BabyPenguin.AI

"Best of" pages are the workhorses of commercial AI search. When a user asks ChatGPT, Perplexity, or Gemini "what's the best CRM?" or "which AI visibility tool should I use?", the engine doesn't go scrambling through individual product pages, it reaches for ranked comparison content that already exists. The brands that show up in those answers are the brands that appear in well-structured "best of" listicles, written by anyone, anywhere.

This is one of the highest-leverage GEO formats in any commercial category, and it's one most teams build incorrectly. Here's how to build a "best of" page AI engines actually use, and a critical second tactic that matters even more.

Build the page as a list, not as an article

The first structural decision is the format. "Best of" content gets cited at far higher rates when it's structured as an explicit ranked list rather than as a flowing essay. The GEO research consensus is direct: "Include more lists" in your content, and add "summaries of long text paragraphs, lists of reviews, tables, videos, and other types of difficult-to-cite content formats."

For "best of" content, that means:

One numbered entry per option, formatted as an H2 or H3
The option's name, ranking position, and one-line summary in the heading
A consistent scaffolding under each entry, not a different format for every option
An overview comparison table at the top of the page

The numbered structure is what AI extractors latch onto. Each numbered entry is its own self-contained chunk the model can pull, and the ranking position is itself a signal the AI uses when constructing recommendation answers. A "best of" page written as 2,000 words of essay-style discussion gets cited at a fraction of the rate of the same content broken into 10 numbered entries.

Add pros and cons for every option

One tactic that consistently improves citation rates: every entry in a "best of" list should include explicit pros and cons. As the GEO citation research notes, to be cited, content benefits from "pros/cons" alongside reviews, expert quotes, and quantitative data.

The format that works:

1. [Tool Name], Best for [primary use case]

Pricing: Plans start at $X/month.

[Two-sentence overview of what the tool is and who it's for.]

Pros:

[Specific strength with concrete detail]

[Another specific strength]

[Another specific strength]

Cons:

[Specific weakness, honestly stated]

[Another specific weakness]

Verdict: [One sentence on who should pick this option.]

This scaffolding does two things simultaneously. It gives AI extractors a predictable set of fields to pull for each option, which dramatically improves the AI's ability to compare options across the page. And it signals editorial neutrality. The cons section is what tells AI engines that the listicle is honest analysis rather than a marketing piece. Listicles without cons read as advertising and get downgraded.

Lead with a comparison table at the top

Above the numbered entries, place a single comparison table summarizing all the options on 3-5 key dimensions. This table becomes the most-extracted element on the page, because tables are pre-structured for AI parsing in a way prose isn't.

The columns that work for most "best of" tables:

Tool / Option name (linked to the detailed entry below)
Best-for descriptor (one short phrase)
Starting price
Key differentiator
Free trial / free tier availability

One row per option. Sort the rows in the same order as the numbered entries below. AI engines often pull this table as their primary source for "compare X" prompts, while pulling individual entries for prompts about specific options.

Include the right kinds of evidence

"Best of" pages that get cited consistently combine four types of evidence:

Quantitative data, pricing, ratings, review counts, performance benchmarks
User reviews, quoted from real platforms (G2, Capterra, Reddit) with attribution
Expert quotes, from named practitioners or analysts in the space
Direct experience signals, "we tested X for Y weeks" or "we used X to do Z"

Pages with all four are dramatically more citable than pages with only the writer's opinion. Each piece of evidence is its own potential extraction unit, and the combination establishes credibility that single-source listicles can't match.

Stay neutral, even when ranking your own product

The single biggest mistake teams make on their own "best of" pages is making themselves the obvious #1, dunking on competitors, and burying anything inconvenient. This pattern is so common that AI engines have learned to filter for it, and self-promotional rankings get cited far less often than neutral third-party listicles.

If you're writing a "best of" on your own brand's blog, the discipline is to acknowledge real strengths in every option (including competitors) and real weaknesses in every option (including your own). When the page reads as honest analysis with a clear methodology, AI engines treat it almost as if it were a third-party editorial piece. When it reads as a sales pitch, they route around it.

Don't just build the page, get featured on other people's pages

Here's the tactic most teams miss, and it might be more important than building your own "best of" pages at all. AI engines pull from established publishers' "best of" lists, not just from brand-owned content. If your competitor is regularly featured on top 10 lists where you never make it into the top 5, the citation gap isn't going to close just by writing your own page.

The advice from one widely-quoted GEO source is bluntly tactical: "Is your competitor regularly featured on top 10 lists where you never make it to the top 5? Offer the publisher who created the list an affiliate deal they cannot decline."

The point isn't cynicism. The point is that getting featured in third-party "best of" lists is a higher-leverage GEO investment than writing your own. Third-party content gets cited at much higher rates than self-promotional content, sometimes 4x the rate, and the third-party listicles that already exist in your category are more important to your AI visibility than any page you build on your own blog.

The right two-pronged strategy:

Build your own "best of" pages for your category, with you ranked honestly somewhere in the mix, written as neutral editorial. This is your contribution to the category.
Aggressively pursue inclusion in third-party "best of" pages for the same category. Reach out to publishers, offer to be reviewed, provide free trial access, share data, accept honest feedback. This is where the big citation lift comes from.

The brands winning AI search aren't only the ones publishing listicles, they're the ones appearing in everyone else's listicles.

Update the page on a real cadence

"Best of" content decays fast. Pricing changes. Features ship. Companies get acquired. Tools that were great a year ago aren't anymore. AI engines weight freshness heavily, and a "Best CRM 2026" page that hasn't been touched since 2024 looks stale to both humans and extractors.

Quarterly updates are the right cadence for most "best of" pages in fast-moving categories. For each update:

Re-verify pricing for every option
Re-check review counts and aggregate ratings
Add any new entrants worth including
Remove or downgrade options that have lost relevance
Bump the "Last updated" date prominently

Pages refreshed quarterly outperform stale pages by a wide margin in citation behavior, and the cost of a quarterly refresh is minimal compared to writing a new page from scratch.

Use the right schema

For "best of" content, the schema combination that matters:

ItemList, to mark up the ranked list of options as an explicit list with order
Product, for each option in the list, with name, description, brand, and aggregateRating
Review, if you're including written evaluations
FAQPage, for any FAQ section at the bottom

The combination tells AI engines unambiguously: "this is a ranked list of products, with these ratings, in this order." Schema isn't a substitute for good writing, but it accelerates parsing and makes the page legible to extractors that would otherwise have to guess at the structure.

The format AI commercial answers depend on

"Best of" pages are the format AI commercial answers depend on. They're the canonical source for the highest-intent prompts in any category, and they reward discipline, structure, and editorial honesty in a way that few other formats do.

Build the page as a list. Add pros and cons for every entry. Lead with a comparison table. Include real evidence. Stay neutral. Update quarterly. Schema it properly. And, more importantly, pursue inclusion in other people's "best of" pages, because that's where the bulk of your AI citation potential lives. The best-of strategy that wins treats both your own content and your earned third-party features as parts of the same investment.