This week I read a lot of honest numbers from small founders, and my favourite ones weren't the big ones. A Product Hunt launch with no audience, a free tool that quietly brings in sign-ups, and ChatGPT sending app downloads. Here's what worked, what didn't, and what I'd try this week.

What worked this week, in one minute

  • A Mac app's Product Hunt launch sent 4 visitors. Google sent 17 in the same stretch.
  • A free, no-signup tool has been used by more than 50,000 people and keeps feeding a paid product.
  • One iOS developer saw 59 downloads that came from ChatGPT, against 5 from Google.
  • Moving a sign-in screen later in an app stopped a 10 to 15% drop on the very first screen.
  • Five comments from strangers changed one product more than 100 sign-ups did.

Is a Product Hunt launch with no audience worth it?

Jeyanth, the developer of Textfold (a text expander for Mac), launched on Product Hunt on October 27 and then wrote up every number. The launch got 5 upvotes, sent 4 people to textfold.app and made 0 new sales. Over the same period Google sent 17 visitors and Indie Hackers sent 10. On X, Jeyanth summed it up: "the quiet, slow stuff beat the big day."

What I liked most is that there's no bitterness in it. The launch images and copy now live on the Textfold site and listings, so nothing was wasted. The Indie Hackers and X visitors came from replying to people, not from posting and hoping.

What I'd try this week: before you pick a launch day, spend an hour replying to real questions in the places your buyers hang out. Then check where your visitors actually came from. If you can, write one useful guide that answers a question people search for.

Source: Jeyanth's write-up on the Textfold blog (self-reported).

Do free tools actually bring in paying customers?

Sharifur Rahman shared how Taskip got its first paying customers, and the free tool part stood out to me. Taskip has a free quotation generator with no sign-up. The post says more than 50,000 people have used it, and about 2% of them go on to create a Taskip account.

The same post has the flops too, which I always appreciate. About $1,500 on Facebook and Google ads brought 40 sign-ups and 2 paying customers. 600 cold emails to agencies got 9 replies and zero customers. Emailing past buyers who already knew the team from earlier products worked much better: about 300 trials and 34 paying customers in the first two months.

What I'd try this week: pick the smallest annoying job your buyers do right before they need your product, and make a free page that does just that one thing. No account, no email gate. It's slow, like planting a tree, but it keeps working while you sleep.

Source: Sharifur Rahman on Indie Hackers (self-reported).

Can ChatGPT really send you customers?

Blaida (@kedytcom) posted a screenshot from App Store analytics: 94 impressions from ChatGPT, 59 app downloads and $112 in revenue. Google sent just 5 downloads in the same view. The post says $0 was spent on ads for that traffic.

That's one developer and one screenshot, so I wouldn't build a whole plan on it. It still made me go and ask ChatGPT a few questions about my own niche, just to see what it says.

On my own site, from Oct 26 to Nov 1, ChatGPT sent 20 visits (19 of them to my sounds page) and Google sent 3.

What I'd try this week: ask ChatGPT and other AI assistants the exact question a buyer would ask ("what's a good app for X?"). If you don't show up, make sure your site says plainly what your product does, who it's for and what it costs, in words a person would actually type.

Source: @kedytcom on X, Oct 25 (self-reported).

Where are new users dropping off?

Stacy (@stacy_siz) wrote a long, generous post about funnels. The app opened straight to a plain sign-in screen, which seemed fine, since people had already installed it. A consultant insisted on measuring it, and around 10 to 15% of people were dropping off on that first screen.

Sign-in moved inside the onboarding, and the numbers improved. Then it moved after the paywall, and they improved again. The post doesn't share the after numbers, but the lesson is easy to copy.

What I'd try this week: write down every step from "opened the app" to "paid", and count how many people reach each one. Then look hard at anything you ask for before people see the good part. Sign-in is usually the first suspect.

Source: @stacy_siz on X, Oct 23 (self-reported).

How do you find out why people don't stick around?

Youssef got 100 sign-ups for FollowToDM in 40 days, and almost none of them said anything. Then five strangers commented on Indie Hackers, and each comment changed what got built.

One tester asked whether it was worth trusting with a real Instagram account. That one sentence showed it was a trust problem, not a setup problem. Others pointed out that people signing up on a phone can't install a Chrome extension at all. Now they can email themselves the link for later. Shipped. Another commenter's advice about a "disconnect in one click" line was simple: only write it if it's true. It wasn't fully true yet, so the fix comes first.

What I'd try this week: post one specific question about your product in a community where people will be honest, and reply to every answer. Silence in your analytics can mean five different things. A stranger can tell you which one in two minutes.

Source: Youssef on Indie Hackers.

Is it harder to sell to other indie hackers?

Om Patel (@om_patel5) says the analysis pulled more than 7,800 startups with verified revenue and split them by who they sell to. In those numbers, 74.5% of tools sold to indie hackers and founders make $0 a month, against 60.5% of tools sold to small businesses and teams. Only 4.5% of indie hacker tools ever reach $1k MRR, and average MRR was about $530 against about $1,870.

The post starts with a story: someone paying $39 a month for an analytics tool cloned it instead and sold the clone for $29 lifetime, to the same crowd. That's a bit funny and a bit painful.

What I'd try this week: if your buyers are other builders, ask yourself honestly if a business with a budget has the same problem. I'm not changing anything yet, but I'm glad I read it.

Source: @om_patel5 on X, Oct 24 (the poster's own analysis; the post doesn't name the data source).

Fun little apps I found this week

TitleArt. davidbarker made a faithful copy of old-school WordArt that runs in the browser, rainbow gradients and all. It got 404 points on Hacker News, and the comments are full of people remembering school computer rooms. It even animates. Try TitleArt or read the Show HN thread.

plushies. prathje built squishy 3D plush characters for the web. Pick one of 17 shapes, like a heart, star or cloud, and it puffs up into a little pillow with a seam and fur. Some of them even work as cursors for AI helpers, which made me smile. Meet the plushies.

find the runners. Isabel von Stauffenberg is running the NYC Marathon with a sister and built a small site to track them both. Then it got opened up for everyone. You start a group, add runners by bib number, and share one link, no sign-up. Find the runners.

Taffy. PaweG calls it a dumb little site, and I love it. Paste some text, then drag the corner of the box. Pull it bigger and the text gets longer, push it smaller and you get a summary. Where was this when I had 500-word essays due? Stretch some text with Taffy.

That's the week!

A lot of this week's wins were small and quiet, and I think that's the good news. If someone on this list just got their first sale, I hope they stopped and celebrated it. I wrote about why that first sale feels so big if you want the long version. Mine get a pixel monkey who swings down and slaps a sticky note onto my screen. That's Slap Alert, the silly little thing I built.

Next week's roundup: what's working, week of Nov 5.

Happy building, friends.

Newt

More from my blog

Download Slap Alert free