A 301 does not lose 15% of your link equity.

Where the 15% came from

The original PageRank paper included a damping factor, usually given as 0.85, describing the probability that a notional random surfer keeps clicking rather than stopping. That damping applies to links in general, as part of how the whole calculation works. It was never a redirect tax.

Somewhere along the way it got restated as "a 301 loses 15% of the value", and the number stuck because it was specific enough to sound researched. Google has since said plainly that 30x redirects do not lose PageRank. That covers 301, 302 and the rest: the redirect type no longer changes how signals pass, which also quietly killed the older habit of agonising over 301 versus 302 for equity reasons.

What actually costs you

The redirect is not the risk. The destination is. Four things do real damage, and none of them is a percentage shaved off in transit.

Redirecting to something that is not equivalent. Point a retired product URL at the homepage and Google frequently treats it as a soft 404, because the destination does not answer what the original did. You have not moved the signal, you have thrown it away and added a hop. Redirect to the genuine nearest equivalent, or return a clean 404 or 410 and accept the loss honestly.

Chains. A points to B points to C points to D. Each hop adds latency, and long chains risk being abandoned before the end. Chains also rot: someone eventually removes a middle link and the whole thing collapses to a 404 nobody notices for a year. Build the map one-to-one, from every old URL directly to its final destination.

An incomplete map. The most common cause of a genuine post-migration drop is not the redirect mechanism at all, it is URLs that were never mapped. This is the thing worth obsessing over, and it is covered properly in the migration guide.

Leaving everything else pointing at the old URL. Redirects are not a substitute for updating internal links, canonicals and sitemap entries. If those still name the old URL, you are routing every visitor and crawler through an unnecessary hop and sending a mixed signal about which URL you actually want.

Why the myth is expensive

Believing in the 15% leads to worse decisions, not more careful ones. People avoid consolidating two competing pages because they think merging costs them value, so the cannibalisation stays and both pages keep underperforming. People keep old URLs alive on thin duplicate content rather than redirecting them. People argue about redirect type in a migration meeting while the actual redirect map has forty gaps in it.

Redirects are the correct tool for a genuine move, and using them properly costs you nothing. In Search Console, page with redirect is not an error, it is a description of one working.

The general lesson

A precise number is not evidence, and 15% is the purest example of one that survived on specificity alone. Before optimising against a figure, check where it came from and whether the source ever said what it is now quoted as saying. In this case the answer is no, and the effort saved is better spent on the redirect map, which is where migrations are actually won and lost.