We Audited 16 IT Services Companies for AI Visibility. Here Is What We Found.

Why we audited IT services companies

Managed IT is bought on trust. The buyer is usually an IT director, a COO or an owner at a mid-market firm, and the purchase usually follows a trigger: a security scare, a failed audit, a departure from the IT team, a cyber insurance renewal with new requirements. The decision is about security posture and response time, and the person making it is nervous.

If the buyer puts that question to an AI assistant, we wanted to know what it says. So we asked. When someone describes an IT problem or asks for providers without naming one, which companies get named? And does being an established IT company make you one of them?

A caution on this headline number. Four of the five queries written for Softchoice contained its own name, which breaks our rule of leaving the name out. Softchoice scored 80%, one of the two top scores, so it sits inside the 23.8% average and affects it. Four of the 16 companies are also off-category (a technology publication, a sheet-metal software maker, a residential internet provider and a business process outsourcer), so this is a small, mixed sample and not a measure of managed service providers. We have not changed any figure. Treat 23.8% as indicative until the audit is re-run on a cleaner list.

How the audit works

The tool reads each company's homepage, classifies it in the words a buyer would use, writes five buyer-intent queries (one local-style, two category, one problem-first, one comparison) and runs each as a live, grounded AI search. By design the company's own name is left out, but in this audit that did not hold for every query (see the caution above and the methodology section). It is one run, five queries per company, on one day. AI answers change between runs, so read every score as a snapshot. The full method is in our audit documentation, and all twelve industry audits are on our AI visibility research page.

A company is either in the answer or it is not.

The distribution: half were not named at all

Eight of 16 were not named in a single query, including Presidio, Trace3, Unisys, AHEAD, Compucom, Sagiss and ThinkGard. Only two reached 80%, Dataprise and Softchoice, and the second needs a caveat.

A caveat on Softchoice. Our classifier read its homepage as simply "business", and four of the five queries it wrote contained Softchoice's own name, such as "what does softchoice do". A company named in a query about itself is a weak test of discovery. Dataprise, by contrast, was tested only on generic queries such as "best managed IT support for small businesses" and "top managed IT and cybersecurity providers comparison", and was named in four of its five. Treat Dataprise as the clearest strong result in this sample.

What kind of question loses the most

Problem-first queries did far worse than anything else: 1 named company in 16, or 6.2%. And the one that did was Softchoice on "we need help managing our software licenses", a software licensing question rather than managed IT.

Look at the problem queries written for the managed IT and security providers: "our servers keep going down and we need professional IT support", "my company keeps having IT downtime and security issues", "our IT infrastructure is vulnerable and we need better security monitoring". None of them named the company tested. Yet this is the real shape of an MSP purchase. The buyer arrives with a trigger and a problem, not a company name.

Comparison queries did best at 31.2%, and category queries followed at 28.1%. Local-style queries were named 25.0% of the time. The takeaway is not that one type is safe. It is that the situation-driven question, which is closest to how MSP deals start, was where this sample was weakest.

Who got named instead, and where the answers came from

When a company was not named, AI named someone else. The most common name by far was Accenture, which appeared as a named alternative in 14 of 80 answers, mostly for the larger enterprise providers in the sample. Among managed IT specialists, Dataprise was named in 6 answers and Ntiva in 5. CDW, Cognizant, Deloitte, CGI, Capgemini and IBM Consulting also recurred.

The sources behind the answers were widely spread. Across 624 citations on 418 distinct domains, Wikipedia was the largest single source at 15 citations, 2.4% of the total. After that came a mix of types:

• Provider directories and review pages. Clutch appeared 8 times, Gartner's review pages 5 times and CloudTango's MSP listings 4 times. G2 appeared only 3 times, and Trustpilot and Reddit not at all.
• Trade media. BizTech Magazine and Forbes both appeared, BizTech 8 times, though it was also one of the companies we audited.
• Provider-written "top providers" lists. Corsica Technologies, Digacore and Red River each had articles cited, such as "top managed IT service providers" roundups. Dataprise itself was named inside a comparison table on one of them.
• Local provider pages. Pages such as managed IT services in Los Angeles and in Austin were also cited.

For a sector whose websites are mostly service menus, that is the number to sit with. In these answers, the sites AI read were other people's: directories, publications and competitors' comparison articles.

The seven we could not scan

Seven of the 23 domains we tried, connection.com, ctts.com, insight.com, pcm.com, sentinel.net, shi.com and wwt.com, could not be audited. Our crawler was shown a bot-verification page or an error instead of the homepage. That is 30% of the domains we attempted, and it is notable.

What can we conclude? Very little. Several of these are large companies, and heavy bot protection is common for sites of that size. Our crawler is one kind of automated visitor, and AI assistants use others. We do not know whether any of these sites also blocks AI crawlers. What we can say is that if a provider's security layer does block them, its content is not being read, and that is worth checking.

What to do with this, if you run an IT services company

1. Describe the buyer, the sector and the framework in one plain sentence. "Managed IT and security for 50 to 500 person healthcare groups, with HIPAA-aligned support" is easy to place. "Proactive, secure, responsive IT" is not. Vertical specialization is the lever we would pull first, though we have not measured it here, and you should only claim sectors you genuinely serve.
2. Write for the trigger, not the service menu. Build pages for "our cyber insurance renewal requires 24/7 monitoring" and "our IT person just left", with what happens, how long it takes and what it involves.
3. Publish an honest pricing and scope page. Buyers ask what managed IT costs per user almost at once. Explain what drives price and what is included, so there is something to quote.
4. Check where you appear. Compare your name, services, locations and sector focus across Clutch, any MSP directory you are on, trade media and partner listings. Inconsistent or missing entries make you hard to place.
5. Evidence every security and compliance claim. Name the framework, the scope and the date, and avoid promising outcomes. A SOC 2 report is an independent examination of controls, not a certification, so describe it that way. Nothing here is legal advice.
6. Check whether your site can be read. Test what your bot protection shows a plain automated visitor.
7. Re-test regularly. AI answers change, so a single score tells you little.

For providers who know the accounts they want, a named list of target firms and outreach timed to public triggers is a natural companion to this work: see signal-based outreach. If you want the inbound work done as a program, see our AEO, GEO and SEO service for IT services companies and MSPs.

Every company we audited

Not scanned (bot-verification page or error): connection.com, ctts.com, insight.com, pcm.com, sentinel.net, shi.com, wwt.com.

Check your own number

Every company above was run through the same tool below. It takes about ninety seconds, and your company's name is never in the query.

Do AI assistants recommend you?

Five buyer-style queries for your category, run through live AI search, show who gets named instead of you. Takes about ninety seconds.

Your website address

Run the free audit

Free. The full report arrives by email so the tab can close.

Methodology and honest limitations

What we did. On October 3, 2026 we ran the live audit tool that powers our free checker against 23 IT services domains: one run, five queries per company, visibility measurement only. The step-by-step method is in our audit documentation. A company that disputes its result can ask us to re-run it.

The sample is mixed, and it is not a sample of small MSPs. We chose real, identifiable IT services domains, but the list mixes IT solutions providers, resellers and integrators, and several are large. The classifier's own labels show the spread: five were read as managed IT services providers, one as a managed cybersecurity and disaster recovery provider, and the rest as IT infrastructure, consulting, enterprise managed services, AI consulting and a business process outsourcing provider. Four of the 16 turned out to be off-category: a technology publication (BizTech Magazine), a sheet metal manufacturing software company (Lantek), a residential internet and TV provider (Blue Stream Fiber) and a business process outsourcer (Netfor). We left them in rather than quietly removing them, and the table lets you read around them. The average would look somewhat different with a different list.

The blocked sites. Seven of 23 domains, 30%, blocked our crawler. We did not score them. Because the crawler is only one kind of automated visitor, this tells us nothing about whether AI assistants can read those sites, and we draw no conclusion about their visibility.

Branded queries. Because the classifier writes queries from each homepage, some queries contain the company's own name. Softchoice is the clear case: four of its five queries named it, and its 80% is included in the 23.8% headline average. This affects the headline figure. We have not removed or re-run it, and a re-run on a cleaner list is on our list. Recomputed from the table above, as arithmetic and not a re-run: without Softchoice the average is 20.0% (15 companies, 75 queries); without the four off-category entries it is 21.7% (12 companies); without both it is 16.4% (11 companies). The conclusion does not change: most of these companies were rarely named.

What it does not cover. Only the homepage was read, so a company with a thin homepage and strong inner pages may be undercounted. Five queries per company is a small sample of real buyer language, and none of the queries targeted a specific sector such as healthcare or legal, so we cannot say how vertical specialization changes the result. Results come from one run on one day, and AI answers change over time. Citation counts of a few instances are small numbers. Own-site citations were matched by the company name appearing in the cited domain. We did not test whether any company pays for directory or list placement.

On claims and compliance. Nothing here is legal advice. Security and compliance claims should be evidenced, and your own counsel should make the final call.

FAQ

Related reading

• AEO, GEO and SEO for IT services companies and MSPs
• AI visibility research: every audit in one place
• What is answer engine optimization (AEO)? The complete guide
• Signal-based outreach: how it works