Monty.fast

Two weeks ago I posted my baseline: 164 Google impressions a month, 6 clicks, $0 revenue.

I spent the next two weeks submitting Monty to 15 startup directories. PostHog showed zero referral visits across 90 days. I was ready to call the whole exercise useless.

Then I checked the raw server log. It caught eight visits in its recent window: three from StackShare, three from Product Hunt, and two from Indie Hackers. Cookie consent had hidden every one of them from PostHog.

Eight visits from 15 listings is not a growth channel. But zero and eight teach different lessons. My analytics dashboard was not the source of truth.

While I was watching the referral column, Google impressions moved from 164 to 384, and 37 people joined the waitlist in 28 days, 24 of them confirmed. Revenue is still $0. I cannot honestly attribute any of that movement to directories.

The best question under my last post was: are users getting value in their first session?

Honest answer: I do not know yet. I have a working video pipeline and a waitlist. I do not have activation data.

So the next experiment is smaller and less comfortable: personally onboard 10 target users. One phone clip goes in. A publishable vertical video comes out. I watch every step.

Success means they get a usable result in their first session without a founder explaining the product. My next update will report how many of the 10 reach that point, and where the rest get stuck.

Directories are done for now. Conversations next.

Two weeks ago I posted my baseline: 164 Google impressions a month, 6 clicks, $0 revenue. Weekly update, same honesty.

Since July I've been filing Monty into every startup directory with a free listing: G2, SaaSHub, StackShare, SaaSWorthy, Crunchbase, SourceForge, Wellfound, and this one. Official goal: links and citations. Secret goal: visitors.

The number of the week: referral visits from all those directories combined, over 90 days of tracking, is zero. I checked twice.

What actually moved while I wasn't getting those clicks:

• Google impressions went from 164 to 384 a month. Non-brand queries appeared for the first time: "submagic alternative" at position 65, "best AI video editors 2026" at 91. Fame.

• Digging through analytics I found two visitors tagged utm_source=chatgpt.com. Someone asked ChatGPT, it browsed the web and handed them my site. Directories sent zero humans. A language model sent two.

• 8 people joined the waitlist this month.

Field note if you plan the same run: free AI directories are nearly extinct. Toolify wants $99, Aixploria $79, TopAI.tools $47, Easy With AI $125, OpenTools $199, Tekpon $249, Futurepedia up to $497. One of them estimates its own $49 listing brings 50-100 clicks in the first week. I'm pre-launch with a waitlist page, so I passed on all of them.

My takeaway: directories send no traffic, and they still earn their keep. Every listing is a link and an entity mention, Google and LLM browsing both pick those up, and that's exactly where the new impressions and the ChatGPT visits came from. If you're early: take the free listings for the citations, keep the money, and stop watching the referral column.

Revenue is still $0. See you next week.

I've been building an AI video agent: you throw in one clip you filmed on a phone, it writes the script in your voice, cuts the pauses and bad takes, times captions to your speech, adds music and b-roll, and posts the result to YouTube Shorts, Reels, TikTok and Telegram.

The engine works. It has produced a few hundred reels for my own channels. I do not open an editor anymore.

Here is the part I keep staring at. Last month Google showed my site 164 times and sent 6 clicks. Six. My blog has 55 topics in five languages, the markup is clean, the sitemap is fine. Nobody knows the thing exists.

So today I stopped building and spent the whole day on distribution instead: listed the product on eight directories, put 19 videos on a fresh YouTube channel, and wrote a comparison piece against the closest tool I could find. First time in months I shipped nothing to the codebase.

Two things I already learned the hard way:

1. One directory rejected me as a "generic AI video generator". I do not generate video, I edit footage people actually filmed. Wrong category is not a small thing, it decides which shelf you die on.

2. I nearly wrote "trained on a million videos" in my own comparison article. It is not true, so it went out. In a niche this noisy, the checkable claim is the only one worth making.

Next month I will post the same numbers again. If they do not move, the problem is me, not the algorithm.