Technical readiness and being used as a source by AI
The most striking finding in the dataset — and the only area where technology alone makes a noticeable difference.
Among the 822 companies with technical code GEO-score 85, 54.5 per cent were used as a source in at least one AI response. Among the companies in the 55–75 score range, the figure was 24.0–39.4 per cent. The difference is statistically significant (p < 0.0001), but the effect is moderate — technical expertise explains part of the difference, but not all of it.
The three figures that support the finding
All score bands with a sufficient number of companies
Note step 100: it is LOWER than step 85. The relationship is clear, but it is not a straight line.
The table shows the 1771 for the 1912 companies. The remaining 141 (7.4 %) are distributed across score bands with fewer than 40 companies — the cells are too small to be meaningful, so they have been omitted from the table. The statistics further down, however, are calculated on the basis of all 1912.
| Technical GEO-score | Companies | Predicted the outcome | Rate |
|---|---|---|---|
| 55 | 88 | 25 | 28.4 % |
| 60 | 151 | 53 | 35.1 % |
| 65 | 75 | 18 | 24.0 % |
| 70 | 289 | 114 | 39.4 % |
| 75 | 115 | 39 | 33.9 % |
| 80 | 83 | 30 | 36.1 % |
| 85 | 822 | 448 | 54.5 % |
| 100 | 148 | 70 | 47.3 % |
What you can use the figure for
Technical aspects are the part of AI visibility that you have control over. You cannot determine what the AI thinks of your business, but you can determine whether your page can even be read, crawled and understood by the systems that compile the results.
The figure shows how much you get out of that work: a company at the bottom end of the technical spectrum is cited as a source in around one in three responses, whilst one with sound technical capabilities is cited in more than one in two. It’s not a guarantee — it’s simply doubling the odds of something that costs working hours, not advertising spend.
What the figure doesn’t reveal
We set out the reservations on this page, not in a footnote at the bottom.
- The relationship is not a straight line. The 148 companies with a top score of 100 were cited as a source in 47.3 per cent of cases — that is, less often than the group with a score of 85. A perfect technical score is no guarantee of anything.
- The effect size is 0.21 on a scale from 0 to 1. This means that if you were to select a random cited company and a random non-cited company, the cited company would have the best technical score around 6 times out of 10 — not 10 out of 10.
- It is a correlation, not a cause. We check whether the company had any chance at all of being mentioned — not the size of the company or how well-known the brand already is. A large, well-known company often has both better technology and a better reputation.
- The citations are measured using Perplexity, which is the only search engine that displays actual sources. Other search engines may behave differently.
How the figure was arrived at
1912 Companies from 15 industry panels in Denmark, Norway and Sweden, as measured between 2026-07-26 and 2026-08-24. The technical score is calculated deterministically without any AI assessment. The outcome is measured based on correct AI responses to unbranded questions.
The test used is the Mann-Whitney U test, with corrections for ties and continuity; this has been cross-checked against a permutation test with 10000 rounds and a fixed starting value, so that the result can be replicated.
Mann-Whitney U: z = 8.39, p < 0,0001, rank-biserial effect size 0.214. The median technical score was 85 amongst the cited articles and 75 amongst the non-cited articles.
View your own technical score
The same measurement as in the survey, on your own domain.
Assess my business for free