Benchmarks · B2B SaaS

DevTools & Cloud: AI readiness benchmark

48 platforms scanned on July 31, 2026. The average score is 74 out of 100. Homepages only, from a hand-picked list.

48sites scanned
74average score
40–100score range
Jul 31, 2026scan date

Key findings

  1. 1

    The average score in the DevTools & Cloud category is 74/100, 1 point above the all-category average of 73.

  2. 2

    Of the three pillars, Technical foundation is the strongest in the DevTools & Cloud category at 17/20 (85% of the maximum), and Discoverability is the weakest at 28/40 (70%).

  3. 3

    In the DevTools & Cloud category, Content schema (Article, FAQ…) is the most-failed check: 90% of sites did not earn full points on it.

Where the scores fall

The lowest score is 40, the highest is 100 and the average is 74, out of 100.

Fig. 1

Score range for DevTools & Cloud DevTools & Cloud sites range from 40 to 100 out of 100, averaging 74, across 48 sites. DevTools & Cloud · 48 platforms 74 40 100 0 25 50 75 100 AI visibility score (0–100)
The bar spans the lowest to the highest score in the category. The marker is the average.

Source: Citehound scan, July 31, 2026, 48 platforms

The three pillars

Each score is built from three pillars worth 40, 20 and 40 points. These are the category averages.

Fig. 2

Pillar averages against their maximums Average score per pillar for DevTools & Cloud. Discoverability, maximum 40 points: 28. Technical foundation, maximum 20 points: 17. Content & trust, maximum 40 points: 29. Discoverability · max 40 28/40 Technical foundation · max 20 17/20 Content & trust · max 40 29/40 Filled = points earned. Unfilled = gap to maximum.
Filled is the average points earned; the track is the maximum for that pillar.

Source: Citehound scan, July 31, 2026, 48 platforms

The checks sites miss most

Share of sites that did not earn full points, for the 8 most-failed checks.

Fig. 3

The 8 most-failed checks for DevTools & Cloud Share of sites that did not earn full points, most-failed first. Content schema (Article, FAQ…): 90%; Organization / WebSite schema: 40%; llms.txt present: 33%; Structured data (JSON-LD): 31%; Meta description: 29%; Single H1 heading: 21%; Author / about signals: 13%; Canonical tag: 10%. Content schema (Article, FAQ…) 90% Organization / WebSite schema 40% llms.txt present 33% Structured data (JSON-LD) 31% Meta description 29% Single H1 heading 21% Author / about signals 13% Canonical tag 10%
Ranked from most to least failed. A site fails a check by earning less than its full points, which includes partial credit. AI crawler access is shown by crawler in the next figure instead. Checks tied at 10% are listed in scan order, so a tied check may fall just outside the eight.

Source: Citehound scan, July 31, 2026, 48 platforms

Which AI crawlers can get in

Counts of sites, for each of 10 AI crawlers, where it is open, limited or blocked.

Fig. 4

Access by AI crawler across 48 sites Number of DevTools & Cloud sites where each crawler is open, limited or blocked. GPTBot: 10 open, 38 limited, 0 blocked; ChatGPT-User: 10 open, 38 limited, 0 blocked; OAI-SearchBot: 10 open, 38 limited, 0 blocked; ClaudeBot: 10 open, 38 limited, 0 blocked; anthropic-ai: 10 open, 38 limited, 0 blocked; PerplexityBot: 10 open, 38 limited, 0 blocked; Perplexity-User: 10 open, 38 limited, 0 blocked; Google-Extended: 10 open, 38 limited, 0 blocked; CCBot: 10 open, 37 limited, 1 blocked; Bytespider: 10 open, 37 limited, 1 blocked. GPTBot 10 38 0 blocked ChatGPT-User 10 38 0 blocked OAI-SearchBot 10 38 0 blocked ClaudeBot 10 38 0 blocked anthropic-ai 10 38 0 blocked PerplexityBot 10 38 0 blocked Perplexity-User 10 38 0 blocked Google-Extended 10 38 0 blocked CCBot 10 37 1 blocked Bytespider 10 37 1 blocked Open Limited Blocked
Open: no rule keeps the crawler out. Limited: at least one Disallow rule applies to the crawler, mostly an ordinary path such as /wp-admin/, and the rest of the site is open to it. Limited is not a block, although the scan gives it half credit. Blocked: the whole site is disallowed.

Source: Citehound scan, July 31, 2026, 48 platforms

Method and limits

What we scanned

48 platforms in this category completed a scan on July 31, 2026. 4 more on the list could not be reached or timed out. They are excluded from every figure on this page.

What we pulled

For each site: robots.txt, sitemap.xml, llms.txt and the homepage, retrieved live. Homepages only. Inner pages were not scanned, and each site was scanned once.

What we scored

One score out of 100 per site, split across three pillars: discoverability (40), technical foundation (20), content & trust (40). The checks and their point values are on the methodology page.

How the sites were chosen

The lists are hand-picked, well-known sites in the category. They are not a random sample, so these figures describe the sites we scanned, not the category as a whole.

Repeat scans

In one October 2026 rescan of the same lists, category averages moved by one point or less. Individual sites moved more: about a third changed by 3 to 15 points.

What this measures

Whether AI systems can reach a site, and whether its homepage carries the signals machines read. It does not measure whether any assistant names a brand in an answer. Model responses vary by prompt, session and training data, and no scan of public files can predict them.

What this doesn’t tell you

  • Anything about pages other than the homepage.
  • Why the unreachable sites could not be scanned, or how they would have scored.
  • Whether any of these signals relates to being named in an AI answer. This page measures inputs, not outcomes.

How to cite this

Citehound, “AI readiness benchmark: DevTools & Cloud,” July 31, 2026. https://getcitehound.com/benchmarks/devtools

Canonical URL: https://getcitehound.com/benchmarks/devtools

To see where your own site lands against these numbers, run the free scan. All benchmarks.

Scan your site free