What is Siteimprove is a harder question in 2026 than it was two years ago, because the company spent the last eighteen months repositioning away from the thing most people know it for.
The short answer people usually give to what is Siteimprove is that it is an accessibility scanner. That is one module of several, it is no longer the top of the marketing hierarchy, and the detection engine underneath it is free open source that Siteimprove itself publishes. If you buy Siteimprove for the scanning, you are paying several thousand dollars a year for something you can get for nothing.
What you actually buy is governance: scheduled crawls across thousands of pages, issues assigned to named editors, a trend line, and a record of what changed last week. That is a real product with real value, and it is worth a very different amount of money to a fifteen-thousand-page university than to a two-hundred-page association site.
This piece covers what the company is and what changed in 2025, the modules and which ones cost extra, what it actually costs and where those numbers are published, how much automated scanning really catches, the score that can mislead the people you report to, why it is not an overlay and why that distinction matters more than price, the CMS and developer picture including a limitation that will surprise you, the free engine Siteimprove gives away, how it compares, and who should and should not buy it.
The short version: the cheapest documented list price is $7,100 a year for one thousand pages of accessibility checking alone, before implementation. Siteimprove’s own published figure for what automated testing catches is thirty to forty percent of accessibility issues. Both of those facts are fine. They only become a problem if you bought the tool believing it was a compliance machine.
What is Siteimprove, and what changed in 2025
Siteimprove was founded in Denmark in 2003 by Morten Ebbesen and is headquartered in Copenhagen, with offices in Bellevue, Minneapolis and London. It is still private. Nordic Capital took majority ownership in 2020, and Nordic Capital’s own portfolio page still lists it as a current holding with 2024 revenue of EUR 105 million.
Two things reshaped the product recently.
In October 2024 Siteimprove acquired MarketMuse, an AI content strategy and SEO platform. In August 2025 it relaunched the whole thing as "Siteimprove.ai," described by chief executive Nayaki Nayyar as "the only enterprise platform that delivers both content compliance and performance on a unified platform."
That repositioning is the thing to understand before you evaluate anything else. Accessibility is now one of four pillars on the marketing site, alongside Analytics, Search and Content Strategy. The company is no longer positioning itself primarily against accessibility specialists. It is positioning against content and search platforms.
There is a practical consequence for buyers. The marketing taxonomy and the contract taxonomy no longer match. The website organizes around four pillars. The help center, the price sheets and the actual contracts still use the older module names, and Quality Assurance and Policy have disappeared from the top-level navigation while remaining documented and sold. When you get a quote, read the module names in the quote, not the ones on the homepage.
The modules, and which ones cost extra
This is the part vendors are vague about and it determines your bill.
Included in what the page-count price buys:
- Accessibility. Automated checks against WCAG 2.2 A, AA and AAA, built on W3C ACT rules. Results are split into "Issues" and "Potential Issues," the latter flagged for human review. PDF scanning included, which matters more than it sounds, since PDFs are where most organizations’ worst accessibility problems live. We went through that separately in the accessible PDF checklist.
- Quality Assurance. Broken links, misspellings, content freshness, basic security and UX checks.
- Policy. Custom rules that flag content breaching your own standards: banned phrases, required disclaimers, brand terms, file types. This is the module most people do not know exists and the one that quietly justifies the purchase for organizations with a lot of editors.
Sold separately, each on its own meter:
- SEO, by credits. Worth weighing against what you already do by hand, which we set out in the on-page SEO checklist.
- Analytics, by yearly page views. First-party analytics, behavior maps, and visual session playback added in late 2025.
- Performance, page speed and Core Web Vitals, by credits.
- Ads, paid search analytics, priced as a percentage of the annual ad spend you manage through it.
- Response monitoring, per check point.
- Manual Testing as a Service, human expert testing on a cadence. This one is expensive and its existence is informative, which we will come back to.
The AI layer, shipped through 2025 and 2026, includes accessibility, analytics, search and content agents, a conversational analytics agent, a PDF and image accessibility agent, a PDF remediation agent that attempts auto-tagging and heading structure, and as of June 2026 an MCP server that plugs Siteimprove into Claude, VS Code and Figma. In August 2026 the federal platform reached FedRAMP Ready status, which is a marketplace listing rather than full authorization.
We could not find published pricing or packaging for any of the AI agents. If they matter to your decision, get it in writing.
What it costs, and where the numbers are actually published
The received wisdom is that Siteimprove does not publish pricing. That is true of its commercial website and false in general, and the exception is the single most useful document available to a buyer.
Siteimprove publishes a full list price sheet on its own domain for its Texas DIR contract. The sheet shows both the discounted public-sector price and Siteimprove’s own list price, roughly nine percent higher.
Accessibility "Complete Package," scaling by pages crawled:
| Pages | Texas DIR price | List price |
|---|---|---|
| 1,000 | $6,437.93 | $7,100 |
| 5,000 | $10,518.30 | $11,600 |
| 10,000 | $13,873.28 | $15,300 |
| 25,000 | $20,673.90 | $22,800 |
| 50,000 | $30,920.18 | $34,100 |
| 100,000 | $46,516.28 | $51,300 |
Note the unit. You are billed by pages crawled, not by domains, not by users, not by issues found. Analytics bills by page views, SEO and Performance by credits, Ads by the spend you run through it.
Two other published sources are worth knowing. UK G-Cloud listings from May 2024 put the accessibility and digital governance license at £8,000 to £28,700, content analytics at £1,500 to £11,100, and standard implementation at £6,300. Every one of those listings states that there is no free trial. And a Texas A&M University System contract runs at $199,995 a year for a large enterprise configuration covering five million pages.
Aggregated reseller data suggests an average contract value around $15,600, with typical bands running from $15,000 to $40,000 for one to two modules at small page counts, up to $150,000 and beyond at scale. Treat those as indicative rather than authoritative, because the sample is not disclosed.
The number to carry away: the floor is about $7,100 a year at list for a thousand pages of accessibility alone, and implementation is a separate mandatory line that has been published as high as £6,300. A realistic first year for the smallest configuration is somewhere around $11,000 to $13,000 for one module.
One caveat on all of it. The Texas price sheet carries no revision date, and the contract predates the August 2025 repackaging. Use it to calibrate, not to quote.
What automated scanning actually catches
Here is where Siteimprove deserves more credit than it usually gets, and where the buying risk actually lives.
Automated accessibility testing cannot detect most WCAG success criteria. This is not a vendor failing, it is a property of the problem. Whether alternative text is accurate, whether a heading structure is meaningful, whether a form error message is understandable are judgments no crawler makes.
The independent evidence:
- Peer-reviewed work published in April 2026, testing all fifty US state homepages, puts automatic detectability at roughly 13 to 30 percent of WCAG 2.1 Level AA success criteria.
- The UK Government Digital Service tool audit found the best of thirteen tools caught 40 percent of deliberately introduced barriers on a strict pass or fail basis, and 50 percent if you count flags requiring human review. The worst caught 13 percent.
- Deque’s widely cited 57 percent figure measures something different: the share of issue volume, not the share of success criteria. Common issue types like contrast and missing alt text are numerous, which lifts the number. Deque’s own repository has an open issue noting the inconsistency between that 57 percent and an 80 percent figure used in its marketing.
Siteimprove’s own published position is the lower one. Its guidance states that "automated testing alone only covers 30 to 40 % of accessibility issues," and that "manual testing is an unavoidable method for catching accessibility issues that automated solutions are not able to detect."
Three things follow, and they are to the company’s credit. It publishes a figure lower than a competitor’s. It splits results into Issues and Potential Issues instead of a single clean verdict. And it sells manual testing as a priced service, which is a commercial admission that the scanner is not sufficient.
For scale, WebAIM’s February 2026 survey of the top million home pages found detected WCAG failures on 95.9 percent of them, averaging 56 errors per page, both up year over year. WebAIM’s own caveat is the sentence to keep: "Absence of detected errors does not indicate that a page is accessible or conformant."
The score that can mislead the people you report to
Siteimprove produces a Digital Certainty Index, a single number weighted one third each to Quality Assurance, SEO and Accessibility.
The help documentation defines the accessibility component carefully, as measuring how well a site tests against standards rather than whether it conforms to them. That wording is accurate. What we could not find is an explicit statement anywhere in that documentation that a high score does not mean compliance.
The practical risk for a nonprofit or an association is not that the vendor lies. It is that a green number travels up an org chart and arrives in a board meeting as "we’re compliant." Two things make that easy. Accessibility is only a third of the composite, so a team can raise the score by fixing broken links and meta descriptions while accessibility stays flat. And the underlying scan, by the vendor’s own published figure, sees thirty to forty percent of the problem.
If you buy this, decide in advance what you will report and to whom, and pair any score you circulate with the coverage caveat. That is a five-minute decision that prevents a bad year.
Siteimprove is not an overlay, and that matters more than the price
There are three categories in this market and they get blurred constantly. Getting them straight is the most useful thing an agency can do for a client.
Scanning and governance platforms crawl your site, report issues, assign work and track it over time. They change nothing on your site. Siteimprove is here, along with Silktide, Acquia Web Governance (which was Monsido until Acquia bought it, then Acquia Optimize, then renamed again in September 2025), Crownpeak DQM and TPGi ARC.
Developer testing tools run in the development loop. Deque’s axe, the axe DevTools extension, Lighthouse in Chrome DevTools and WebAIM’s WAVE. Mostly free, page at a time, no governance layer.
Overlays inject JavaScript that attempts to alter the page at runtime for the end user. accessiBe, AudioEye, UserWay, EqualWeb and thirty-odd others.
Siteimprove is unambiguously not an overlay. It sells no widget, is not listed on the Overlay Fact Sheet, and argues against them in its own glossary: "Overlays are not effective."
That matters because of price. accessiBe publishes a rate card starting at $490 a year. Siteimprove’s floor is $7,100. A nonprofit comparing those two is not comparing two versions of the same thing, and the cheap one is the category the accessibility community has organized against. The Overlay Fact Sheet, with over a thousand signatories including WCAG contributors and accessibility staff at major platform vendors, cites a WebAIM survey in which "a strong majority (67%) of respondents rate these tools as not at all or not very effective," with disabled respondents less favorable still at 72 percent.
One complication worth knowing, because it makes the categories harder to explain than they used to be. Level Access, an audit and governance company, acquired UserWay, an overlay company, in March 2024. One vendor now spans two categories that most practitioners consider opposed.
The CMS and developer picture
Plugins exist for a long list of systems: WordPress, Drupal, Adobe Experience Manager, Optimizely, Sitecore, Sitefinity, TYPO3, Umbraco, Kentico, Magnolia, Plone and more. Contentful is not on the published list and we found no plugin for it.
The most useful of these is the prepublish plugin, which checks accessibility, SEO, policy and QA against draft content inside the CMS before it goes live. Supported on Drupal, AEM, Optimizely, Sitecore, WordPress and several others, so it meets editors inside the tools they already use, whether that is the block editor or a Drupal authoring screen. For an organization with many editors, that is where the behavior change actually happens, because it catches the problem at the moment someone can still fix it cheaply.
There is a documented REST API, a CMS plugin SDK, and connectors for Jira and Azure DevOps. We could not verify whether API access is included in a base subscription or gated.
Now the limitation that will surprise developers. Siteimprove’s crawler can reach non-public sites, but with hard constraints. It supports HTTP basic authentication, form-based login and IP allowlisting. It does not support any form of two-factor authentication. And the site "must be available over the internet, i.e. not only available on your internal network."
In plain terms: a localhost or VPN-only staging environment cannot be scanned. Cornell tells its own users flatly that "Siteimprove is available for public sites only. It cannot be used for secured or locked-down websites."
So if you were picturing this as a pre-deployment gate in CI, adjust. The crawler is a production monitoring tool. Shifting left happens through the prepublish plugin for content, and through the open-source engine below for code.
The free engine Siteimprove gives away
This is the fact that reframes the whole purchase, and almost nobody mentions it.
Siteimprove publishes Alfa, its accessibility conformance engine, as MIT-licensed open source on npm, implementing W3C ACT rules and outputting standard EARL reports. It is actively maintained. You can put it in a pipeline today, for nothing, with no Siteimprove contract.
There is also a genuinely useful free layer on the website: an accessibility checker, a browser extension for Chrome and Firefox that handles HTML but not PDFs, a color contrast checker, an accessibility statement generator and an SEO checker.
A company that gives away its detection engine has told you where it thinks the value is. It is not in finding the issue. Lighthouse, WAVE, axe and Alfa all find the same class of issue, largely because several of them implement the same rule set. The value is in crawling ten thousand pages on a schedule, assigning each finding to a named person, catching regressions at publish time, and producing something you can put in front of a board.
The pricing unit says the same thing. You are billed by pages, which is a measure of coverage and continuity, not of detection quality.
How it compares
For an agency advising a client, the honest comparison is not feature by feature.
Against Silktide, the closest direct analog: similar category, similar shape, also does not publish pricing. Worth putting in any competitive process precisely because it is comparable.
Against Acquia Web Governance, formerly Monsido and briefly Acquia Optimize: relevant if the client is already in the Acquia ecosystem or buys through CivicPlus, and the two rebrands in two years mean older research under either old name still applies.
Against Deque: a different purchase. Deque is where developers live, axe-core is the de facto open-source engine, and Deque also does not publish per-seat pricing. If your client’s problem is that developers ship inaccessible components, Deque is the better answer. If the problem is that two hundred editors publish content nobody checks, it is not.
Against free tooling: Lighthouse, WAVE and axe DevTools cost nothing and find the same categories of issue on the page in front of you. They do not tell you what happened across four thousand pages overnight.
Forrester named Siteimprove a Leader in its Q4 2025 Digital Accessibility Platforms Wave, noting that it "is unique from other Leaders in this market because it provides accessibility as part of a broader unified platform that includes SEO, analytics, and content strategy." Read that label carefully, though: Deque and Level Access were named Leaders in the same Wave, so it separates the top tier from the rest rather than separating these three from each other. That breadth is genuinely the differentiator, and it cuts both ways. One competing platform vendor argues that "accessibility-specific depth often suffers when competing for R&D priority with non-accessibility features," which is a self-interested claim from a rival and also a reasonable question to ask given the 2025 repositioning toward content and search.
Who should buy this, and who should not
Do not buy it if you are a small agency buying for your own sites. The free extension, Lighthouse, WAVE and axe DevTools cover your own quality assurance, and Alfa covers your pipeline. Spending eleven thousand dollars in year one on dashboards nobody is assigned to drive is the most common way this purchase fails.
Do not buy it if you run a two-hundred-page site with one part-time web person. One paid manual audit from a specialist, plus free tooling and a habit, will improve accessibility more than a subscription will, because the subscription buys governance and you have nobody to operate it.
Do consider it when four things are true at once: more than a thousand pages, more than one content editor, a named person accountable for accessibility, and a reason that an auditable trail of scans and assigned fixes has value beyond the scanning itself. In practice that describes mid-size universities, city and county government, health systems, large member associations and multi-brand publishers.
Know it either way if you work with clients in higher education, government or large nonprofits, because many of them already own it and will hand you its reports. Reading those reports correctly, including knowing what the score does not cover, is a service you can provide whether or not you ever buy a license.
So the useful version of what is Siteimprove, for a buyer, is not a feature list. The test that decides it is not a feature comparison. It is one question: who, by name, will open this tool on a Tuesday and act on what it says? If you cannot answer that, you are buying a report generator.
Frequently Asked Questions
What is Siteimprove?
A Danish web governance platform founded in 2003, majority owned by Nordic Capital since 2020, that crawls websites and reports accessibility, content quality, policy, SEO, analytics and performance issues, and assigns them to people to fix.
Is Siteimprove just an accessibility tool?
No, and less so than it used to be. Following the MarketMuse acquisition in 2024 and an August 2025 relaunch as “Siteimprove.ai,” accessibility is one of four marketing pillars alongside analytics, search and content strategy.
What does Siteimprove cost?
The cheapest documented list price is $7,100 a year for 1,000 pages of accessibility checking, from the price sheet Siteimprove publishes for its Texas DIR contract. Implementation is separate and has been published as high as £6,300.
How is it priced?
By pages crawled for the core package. Analytics bills by yearly page views, SEO and Performance by credits, Ads as a percentage of managed ad spend.
Is there a free trial?
The published UK G-Cloud listings all state no free trial. There is a free toolkit on the website, including an accessibility checker and a browser extension, which is separate from the paid platform.
How much of an accessibility problem does automated scanning actually find?
Siteimprove’s own published figure is 30 to 40 percent of issues. Peer-reviewed work puts automatic detectability at roughly 13 to 30 percent of WCAG 2.1 AA success criteria. Deque’s 57 percent figure measures share of issue volume, which is a different measurement.
Does a good Siteimprove score mean my site is compliant?
No. The Digital Certainty Index weights accessibility at one third, alongside SEO and quality assurance, and the underlying scan sees a minority of WCAG criteria. Treat the score as a trend line, not a verdict.
Is Siteimprove an accessibility overlay?
No. It sells no widget, injects nothing into your pages, is not listed on the Overlay Fact Sheet, and argues against overlays in its own glossary.
Can it scan my staging site?
Only if staging is reachable from the internet. Siteimprove supports basic authentication, form login and IP allowlisting, but does not support two-factor authentication, and a site available only on an internal network cannot be crawled.
Does it integrate with WordPress and Drupal?
Yes, both, along with AEM, Optimizely, Sitecore, Sitefinity, TYPO3, Umbraco and others. The prepublish plugin, which checks draft content before it goes live, is the integration that changes editor behavior.
Is there a free version of the scanning engine?
Yes. Siteimprove publishes Alfa, its accessibility conformance engine, as MIT-licensed open source on npm. It implements W3C ACT rules and can run in a pipeline with no Siteimprove contract.
What are the main alternatives?
Silktide and Acquia Optimize in the same governance category, Deque for developer-native testing, and free tooling including Lighthouse, WAVE and axe DevTools for page-level checks. Overlay products are a different category and are not a substitute.
Is it worth it for a small nonprofit?
Usually not. Below roughly a thousand pages and one accountable owner, a manual audit plus free tooling delivers more accessibility improvement per dollar, because the subscription’s value is governance and a small team has nobody to run it.
