How Many QHSE Vendors Publish a Price? What the Data Shows

A buyer who asks what QHSE software costs can meet two situations: a published number with conditions, or no number at all. This article counts both. Of the 212 platforms researched on this site, 78 publish a current price on their own pages or another public listing and 134 do not show a paid amount that could be verified as current. For the 78, the more useful finding is what the price is charged on, because a price per user, a price per site and a price per module cannot be set side by side without arithmetic.

The article names no vendor as cheapest, dearest or best, and it publishes no average. The figures come from the dated observations behind the QHSE Software Pricing Index 2026, checked between 28 September and 14 October 2026.

How many publish a price

The rule behind the count is simple. A platform counts as publishing a price when its own pages, or another public listing, show an amount that the research record does not call legacy, historical or unverified. A free plan alone does not count, because it is not a paid amount. A counted amount can still lack a currency, a billing period or a full description of what the package includes.

On that rule, 78 of 212 platforms (37%) publish. The other 134 split into three groups. For 123 of them the review recorded no paid amount at all. Four offer a free plan without a published paid amount. Seven have an amount on record that the evidence itself calls old, legacy or unverified, so it was left out of the count.

The group of 123 is the largest, and its meaning is limited. A missing price may reflect configurable scope, a sales-led model, a page behind a form, or simply a pricing page the review could not read. It is not evidence that the product is expensive, and it is not evidence of how a vendor treats a buyer. The existing price transparency article deals with that inference in more detail.

Price transparency by category

The share of platforms that publish a price varies sharply by category. These counts use categories with at least three researched platforms, as the pricing index does. The table shows seven of the 18 category rows in the index; the lowest share there is risk management, with none of its 3 platforms.

The records do not establish why the shares differ; the counts only describe the researched set. Categories here are the ones recorded on the pricing observations, which differ slightly from the categories used in the capability map, so do not combine the two sets of numbers.

What the published prices are charged on

For the 78 platforms with a published price, the pricing model recorded in each observation says what the amount is based on. The next chart counts them by an approximate keyword rule on that text, described in the method note. A platform can appear in more than one bar, for example a per-user price plus a charged module, so the bars do not add up to 78.

Three things stand out.

First, by the keyword rule, per user or per seat is the most common unit, at 27 of the 78, but it is not the majority. Per site or per location is named by 6, and 18 name a module, a device, a tenant, a form, a document or an organisation as the unit. Another 33 describe plans or tiers without naming a billing unit in the model text, so the unit has to be checked on the vendor's own page.

Second, 41 of the 78 have model text that mentions an enterprise plan, a quote, custom pricing or on-request pricing. In these cases the published price may be the entry point of a ladder whose upper rungs are quoted. Fifteen mention a free plan or tier next to the paid plans.

Third, the amounts themselves carry caveats. Of the 78, 60 have an entry-price statement recorded and 18 are counted because a plan amount is published even though the record has no one-line entry price. Eleven carry a note that the currency was not confirmed, and 15 mention a minimum number of users or a user cap. Seventy-six link to the page the amount was read from.

Why there is no average

The tempting next step is a median. The records do not support one. The published amounts are per user per month, per user per year, per site per month, per location per year, per module, per tenant, per document and per emission unit. They appear in at least five currencies, and some record that the displayed currency symbol was never explained. Converting them to a common unit would require assumptions about team size, billing term and scope that differ for every buyer, and the result would look like a market price without being one.

What can be compared is the shape of the sale: whether a price exists, what it is charged on, what the minimum commitment is and what is excluded. Those four questions are the ones to put to every vendor in writing.

How to turn a published price into a comparable figure

Take a worked example with invented numbers, so that no vendor is implied. Suppose a team needs 12 inspectors and 2 supervisors, and two published prices look similar. Vendor A shows a per-user price of 20 a month, billed annually, with a five-user minimum. Vendor B shows 250 a month per site, unlimited users, billed annually.

• Vendor A costs 3,360 a year if all 14 people need a full seat (14 users at 20 for 12 months), or 2,880 a year if only the 12 inspectors do. The five-user minimum does not bind here.
• Vendor B costs 3,000 a year for one site (250 for 12 months) and 9,000 for three sites.

Neither figure is better in general. The per-site model gets more attractive as the number of users grows, and less attractive as the number of sites does. The same arithmetic is what the pricing guide and its quote worksheet are for. Always ask for these items in writing:

1. The unit the price is charged on and what counts as one unit.
2. Minimum users, sites or modules, and the minimum term.
3. Whether the amount is before tax, and in which currency.
4. Which features sit in the published plan and which need a higher tier.
5. Set-up, training, integration and support fees that are not in the subscription.
6. Price protection at renewal and the notice period for leaving.

The RFP template puts these into a form every vendor can answer on the same terms, and the demo checklist keeps the answers beside what you saw. The software ROI calculator takes your own quoted figures, and the article on free trials counts the other kind of openness a vendor can offer.

How to cite these figures

Quote the count and the rule: "78 of 212 researched QHSE platforms publish a current price, as of the observations checked between 28 September and 14 October 2026." Avoid "37% of QHSE software is free to price", which mistakes a researched set for the market. The model counts describe the keyword rule applied to the pricing model text and should be quoted as approximate. The row-level data, with the source page and date of each observation, is on the pricing index, and the datasets page explains reuse under CC BY 4.0.

Method and limits

Each platform was reviewed against its own published pages or other public listings and the date of each check is recorded; no vendor was contacted and no quote was requested. The headline counts come from the generated pricing observations and charts, and the count of 78 follows the published-price rule in the pricing index.

The model counts apply a keyword rule to the pricing model text of the 78 published observations. A text counts as per user when it names a user, seat, worker or employee unit; as per site when it names a site or location; as other unit when it names a module, application, device, tenant, form, document or organisation; and as higher tier quoted when it uses enterprise, quote, custom or on-request wording. A text that names a free plan or tier counts as free. This is a reading of a short description, not a reading of the pricing page, so a count may move by a few platforms when the text is refined.

Limits to keep in mind:

• Researched set, not the market. The 212 platforms were chosen for research by this site.
• Published is not the price you pay. Amounts are entry points, recorded as the vendor states them, and are not converted.
• Pages change. A price seen on one date may differ later, so check the date on each observation.
• Documentation review. Nobody tested checkout, discounts or negotiated rates.

One of the 212 profiles belongs to a vendor that pays for sponsored placement on this site, and the editor works at that vendor. Its record was read by the same rules as every other and is counted in every figure above like any other profile; it is not named here. The methodology page holds the full disclosure, and the 2026 report uses the same records.