Comparison pages, and the honesty that makes them work

Why this page and not another blog post

Almost everyone who lands on “YourProduct vs BigCompetitor” is at the end of their decision, not the start. They know the category, they know at least two names in it, and they are looking for a reason to pick. Nothing else you publish gets that audience.

There are three page shapes and they are not interchangeable:

• You vs them. “Acme vs BigCorp”. For people already choosing between exactly the two of you. Highest intent, lowest volume.
• Alternatives to them. “BigCorp alternatives”. For people actively leaving, usually over price or a specific frustration. Higher volume, and the visitor expects a list rather than a pitch, so give them a real one.
• Best X for Y. “Best invoicing tool for freelance designers”. Broadest, and the shape answer engines reach for most when someone asks for a recommendation.

Write the alternatives page first if you are picking one. It gets more traffic than the head-to-head, and it is the format an answer engine is most likely to quote, because it already looks like the answer to the question being asked.

The template

Seven sections, in this order. The order matters more than the wording: it front-loads the answer, which is what both a skimming buyer and a language model need.

1. The verdict, in the first 60 words

Say who should pick each product, immediately, before any argument. Something like: “Choose Acme if you invoice fewer than 50 clients a month and want flat pricing. Choose BigCorp if you need multi-entity accounting and have someone to administer it.”

This feels like giving away the ending. It is the single highest leverage thing on the page. A buyer who reads that and self-selects out was never going to convert, and a buyer who reads it and recognises themselves has just been sold to by someone who appears not to be selling.

2. The comparison table

Real rows, real values, no ticks and crosses where a number would do. Include price. Include the row you lose.

Keep it under about ten rows. A forty-row table is a document nobody reads, and it usually exists because someone padded it with rows they win rather than rows that matter. Pick the criteria a buyer actually decides on and let the loser rows sit there.

3. Where they are genuinely better

An explicit section, with a heading, saying what the competitor does better. Not a grudging aside buried in paragraph nine.

This is the section that makes the whole page work, and it is the one everybody cuts. It is also the section most likely to get quoted, because it is the only part of the page that reads as information rather than marketing.

4. Where you are genuinely better

Now you have earned it. Two or three things, specific, with the evidence attached. “Setup takes about four minutes because there is no chart of accounts to configure” beats “easier to use” by a distance.

5. Pricing, side by side, with the awkward part included

Actual numbers and actual plan names, including the point where you get more expensive. If you are cheaper at 5 seats and more expensive at 50, say so and say where the line is. Someone will work it out anyway, and finding it themselves after they trusted your table costs you the sale twice.

6. Who should not choose you

One short paragraph. The strongest credibility signal available, and almost nobody uses it.

7. FAQ, using the questions people actually ask

Pull them from your support inbox and sales calls, not from a keyword tool. Three to six, answered in two to four plain sentences. Mark this section up with FAQPage JSON-LD, since those are real questions really on the page.

The honesty argument, properly

The instinct is that conceding a row loses the sale. Three reasons it does the opposite.

The reader has already used the competitor. Very often they are on your page because they use BigCorp today. When your table claims BigCorp lacks something they used this morning, you have not damaged BigCorp. You have told the reader you are either careless or lying, and every other row on the page is now suspect.

Conceding is a credibility trade. Admitting the competitor wins on integrations buys belief in the three rows where you say you win. A page where you sweep every category reads as an advert, and people discount adverts automatically.

Answer engines lean on the honest ones. This is the newer reason. A model synthesising “what is the best X for Y” is looking for text that reads like a considered comparison, and a page that names conditions under which a rival wins looks far more like a source than a page that does not. Every major AEO writeup lands on the same conclusion from a different direction: balanced comparative content gets cited, promotional content gets skipped.

The legal bit, briefly

• Naming them is fine. Using a competitor’s name truthfully to identify their product is nominative use and is normal comparative advertising in the US, UK and EU.
• Their logo is a different question. Using it in a way that suggests endorsement or partnership is where trademark trouble starts. Plain text names are the safe default and look cleaner anyway.
• Every claim needs to be checkable, and dated. “BigCorp charges $49 per seat” wants a “as of March 2026” next to it. Pricing changes, and an out-of-date claim on a live page is a false claim.
• Never invent ratings for anybody. Publishing AggregateRating markup for scores you made up violates Google’s structured data policies and is a manual action risk. This applies to your own product too.

Maintenance, which is where these die

A comparison page is a promise to keep checking. Competitor pricing moves, features ship, and a page that was accurate in 2025 is an embarrassment in 2027.

Put a visible “last checked” date on the page, review it quarterly, and write down where each number came from so the review takes twenty minutes rather than an afternoon. If you will not commit to that, write fewer pages. Two accurate comparisons beat nine rotting ones, and the rotting ones actively cost you the trust the accurate ones were building.

Who should not write these

If nobody in your category has a name yet, there is nothing to compare against and no search demand to capture. If your product is genuinely worse than the alternative for every buyer, a comparison page is a well-optimised advert for your competitor on your own domain. And if you cannot state a segment you are the right answer for, the problem is positioning, and no page will fix it.

Common questions

The hardest part is the “where they are better” section, because you are too close to it. That is a good thing to hand to another founder who has never seen your product. It is also exactly the kind of review the exchange exists to arrange.