What the March 2026 Core Update Means for Comparison Sites

Google's recent core updates, combined with its Scaled Content Abuse policy, reward content that adds genuinely new information and penalise templated, value-thin pages. Affiliate and comparison sites built on aggregated feature lists and data-swap templates have been most exposed; sites with original data, hands-on testing, named expert reviewers, and useful tools have held up better. This is qualitative analysis of a real policy direction, not a measured impact study, so it does not quote a specific percentage of sites affected.

What Google changed

The March 2026 update combined two major enforcement mechanisms:

Information Gain scoring

Information Gain measures whether a page adds new information to what already exists on a topic. If your page says the same things as the other 50 pages ranking for the same query, just reworded, it gets a low score. This penalises content that exists to rank rather than to inform.

Scaled Content Abuse enforcement

Scaled Content Abuse now covers any content produced at scale that does not add substantive value. This applies to AI-generated text and human templating alike. It includes human-written content produced by swapping data into templates. That describes the majority of affiliate comparison content on the web.

What died

### Content types most exposed under Google's scaled-content enforcement

| Content type | Why it was hit | Example |
| --- | --- | --- |
| Thin comparison listicles | No original evaluation, just aggregated feature lists | "10 Best CRM Tools 2026" with copied feature descriptions |
| AI-generated content dumps | Zero Information Gain - rephrased existing content at scale | Hundreds of city-specific pages with identical structures |
| Template + data swap | Same template with different data points inserted | "Best [X] in [City]" pages with only the city name changed |
| Affiliate review sites without testing | Reviews based on marketing materials, not hands-on use | "Full review of [Product]" based entirely on the product's website |
| Programmatic SEO at scale | Thousands of pages from a database with minimal editorial value | Auto-generated pages for every postcode/product combination |

Qualitative analysis of a policy direction, not a measured impact study.

What survived

The affiliate and comparison sites that maintained or grew traffic tend to share common characteristics:

• Proprietary data - original research, surveys, or testing data that cannot be found elsewhere
• Hands-on testing - genuine product reviews with screenshots, video evidence, and specific findings
• Interactive tools - calculators, comparison generators, and assessment tools that provide personalised value
• Expert attribution - named authors with verifiable expertise and Person schema
• Unique methodology - documented evaluation frameworks with transparent scoring criteria

Lessons for comparison sites

The comparison site model is not dead. But the low-effort version is. Comparison sites that survive after March 2026 need proprietary evaluation methods, original data, named expert reviewers, and interactive tools. The bar has moved from "aggregate and present" to "test, evaluate, and add genuine insight."

For any site in the comparison or review space (including this one), the survival checklist is:

1. Document your methodology - how do you evaluate? What criteria? How is scoring determined?
2. Generate proprietary data - run your own tests, surveys, or audits
3. Name your reviewers - attribute content to real people with verifiable expertise
4. Add interactive elements - tools that provide personalised value beyond static content
5. Update regularly - stale comparison content signals low editorial commitment
6. Show your working - screenshots, data tables, test results that prove original evaluation

The AI search connection

Sites hit by the March 2026 update will see compounding losses as AI search grows. AI platforms prefer to cite authoritative, original sources. That is the kind of content that survived the update. The sites that lost Google traffic are also the least likely to be cited by ChatGPT, Gemini, or Perplexity. The update and the AI search shift reinforce the same message. Original, expert, evidence-based content is the only viable long-term strategy.

See all AI search statistics · See our evaluation methodology