Capterra product listing: eligibility, submission, and free versus paid options, with a dated count you make yourself

You get a product on to Capterra by submitting it as a vendor, completing the product details the form asks for, and passing the platform's own review, and the useful number on a page like this is not a processing time but the count of eligibility criteria the current guidelines actually state. That count is reproducible, because the document is public: two people can read the same version on the same day and agree. As of 29 September 2026 we have not completed a dated count of Capterra's current vendor guidelines, so we publish no count. What follows is the counting method in full, the eight fields to record per criterion, the rules that stop two readers getting different totals, and the nearest real fraction we do hold about whether directory listings reach AI answers, which is from legal research software in India.

The answer, first

Four statements.

A criterion count is the safest number on this subject. A processing time needs a stopwatch and repetition. A criterion count needs a document, a date and a counting rule, and anybody can check your work by opening the same document. That is why the number this page is built around is a count of stated criteria rather than a benchmark for how long anything takes.

Our count is not published, because we have not made it on a dated basis. As of 29 September 2026 the cell is empty. We are not going to write a plausible number in it, and we are not going to quote another article's number, because that article does not say which version of the guidelines it read.

We do not restate the guidelines. The document to read is Capterra's own vendor material on adding a product and its vendor terms, together with its review guidelines. Read the current version, record its URL and the date, save a copy, and count from your copy. A paraphrase on our page would be stale within a year and would then be a false claim about a company that did not make it.

Free versus paid is a question about placement, not about presence. Directories of this kind generally let you have a profile without paying and sell placement or lead flow on top. What each tier includes on the day you read this is on the platform's own pricing and vendor pages. Record what you find with the date and the URL on the same line, and record the currency and the billing basis too, because those change the meaning of any figure.

How it was measured

This is a documentation audit, and it has its own discipline. The whole point is that a second person can repeat it and get your number.

One. Fix the source set before you read. Name the exact pages that count as the guidelines, and exclude everything else. We would use four: the vendor sign up page, the vendor help article on adding a product, the vendor terms, and the review guidelines. Anything in a blog post, a webinar, a sales email or a support chat is excluded and recorded in a separate list as unofficial. A count that mixes official pages with a support agent's reply cannot be reproduced.

Two. Capture the version. For every page in the source set, record the URL, the date and time you read it, any last updated stamp printed on the page, and save a copy you can open later. Without a saved copy your count cannot be defended six months from now, because the page will have changed.

Three. Apply five counting rules, and publish them with the number. These are the rules that decide whether two honest readers agree.

• One criterion per testable condition. If a sentence contains two conditions joined by "and", it is two criteria.
• A requirement about you is a criterion. A requirement about your behaviour later is not. A rule about who may submit counts. A rule about how you may ask for reviews afterwards is a separate list.
• A recommendation is not a criterion, unless the document says it is required. Keep recommendations in a second column and report both totals separately.
• A restatement is not a new criterion. If the same condition appears on two pages, count it once and note both locations.
• Anything conditional on a tier, a region or a category is counted once and flagged, with the condition quoted.

Four. Record eight fields per criterion. The criterion in the document's own words. The page it came from. The date read. Whether it is required or recommended. Whether it is testable by you before submitting, yes or no. What evidence would satisfy it. Whether your product currently satisfies it. The date of that self assessment.

Five. Two readers, then publish the disagreement count. Have a second person count the same saved copies without seeing your sheet, then publish both totals and the number of rows you disagreed on. A criterion count with no disagreement count behind it is one person's reading of a document. In our experience the disagreements are almost never about the rules, they are about the first counting rule above.

Six. Print the denominator arithmetic. If you assess your product against the criteria you counted, report it as a fraction with both numbers visible: satisfied now, out of criteria counted, as of a date. If you counted 19 required criteria and satisfy 15, the line reads 15 of 19 satisfied as of the date, with the 4 unsatisfied rows listed. Never report a percentage on its own, because a percentage hides whether the denominator was 19 or 90.

Seven. Set a recheck date and keep the old sheet. Directory guidelines change quietly. Recheck quarterly, keep every previous sheet, and publish what changed between versions. That change history is the most quotable thing a page like this can own, because nobody else keeps it.

Eight. Treat anything a machine tells you about the rules as a claim. If you ask an assistant to list the eligibility criteria, save the answer as a claim with its date and then check every line against your saved copy. The reason is on the record: in the aiknowsus.com audit of September 2026, Perplexity withdrew its own earlier statement about how many questions it had searched for, saying it could not honestly substantiate the claim, and in another batch that its claim to have searched all five was not adequately supported. Three separate batches.

What the numbers were

Eligibility criteria counted by us in Capterra's current guidelines: none, as of 29 September 2026. The audit has not been run. There is no count on this page and there is no estimate.

What we deliberately refuse to publish instead. We will not name a number of criteria from memory. We will not repeat a count from another article, because no article we have seen names the version it read. We will not restate the platform's prices or tier contents. And we will not convert a figure in one currency into rupees and present it as an India price.

The nearest real fraction we hold, with its sector named. Across six questions in the clawlaw.in programme on 17 September 2026, no third party review or directory source made it into any answer at all: 0 of 6, in Claude's run. One well known review site did appear in the raw results and was discarded, because the list it offered was of American products and so was not an answer to an India question. The sector is legal research software sold in India. That is a finding about six questions on one day in one sector, and it is not a finding about directories in general.

Two further dated results worth having before you spend money on placement.

• Across the whole aiknowsus.com capture of September 2026, 24 batches and 72 conversations, the most cited domains were the assistants' own documentation and the vendors' own websites, with a single well known review site far down the list. Perplexity.
• On the questions about how to look something up, official portals took every position above any commercial product, and Claude said plainly on 17 September 2026 that no commercial product should rank above the official portal for a question about using that portal.

The practical reading is that a directory profile is a buyer facing asset first. If your reason for buying placement is to be named by an assistant, measure that separately and blind, before and after, with the dates written down.

What this cannot tell you

• It cannot tell you the current criteria. Only the document can, on the day you read it, which is the whole point of the method above.
• It cannot tell you what a paid tier is worth. That depends on your category's traffic on that platform, which the platform measures and you do not.
• A criterion count is not a difficulty score. Nineteen easy criteria are easier than four hard ones, and a count says nothing about which you face.
• Our 0 of 6 is legal technology in India. It is the nearest real fraction we hold and it is not evidence about your sector or your country.
• Nothing here promises a position in an AI answer, on a directory or anywhere else.
• Our own counts over capture files cannot be checked by you yet, because the captures are not published. The clawlaw.in pricing and app store findings can be checked from outside today.

Sources and change log

Sources. The counting method, the five counting rules and the eight fields are ours. The dated results come from the clawlaw.in programme of July to September 2026, recorded in Tier_1/claude_response_17_09_audit.md and Tier_1/GEO_BASELINE_RESULTS_2026-07-27.md, and from the aiknowsus.com audit of September 2026 across 24 batches and 72 conversations. No sentence on this page paraphrases Capterra's documentation, and no price from any platform appears on it.

Change log. 29 September 2026, first publication, stating that no dated criterion count has been made. When the audit runs, this section will carry the source set, the read date, the required count, the recommended count, the disagreement count between two readers, and the self assessment fraction with both numbers visible. Corrections are made on the page and noted here rather than edited in silence.

Common questions

Why count criteria at all, instead of just following the form?

Because the form tells you what to type and the guidelines tell you what will get you removed later. A written count also gives you a checklist you can hand to somebody else, and a dated record of what the rules were when you joined, which matters the first time a profile is changed without notice.

Is the free profile worth the work if we are not paying for placement?

It is another public page that states your facts in a form a buyer and a machine can read, which is worth an afternoon. Keep your expectations tied to evidence: in the one sector we measured, across six questions on 17 September 2026, no directory source reached any answer in Claude's run. Do it for the buyers who browse directories, and measure the AI effect separately.

How often should we recheck the guidelines?

Quarterly, and immediately after any email telling you something about your profile has changed. Keep every old sheet. The quarter on quarter diff is more useful than any single snapshot, and it is the part nobody else bothers to keep.

Can we ask customers for reviews once we are listed?

Every one of these platforms has its own rules about soliciting reviews, incentives and wording, and those rules are the ones most likely to get a profile penalised. Read the review guidelines as a separate document from the listing guidelines, count them separately, and keep the wording you relied on with its date. We do not restate them here for the same reason we restate nothing else.

Should the count include criteria we already satisfy?

Yes. Count everything stated, then report satisfied out of counted as a fraction with both numbers visible, as of a date. A count of only the ones you fail is a to do list, not a measurement, and it cannot be compared with the next quarter's sheet.

What to do first

Open the four pages in the source set today, save a copy of each, and write the URL and read date at the top of a fresh sheet. Count the required criteria using the five rules above, then ask a colleague to count the same saved copies without looking at your sheet. Publish both totals and the number of rows you disagreed on. That gives you a dated, reproducible number by the end of the afternoon, which is more than any article about this platform currently offers.