A 5x publishing increase cost one site 1,000 indexed pages and every impression

The developer behind IELTS Writing Checker raised his programmatic publishing rate from roughly 200 pages a day to 1,000, and shortly afterwards Google Search Console reported zero impressions and zero clicks for the latest day ^[8][9][1]. Over the same stretch the approximate number of pages Google had indexed from the site fell from about 3,500 to about 2,500, according to his write-up on dev.to ^[2]. That is a net loss of roughly 1,000 pages, about 29 percent of the peak footprint, while the site itself was several times larger than before ^[11][20].

The setup matters because it looked healthy. The site sells tools and practice content around the IELTS English proficiency exam, including a Task 2 essay checker, with pages generated programmatically around writing practice ^[3][4][7]. In its first month, Google traffic kept climbing, Ahrefs Domain Rating hit 25, and new pages were often indexed within hours of publication ^[5]. He read fast indexing as evidence of trust, and that reading made him impatient ^[6][19].

Two changes landed at nearly the same time. The publishing rate went from 200 to 1,000 pages a day, a fivefold increase ^[9][21]. Separately, he changed one of the rules the site used to organise its pages, which unexpectedly generated several thousand additional URLs ^[10]. To put the intended output in context, 1,000 pages a day is around 30,000 URLs a month, roughly 8.6 times the entire indexed footprint the site had at its peak ^[22].

The leading indicator was visible before the traffic loss, and it was not a ranking metric. The site was adding thousands of URLs while Google's indexed count refused to follow, and the gap between published and indexed pages widened ^[11][12]. He interpreted the gap as processing lag and kept publishing ^[12]. A few days later Google traffic began to slide, then kept sliding, and then the impressions went to zero ^[13]. By his reading, Google had moved from declining to index new pages to dropping pages it had already indexed ^[14].

He is careful about what he can prove, and the caveats are worth repeating: no manual action notification arrived, approximate index counts are not perfectly reliable, and Search Console data can lag ^[15]. His claim is narrower than a penalty diagnosis. The site lost most of its Google visibility shortly after he dramatically increased its URL count ^[16].

The operational lesson is the confounding, not the volume alone. Because both changes shipped together, he cannot say which did more damage, which is the part he says he regrets most ^[17]. A site with 3,500 indexed pages and a five-day-old spike has no way to isolate a URL-taxonomy bug from a publishing-rate change after the fact. The other lesson is about the assumption underneath the strategy: he believed each page was genuinely useful and better than the alternatives in the niche, and concluded from that Google would tolerate the rate ^[18]. Per-page quality and site-level growth rate are judged separately.

What to watch, if you run generated inventory: the ratio of indexed pages to published pages, tracked daily, treated as a hard stop rather than a lagging report. A widening gap is the signal that arrives before impressions move ^[11][12][13]. And ship template rules and rate changes on separate days, or accept that you will not know which one broke it ^[17].