Grovia-Digital/ai-visibility-index

AI Visibility Index: which agencies do ChatGPT, Perplexity and Gemini recommend?

Open dataset by Grovia Digital (Dubai). We ask ChatGPT, Perplexity and Gemini the questions buyers ask when they look for an agency for AI search visibility (generative engine optimization, GEO), in German and English, and record which providers the assistants name and which sources they cite.

The same method runs every two weeks. Each new measurement is added here.

DOI (latest version): 10.5281/zenodo.23385485

What is in the data

Measurement	Date	Scope	Measurements
Baseline	2–9 Sep 2026	35 prompts, 3 platforms, 5 runs	525
Pulse P1	24 Sep 2026	10 prompts (block A) + 4 GCC prompts (block B), 3 runs	114
Pulse P2	8 Oct 2026	10 prompts (block A), 3 runs	90
Pulse P3	22 Oct 2026	10 prompts (block A) + 4 GCC prompts (block B, ChatGPT and Perplexity), 3 runs	114

Baseline in numbers: 596 firms and 97 software tools named, 3,335 source rows from 1,044 domains. The most-named provider reached 2.2 % of all mentions.

Grovia Digital measures itself with the same method: 0 mentions in the 525 baseline answers. In Pulse P2 (8 Oct 2026), Perplexity named Grovia Digital on one buyer question in 2 of 3 runs. In Pulse P3 (22 Oct 2026), ChatGPT named it in 6 of 12 measurements on the GCC prompts (block B), on "ai visibility agency uae" in 3 of 3 runs. In P3 block A (90 measurements) and on Perplexity in block B (12) it was not named. Gemini has not named it in any measurement so far. Details per pulse: CHANGELOG in the source repository.

Load the data

from datasets import load_dataset

measurements = load_dataset("Grovia-Digital/ai-visibility-index", "baseline_measurements")
sources = load_dataset("Grovia-Digital/ai-visibility-index", "baseline_sources")
pulses = load_dataset("Grovia-Digital/ai-visibility-index", "pulse_measurements")

Available configurations: baseline_measurements (default), baseline_sources, baseline_providers, prompts, limitations, pulse_measurements, pulse_sources.

Files

All files are UTF-8, semicolon-separated.

data/baseline-2026-09-10/

File	Content
..._prompts.csv	prompt ID, language, category, exact wording (frozen)
..._measurements.csv	one row per answer: date, time (CEST where recorded), platform, model, run, location/VPN, providers and tools named, whether sources were captured
..._sources.csv	one row per cited source: measurement ID, domain, URL, source type, position
..._providers.csv	providers in alphabetical order with the number of measurements naming them
..._limitations.csv	every declared limitation of the measurement

data/pulses/

File	Content
..._P1_measurements.csv, ..._P2_measurements.csv, ..._P3_measurements.csv	same columns as the baseline plus pulse, block, grovia_rank, map, paid_placement
..._P1_sources.csv, ..._P2_sources.csv, ..._P3_sources.csv	cited sources per measurement (P2 and P3: Perplexity only)

Method

• Frozen prompt set, exact wording, new empty chat per measurement, web search on, memory and personalisation off.
• VPN exit Frankfurt. Times in CEST.
• A provider counts once per answer. Spelling variants of the same provider are merged. Names of individual persons are replaced by "Individual consultant (name removed)". Software tools are counted separately from agencies.
• Pulses (3 runs) are compared only with runs 1–3 of the baseline, never with the five-run figures.
• Gemini shows sources as display names without URL, so domain frequencies should be read per platform.
• Mention frequency is not a statement about quality.

All limitations are listed in the limitations configuration.

Use and citation

Licence: CC BY 4.0. You may use, share and adapt the data, including commercially, with attribution:

Grovia Digital (Grovia Management Consultancy L.L.C), Knodel, A., Knodel, K. (2026). AI Visibility Index. Zenodo. https://doi.org/10.5281/zenodo.23385485

@dataset{grovia_ai_visibility_index_2026,
  author    = {{Grovia Digital (Grovia Management Consultancy L.L.C)} and Knodel, Alexander and Knodel, Kristina},
  title     = {AI Visibility Index: which agencies do ChatGPT, Perplexity and Gemini recommend?},
  year      = {2026},
  publisher = {Zenodo},
  doi       = {10.5281/zenodo.23385485},
  url       = {https://doi.org/10.5281/zenodo.23385485}
}

Links

• Live index with the measurement diary: https://www.grovia-digital.com/en/geo-visibility-index/
• Case: how Perplexity started naming Grovia Digital: https://www.grovia-digital.com/en/case-studies/grovia-named-by-perplexity/
• Source repository and changelog: https://github.com/Grovia-Digital/grovia-ai-visibility-index
• Also on Kaggle: https://www.kaggle.com/datasets/groviadigital/ai-visibility-index
• Grovia Digital: https://www.grovia-digital.com/en/