Whether you can launch a site with thousands of pages was the question a listener put to David Quaid of Primary Position on Edward Sturm's show. I build large sets of pages for local service businesses, so I listened closely to how the two of them talked it through. You can watch the full episode with Quaid and Sturm. Most of it is about marketplaces and database sites, not plumbers or remodelers. A lot of it still applies.

What the listener asked

The listener wanted to launch a database site, the kind that lists businesses, car models or buildings, with 10,000 to 100,000 web addresses. He said he and his friends had tried it several times. Each time, traffic shot up and then dropped to zero. Sturm said he put one of those sites into Ahrefs and saw that shape: straight up, then straight down.

That is one listener and a few sites. I still take it seriously, because it is the exact risk of putting many pages up at once.

Quaid's first answer: go very slowly

Quaid's first answer was to launch a large site "very slowly." His steps:

  • Run a holding page for a couple of weeks before the pages go up. A "we're building soon" page, a meet the team page, an about page. In his words: "I wouldn't want to buy the domain and in an hour launch a CMS that is capable of mass-producing so many pages."
  • Build up to about 100 pages by hand before automating anything.
  • Go in phases and be "ultra cautious at the start."

He called 100,000 pages in a single go a "definite blacklist." He also said that once a site has about 100 pages indexed, "it's nothing to launch a thousand pages." The captions do not make clear whether he meant a thousand is safe or only easy to do, so I would not lean on that line.

His final answer was more open than his first. Maybe 20,000 pages at once can work "under the right conditions." Then he added: "Could a normal person do it? I don't know."

Too slow has a cost too

This part surprised me. Quaid said publishing many pages together lifts traffic. Build them one at a time and none of them get enough clicks, so they drop. "You'd almost have to be so slow that you're not getting the sort of cumulative effect of publishing so many pages in a go."

So he looks for a middle path between 100,000 pages in one go and slow, hand built pages.

Sturm added the check he runs before he builds more. Look at the bottom of funnel pages you already have and "double check, triple check, quadruple check that it's going to satisfy search intent. And then we can build more." Quaid agreed. Sturm calls himself "slow and steady."

What gets a big site hit

Quaid was clearer about what goes wrong than about safe numbers.

Empty pages. If a new marketplace has 10 listings, "why are you launching with 25,000 URLs?" Sending people to empty pages, he said, "is basically you're building for ranking."

Crossing lists. Category times industry times geography, built out to "80,000 or 800,000 pages," reads as search manipulation. He said these sites get struck down, sometimes after a few months.

Pages that all look the same. If 99% of pages are "exactly the same," he called that "a very very strong signal." Even if half the pages have real data, the empty half "could be classed as doorway pages."

A real business can still be hit. A friend of his ran a marketplace in New York and New Jersey that made money. It crossed its marketplace with places. About two years before the episode it lost "90% 99%" of its traffic. Quaid reads it as doorway pages: empty pages "rolled out to match search phrases."

Pages for places where you are not. A client of his built location pages. They failed because "people were looking for like a physical location" and same day delivery. Sturm called it "search intent mismatch basically."

He also said not to over-optimize a big launch. No "ultra optimized" internal linking or titles at the start. "I'd almost want to dumb it down."

What this means for a local service business

A local business site with a page for each service in each town is a crossing too: service times town. So Quaid's warnings apply. This is how I read them for an owner.

  • Each town page has to say something true about that town. If the only thing that changes from page to page is the town name, it is the cookie cutter page Quaid describes.
  • Do not build pages for towns you do not serve. When people want someone close by, a page for a place you never go is the mismatch his client hit.
  • Quaid thinks signs of a real business help protect a big launch, and that a Google reviewer would look for social profiles, press, people searching your name and paid traffic. He offered that as his guess. Sturm put it as "authority and user signals," and Quaid agreed. Neither showed data.

If traffic does drop, Quaid's advice is practical. In Search Console, compare equal windows before and after the drop. Find the pages that lost the most clicks, then look at click through rate for each search. Take out words that mislead, such as "free" on something that is not free, even if those words bring traffic. And before you blame AI Overviews, check which of your searches even show one. He thinks many people who blamed them were marked down instead.

How we use this at eezyRank

We build many pages for local service businesses, so this episode reads like a checklist for us.

  • Pages stay hidden from Google while we build them. They open to search at go live.
  • We choose which service and town pairs get pages. A pair that makes no sense gets no page.
  • We build one service first and read those pages before we build the rest.
  • We build pages only for towns the client really serves. Each town's pages sit under a town page people can reach from the menu.
  • After launch we read Search Console to see which searches each page shows up for.

We do not promise that a big set of pages will all rank, or that early results will hold through the next Google update. Quaid's stories are the reason.

See what eezyRank builds at https://eezyrank.com.