How this index is built
No mystique, just the criteria.
1. Why rank by growth instead of traffic
Sort by traffic and the top twenty is always the same: ChatGPT, Canva, Midjourney. That ranking is useless to anyone who builds sites, because nobody is going to build another ChatGPT.
The useful signal is the site that nobody had heard of last month and that tripled its traffic this month. What did those sites do in common? Which acquisition channel did they use? Which narrow need did they carve out? That is what can actually be learned. So this index has exactly one sorting rule: month-over-month growth, descending.
2. Where the numbers come from
Domains, descriptions, categories, monthly visits and growth figures come from public third-party traffic estimates and publicly available site information, then each entry is checked by hand to confirm the domain resolves and the description matches what the tool actually does.
- Monthly visits is a third-party estimate rounded to K or M. It is not exact. The same site can differ by 20 to 50 percent between providers, so read it as an order of magnitude, not a measurement.
- Growth is the month-over-month percentage change in visits. Sites starting from a very small base — 30 visits becoming 3,000 — produce absurd percentages, so everything above a threshold is capped at +9900%.
- Registration date, organic search share and domain rating are reserved fields. Where they have not been filled in per site, they are left blank rather than guessed.
One thing has to be said plainly: these are estimates, not audited data. They are good for reading trend and magnitude, and unsuitable for any financial decision.
3. Why only nineteen sites
Because every entry is verified by hand, and a short honest list beats a long padded one. An entry has to satisfy all three: the domain resolves, the product does one clearly defined thing rather than everything, and it grew noticeably in the past month. The list will expand gradually; only at a few hundred entries does an index like this start to be statistically meaningful.