I built a directory, probably four years too late, but I don't care.

I can already hear you. Who, in their right mind, still builds a directory in 2026?

Fair question, because I did exactly that. And I also built this a month ago, before I decided to focus entirely on one app in August.

Directories were hot in the early 2000s. They saw a new surge around 2018, especially the niche ones. Then 2021 hit, and things exploded. Indie hackers and devs building in public, moving a hundred miles an hour, shipping something new every week.

2022 was the real boom. No-code tool directories. B2B directories. Newsletter directories. The list never ended. And then X and Reddit went completely bonkers when OpenAI released ChatGPT.

The AI directory gold rush

That release created an explosion of AI directories. Hundreds, maybe thousands, appeared almost overnight. You know the ones with "Best AI Tools," "AI for Designers," all of it.

People realized the playbook was simple. Scrape some data, buy a directory template, deploy it in a day, then monetize through affiliate links, featured listings, or sponsored placements.

Many of them were pulling thousands of organic visitors a month, because competition was still low and Google hadn't caught up yet.

Then it went downhill. Around 2023 and 2024, the whole thing started to collapse under its own weight.

Google got much better at spotting thin, low-value directory content. Thousands of near-identical directories flooded the web. AI made it trivial to generate directory pages, which further dragged down quality.

People started asking their AI assistant instead of browsing a list of links. And SEO got more competitive, more expensive, harder to win.

Directories aren't dead. They just changed. I could go on about that for a while, but that's not the point here.

So why did I build one anyway?

Call me crazy, but there are more than one reason.

The first is Kiroka. I built a free subscription tracker for people, and I want to keep it free forever. To make that work, I added a small, handpicked, curated directory of things most people already use and pay for.

That got me thinking. I'd been planning to position Ryoka through small experiments and micro-apps, and I was already figuring out how to feed entries into Kiroka's little directory.

Since I'd built my own OS dashboard (RyokaOS), it was natural to add the entries there and push them straight to the subscription tracker.

It didn't take long to see I could kill two birds with one stone.

• Wire a normal, standalone directory into the same setup

• Add an entry once, and decide whether it shows up on both, on one, or on neither.

My workload didn't go up. It got easier. More manageable. And it gave me another shot at building a better position as a content creator and builder, and another way to plant seeds with Ryoka.

That directory is Harova. It lives at harova.xyz.

Keeping it simple

Harova is simple. No fluff. I plan to add two or three entries a day, all picked by hand. Around 90% neutral picks, maybe 2% my own products and builds, and the rest affiliate.

I'm not in the mood to mess around with paid placements or advertising. That's not the goal, not yet. And let's be honest, who the fuck cares about a brand new directory with no traffic? Nobody.

That's fine, because it's not really about today.

The longer game

For me, Harova is one piece of a bigger playbook. And sure, I might be four years too late to the party. But look at the ones who rose back then. Most of them fell into decay, or just quietly vanished.

Here's the difference. Back then, I wasn't familiar with coding. No vibe coding, no boilerplate templates that spin up a directory in an afternoon. I couldn't have built this even if I'd wanted to.

Today the game has changed, even for me. I'm capable now, and I'm confident, and that's really the whole reason Harova exists. Not because the timing is right. It isn't. I built it because I finally can, and because it slots quietly into everything else I'm already doing.

Being late doesn't bother me much. I'd rather be four years late and actually build the thing than keep waiting for a window that already closed.