Guide 08 · Evidence review
Are comparison pages worth building at scale?
A few are worth building. Hundreds are not. Comparison answers account for 2.2% of citations and listicles for 21.88% — so generating a page for every competitor pairing aims most of your effort at a fiftieth of the opportunity.
But the handful you do build have a specific job that nothing else does, and it is not the job most people think they are hiring them for.
Where does the citation share actually go?
| Answer shape | The query behind it | Share of citations |
|---|---|---|
| Listicle | “best X for Y” | 21.88% |
| Comparison | “X vs Y” | 2.2% |
| Alternatives | “alternatives to X” | 0.29% |
The common advice is to generate a /vs/ page for every pairing you can name. Set against that distribution it is close to backwards — and thin programmatic pages are exactly the kind of page an engine has no reason to quote, because there is nothing in them a competitor’s page does not already say.
So why build any at all?
Because of a finding from a different measurement, and putting the two together is what makes this decision clear.
Across ten plugin categories, All-in-One WP Migration — a product with over five million active installs — was named in exactly one answer out of seven. That one was the answer to “UpdraftPlus vs All-in-One WP Migration”, where the question itself supplied the name.
A comparison page does not create discovery. It serves demand that already exists — a buyer who has already got both names in their head. When it works, it works because the query named you, not because the page found you an audience. Judge it on that basis and the maths changes.
That is why the 2.2% figure understates their value while still being the right number to plan with. Comparison queries are low volume and high intent: someone typing “X vs Y” has already decided to buy something and is choosing between two options, one of which is you. A discovery query is worth more traffic; a comparison query is worth more money per visit.
The alternatives shape at 0.29% is the one to leave alone entirely. “Alternatives to X” pages are usually written by X’s competitors, and they read that way to a reader and to an engine.
How many should you build?
As many as you have real, named competitors you actually lose deals to. For most plugin companies that is three to six, not sixty.
| Build a comparison page when | Skip it when |
|---|---|
| Support or sales hears the pairing regularly | You assembled the pairing from a keyword tool |
| You can state where the other product is genuinely better | Every row of the table favours you |
| You will keep both sides current | It will be stale in six months and stay that way |
| The differences are concrete — limits, pricing, what breaks | The differences are adjectives |
Row two is the one that decides whether the page is quotable. A comparison where you win every line is marketing, and an engine writing a balanced answer has no use for it. A comparison that says “their import handling is better if you are moving large multisite installs; ours is better under a memory limit” contains an actual fact worth lifting — and the concession is what makes the rest of the page credible.
What should the rest of that effort go into?
Listicle-shaped content, which carries ten times the citation share — and here the measured source data says something encouraging about who is allowed to write it.
Across ten categories on both engines, 41% of Perplexity’s source slots and 47% of Google’s pointed at domains owned by product vendors. Not independent review sites. Companies with a blog, writing about categories — including categories they sell nothing in. One booking-software network was cited in seven of ten categories on Perplexity and eight of ten on Google, selling nothing in six or seven of them.
So the highest-return page for a plugin company is not “us versus them”. It is a genuinely useful roundup of a category adjacent to your own, written well enough that an engine has a reason to read it. That is the one pattern in this data observed travelling across both engines, and it costs time rather than money.
What this page cannot tell you
The citation-share percentages come from published observational work across a general corpus, not from WordPress specifically. The plugin-category figures are mine: ten categories, two engines, one day. Small samples on both sides.
The All-in-One WP Migration observation is a single case. It is a vivid one — five million installs, named only when the prompt supplied the name — but one product does not establish that comparison pages generally capture existing demand rather than creating it. It is consistent with the citation-share data and it is not a proof.
And “build three to six” is a judgement, not a measurement. The evidence supports “far fewer than the programmatic advice suggests”; the specific number is mine and you should adjust it to how many competitors your buyers genuinely weigh you against.
How this was made: figures come from our own runs in the engines’ public interfaces or from the studies named above. Drafted with an AI assistant (Claude) and reviewed by Alauddin Chowdhury. Method and AI use.
Which shape is winning in your category?
Send me your wordpress.org slug. I will pull the full source list for your category from both engines and tell you what shape those pages are — roundups, comparisons, docs, forum threads — and which of them name you. Free, no call, usually within a day.