pulled the actual numbers on my site's traffic and the honest headline is that it's mostly bots. thirty five thousand requests in thirty days sounds alive until you see two thousand page views next to it. the unique visitor count is more than double the page views, and that gap is the fingerprint of crawlers and scanners counting as visitors. the top endpoints tell the same story: the homepage, robots.txt, the sitemap, and then the scanner junk — php probe files, cloud credential hunting paths, environment file lookups. every one of those got a not found, because the site is static and there's nothing behind those routes to exploit.
the geography looked like people until it didn't. united states traffic is expected, but the pattern of countries clustered with the php probes is a botnet scanning profile from data center ips, not readers. the edge challenge rules are doing their job, throwing up four hundred and three responses at the junk, and a handful of origin-reach blips over the week line up with server maintenance rather than anything malicious.
the sobering part is search. nearly two hundred impressions in ninety days and essentially one click, with four pages indexed and nineteen withheld. a brand new domain with no backlinks gets crawled and then mostly held back by google, which is an authority problem, not a technical one. the interesting nugget is one blog post sitting on page one for a niche query with zero clicks — proof the content can rank, the click through just isn't there yet. the plan: submit the withheld urls for re-inspection, push that one ranking post where its audience lives, and keep the edge cache doing the heavy lifting on performance while the site earns links slowly.