How we compare Pondral to other tools.

What we compare

We compare on the dimensions where AI visibility tools actually differ, not on a feature checklist chosen to flatter us. In practice that means some mix of: which AI engines are covered, whether the scoring model and its weights are published, what evidence sits behind a single score, how the tool collects answers from the engines, how it handles the fact that AI engines return different answers to the same question, what the tool does beyond measurement, and the entry price.

Not every page carries every dimension. We pick the rows that matter for the specific tool, because a row we have nothing useful to say about is filler that makes a page look more thorough than it is. Where a rival does something Pondral does not do at all, that gets its own row rather than being left off.

Where each fact comes from

Facts about another company come from that company's own public materials. Their website, their documentation, their pricing page. We do not source competitor facts from third-party review sites, analyst summaries, or what a sales team told a prospect, because those are second-hand and frequently out of date.

Every comparison page links to the competitor's own site and tells you to check it. That is not a disclaimer we bolt on at the end. It is the point: our claim is that a fact was published at a URL on a date, and you can go and look.

Facts about Pondral come from the product and are checkable in it. Our scoring rubric and its weights are published in full at pondral.com/methodology, with a versioned changelog of every scoring change.

How facts are dated

A claim about a competitor is a claim about a moment. This category moves fast enough that a true statement can go stale in weeks, so competitor facts are written with the month and year they were checked attached to the claim itself, in the row where the claim appears, rather than only in a footer.

Because a page carries the dates from the review that last touched it, pages reviewed at different times carry different levels of detail in their dating. Read the date next to a claim, not the freshness stamp at the top of the page. Where a row carries no date, treat it as a prompt to go and check rather than as a current fact. The whole surface was last reviewed in July 2026 and is re-verified at least every 3 months.

What we do when a competitor publishes nothing

Often the honest answer is that we could not find something. A rival may not publish its scoring weights, or its pricing, or its engine list. There is a large difference between “they do not do this” and “we could not find that they do this”, and only the second one is something we actually know.

So an absence is recorded as a search that failed, with the date and the place we looked, and never as a fact about the competitor. If we have no information at all on a dimension, the cell holds a question for you to ask them rather than an assertion from us. That is why the third column on our comparison tables is headed What to evaluate: much of it is a checklist for your own diligence, not a set of findings.

A tool being quiet about something is not evidence against it. Plenty of good products do not publish their internals.

Hedging, and conceding

Two standing rules apply to every comparison page.

Hedge anything we did not verify ourselves. Competitor claims are attributed to the source and the date, and the reader is told to confirm. We would rather under-claim about a rival than overstate a weakness, because an unfair comparison is both wrong and, on a page whose whole argument is that we show our work, self-defeating.

Concede where a rival is genuinely better. Every comparison page carries a section headed Where they may fit better, and it has to name something real. Broader engine coverage, a lower entry price, a security certification Pondral does not hold, a capability we do not have, or simply a longer track record. If we could not find a single honest reason to pick the other tool, we would not have understood it well enough to compare it.

Pondral does not hold a SOC 2 report. Several tools we compare against do. That appears on the pages rather than being left for you to discover.

What we will not do

We do not build a competitor feature matrix of ticks and crosses across every rival. Those grids look authoritative and are mostly unverifiable: dozens of cells about other people's products, each one able to go stale without anyone noticing, on a page that reads as a claim of fact. We compare one tool at a time, on its own page, against its own published material.

We also do not compare against a version of a competitor that no longer exists. When a rival ships something that closes a gap we used to point at, the row changes at the next review. If that means we lose a differentiator, we lose it.

How to tell us we got it wrong

If you work at a company we compare against, or you are evaluating one and you find a row that is out of date or unfair, tell us and we will check it against your published material and correct it. Corrections do not wait for the next scheduled review.

Contact us with the page and the row.

The pages themselves are indexed at pondral.com/compare.