The Helpful Content Update — rolled out in waves from August 2022 through the core update integration in March 2024 — was the most significant single algorithm event for affiliate publishers in years. Sites that had built traffic on thin product reviews, comparison tables with no original analysis, and SEO-first writing saw rankings collapse. Some never recovered.
That’s the context. What’s more useful now is understanding what the recovery data shows about what actually works for affiliate content in 2025.
What Got Hit and Why
The pattern of sites hit by HCU and subsequent core updates is instructive. The consistent factors:
Product coverage without product experience. Review sites that listed specs sourced from manufacturer pages, wrote generic pros and cons, and had no evidence of hands-on testing were systematically demoted. Google’s quality rater guidelines are explicit about this: E-E-A-T requires demonstrated experience, not just expertise. For affiliate content, that means showing you’ve actually used the thing.
Thin at scale. Sites with thousands of programmatically generated product pages — “best X under £Y” lists with no differentiation, templated review structures with thin unique content — were penalised regardless of domain authority. The penalty was sitewide, not page-level. That distinction matters: a sitewide HCU penalty isn’t reversed by improving individual pages. It requires improving the overall quality signal across the site.
Excessive affiliate density. Sites where the primary purpose of pages was clearly affiliate monetisation, with content serving as context for links rather than value in its own right, were assessed differently from sites where affiliate links were incidental to genuine editorial coverage. The ratio of monetisation to genuine content value became a signal.
What Recovery Actually Looked Like
Sites that recovered after HCU penalties shared several characteristics:
They pruned aggressively. The most consistent recovery pattern was removing or noindexing thin, duplicative, or low-quality pages — often 30–50% of total URLs — and seeing sitewide quality signals improve as a result. The counterintuitive insight is that fewer, better pages outperform more pages with variable quality.
They added original research and testing data. Sites that recovered fastest added content with no equivalent elsewhere: original test results, comparative data from physical testing, real-world measurements. This differentiation from manufacturer specs is the clearest signal of first-hand experience Google can evaluate.
They reduced affiliate link density. Cutting affiliate links per page, adding nofollow to lower-priority links, and improving the content-to-monetisation ratio were consistent recovery steps. Affiliate link density has never been an explicit ranking factor, but the correlation between high density and HCU penalty is strong enough that it’s a reasonable proxy.
They built topical authority, not just product coverage. The affiliate sites that came back strongest weren’t just review sites — they became destination sites for their category. Sites covering running gear that added training advice, race coverage, and community content recovered faster than pure review sites. Category authority matters more than product coverage depth.
What This Means for New Affiliate Content
The practical implications for affiliate content built today:
Write from experience, or commission content from people who have the experience. Generic comparative content produced at scale is not a viable long-term affiliate SEO strategy. It was marginal before HCU and it’s negative-value now.
Be genuinely useful first. Affiliate revenue should follow from being the best resource for a topic, not precede it. Sites where editorial decisions are driven by affiliate potential rather than audience need are producing content that Google is increasingly good at identifying.
Invest in differentiation. Original data, real testing, unique insights, first-person experience — these are the things that AI can’t replicate and that genuinely help users make decisions. That’s a narrow lane for pure-play affiliate content, but it’s the lane that survives algorithm updates.
The HCU didn’t change what good affiliate content looks like. It just raised the penalty for content that looks like it was built for search engines rather than for people who need to make purchasing decisions.