make the number move. #buildinpublic #indiehacker #solofounder
1
18
Ten years of audits and the pattern holds: teams polish the checkout they have instead of removing the step that's killing it. ~24% of abandons trace to a forced account (Baymard). Guest checkout. Ship it before the redesign. #buildinpublic #indiehacker #solofounder
1
22
ConvRadar trial conversion: ~3%, up nearly a third month over month. Turtle pace. But every point walks me closer to flipping the paywall back on. At 10% the beta ends. Seven points to go. #buildinpublic #indiehacker #solofounder
20
Got my first real crash test today. UK e-com, millions of events a day, hundreds of thousands of landing pages. Half my reports won't run at that volume. Half the ingestion hangs. Nothing finds your ceiling like a customer too big for it. #BuildInPublic #SaaS #solopreneur
4
41
So now I rewrite all of it. Every report's architecture rebuilt to survive that volume. More stable, more flexible — the kind of work you only do when something forces you to. Costs me more to run, too. Paying that bill with a smile. #BuildInPublic #SaaS #solopreneur
34
Every audit, same trap: chasing one blended conversion rate. Mobile usually converts at a third of desktop. Blend them and the number describes no one — so you fix a page that was never the leak. Segment first. The leak has a device. #buildinpublic #indiehacker #solofounder
21
No commits, no demo calls, no client pings today. Three options: ship something tiny, write a useful post, or fake-busy. Picked option 4: admit it. Hardest part of solo building isn't the silent days. It's pretending they're loud. #buildinpublic #indiehacker #solofounder
2
25
What stops you from working like that? #buildinpublick #solopreneur
278
The most common funnel leak I see isn't checkout. It's the headline. The ad promises X, the landing page promises Y, and people bounce in 4 seconds. The hardest CRO fix is the one that needs copy, not code. #buildinpublic #indiehacker #solofounder
23
First Google Ads sign-up for ConvRadar. $500 and ~4 hours pruning negative keywords to get there. The channel still works. It just makes you earn it. #buildinpublic #SaaS #GoogleAds
25
4 test users. All happy with the product. All said the same thing in interviews: the landing page doesn't show what it actually does. Building something good and selling it are two different problems. Back to the copy. #buildinpublic #SaaS #CRO
2
1
53
Product Hunt launch flopped. But two people who saw it reached out on LinkedIn and gave me feedback I couldn't have paid for. Sometimes the thing that "didn't work" hands you what you actually needed. Building ConvRadar in public, one weird outcome at a time. #buildinpublic #conversion
2
57