NOAH: Nobody built this chain on purpose, that’s the first thing to see. The used-book shop’s been live for years, URLs changed, categories got reorganized, and each change added a redirect. Now old links pass through three or four hops before landing. Each hop was a reasonable fix at the time. They piled up. This isn’t one bad call, it’s the residue of many good ones nobody ever consolidated.
ELENA: Residue or not, it does real damage, and I want to be specific so it’s not abstract. Every hop is a separate request the browser and crawler make in sequence before reaching the page. Four hops is measurably slower than one.
NOAH: Slower I buy. Is it more than slow?
ELENA: It is, that’s the part people miss. Crawlers follow a limited number of hops and may quit before the final page, so the link signal from that old URL can fail to consolidate onto the live page. The chain doesn’t just drag, it can sever. Fix is mechanical, point every old URL straight at the final destination, one hop.
HANNAH: There’s a second problem tangled in here that the chain is hiding, the redirect type. Are these permanent or temporary?
MARCUS: Does it matter if we’re flattening them anyway?
HANNAH: It matters, Marcus, just not the way the old PageRank-dilution stories claim. A permanent redirect is a strong signal that the destination is now the canonical URL, so it’s the one that should consolidate into the index and own the standing. A temporary one tells search engines the original is still the real page and to keep it around. Mark a genuinely permanent move as temporary and the risk isn’t that signal evaporates, it’s that the old URL lingers as the canonical, the move stays ambiguous, and the destination is slower to take over. Modern engines often sort it out eventually, but you’ve made them guess instead of telling them.
MARCUS: So it’s not “use 301 or lose your rankings,” it’s “use the right type or leave the canonical ambiguous.”
HANNAH: Right. The dramatic version is outdated. The real cost is ambiguity, not lost equity.
SOFIA: And the visitor pays for the chain too, which a book shop should care about. Someone clicks an old link from a forum or an email, already half-impatient, and every hop adds a beat of delay before anything shows. On a slow connection that’s the difference between landing on the book and closing the tab. Flattening to one hop is a courtesy to the reader, not just a signal to the crawler.
MARCUS: Let me put a boundary on this before it becomes a cleanup crusade, though. Not every redirect is worth chasing.
NOAH: Meaning the dormant ones.
MARCUS: Right, a redirect nothing links to and nobody visits is doing no harm sitting there. The ones that matter are the chains on URLs that still get traffic or still have inbound links. Aim the fix, don’t scrub the whole map for its own sake. Friction where friction actually happens.
THEO: So the rule, concrete. Every redirect is a single permanent hop from old URL to final live destination, no stops. Two parts, flatten the chains so each old URL points directly at the current page, and confirm the type so genuinely permanent moves are marked permanent and signals consolidate. Prioritize by Marcus’s filter, what still has traffic or links, and you get most of the benefit from the smallest cleanup.
AIKO: And then stop the next chain forming, because flattening today doesn’t prevent tomorrow’s tower. The reason chains grow is each URL change redirects against whatever the URL currently is, including an existing redirect. So the standing rule is any new redirect points at the final destination, and when a URL that’s already a redirect target moves again, you update the originals to the new endpoint instead of stacking. There’s a sharper failure than a long chain, too, a loop, where A points to B and B points back to A and the page never resolves at all, which the same stacking habit causes. A crawler like Screaming Frog reports redirect chains and loops with the full hop path, so a periodic crawl catches both before they bite. The map stays flat as the catalog keeps reorganizing, which Noah’s right, it always will.
NOAH: So it’s both, the one-time flatten and the standing habit, redirect to the destination, never to another redirect.
DANA: Then here’s the call, two fixable problems plus a habit. We flatten the chains, every old URL with traffic or links points straight at its final destination in one hop, killing the latency and the risk crawlers quit before consolidating. We confirm the type, genuinely permanent moves marked permanent so the destination cleanly becomes the canonical, not left temporary where the old URL lingers and the move stays ambiguous. We prioritize by Marcus’s filter, what actually has traffic or links, not every dormant redirect, and we set the rule Aiko named, new redirects always target the final page, with a crawl catching chains over one hop. The chains weren’t a blunder, they were unconsolidated history, and the fix is to flatten them and stop them restacking.
MARCUS: That spends the effort where the friction actually is, instead of polishing redirects nobody ever hits.
DANA: A redirect should be one clean hop to the real page. Flatten the ones that still carry people, mark the permanent ones permanent, and don’t let the next reorganization build a new tower.