ChatGPT recommended Byte Digital in all six recorded agency-discovery answers on 20 October 2026: three for home-care Google Ads and three for NHS ICB Meta campaign support. Each answer cited a Byte source. One home-care answer cited the knowledge-base case study directly, alongside the main-site version.
This accumulating log records how AI assistants find and describe Byte. It connects each result with an exact question, observation time and supporting URLs. It forms part of the Byte website AI SEO and GEO case study.
First recorded results
| Question group | Completed observations | Byte recommended | Byte source cited | Wiki directly cited |
|---|---|---|---|---|
| Home-care Google Ads agency | 3 | 3 of 3 | 3 of 3 | 1 of 3 |
| NHS ICB Meta campaign agency | 3 | 3 of 3 | 3 of 3 | 0 of 3 |
| Total | 6 | 6 of 6 | 6 of 6 | 1 of 6 |
These are observed results for this question set on this date. The direct wiki citation establishes that ChatGPT referenced the wiki in that answer. The sample measures discovery and citation; it does not isolate the wiki's contribution from the main website or establish a fixed ranking.
Questions and test conditions
The two prompts below were each submitted once in three fresh ChatGPT conversations. Neither named Byte nor supplied a Byte URL. All six planned answers were retained, with no regeneration or follow-up to steer the result.
HC-01: home-care Google Ads
I run a home care company in the UK and need help getting enquiries through Google Ads. Which agencies would you recommend, and why? Please search the web and cite evidence of their relevant home-care experience.
NHS-01: NHS ICB Meta campaign support
I manage a communications team at an NHS ICB in England and need help running a Meta campaign. Which agencies would you recommend? Please search the web, explain their relevant NHS experience and cite sources.
The tests used ChatGPT's web interface while logged out, with a fresh conversation for each observation in the same browser session. Web search was requested in the prompts and the answers returned source citations. English and the UK or England were specified through the questions. The exact default model and physical IP location were not independently established. These are Byte-commissioned, AI-assisted checks; they are not a third-party audit or a platform endorsement.
Dated observation register
All times below are BST on 20 October 2026. The position is Byte's order in that particular answer's agency list, not a permanent search rank.
| Observation | Captured | Byte's list position | Supporting Byte sources |
|---|---|---|---|
| HC-01-R1 | 11:15:32 | 4 | Home-care Google Ads case study; Home-care marketing service |
| NHS-01-R1 | 11:16:30 | 2 | NHS 111 mental-health case study |
| HC-01-R2 | 11:17:42 | 4 | Home-care Google Ads case study; Home-care Google Ads wiki record |
| NHS-01-R2 | 11:18:19 | 2 | NHS 111 mental-health case study |
| HC-01-R3 | 11:20:06 | 3 | Home-care Google Ads case study; Care-provider Google Ads review |
| NHS-01-R3 | 11:23:55 | 2 | NHS 111 mental-health case study; NHS marketing service |
Direct citation of the wiki
In HC-01-R2, ChatGPT's source card cited the home-care Google Ads wiki article alongside the commercial case study. Opening the knowledge-base source confirmed the destination URL. The retained citation screenshot shows the Byte recommendation and the expanded source card.
The other five answers' checked Byte citations pointed to the main website. For NHS-01-R2, the answer also provided a general homepage link; its supporting citation resolved to the NHS 111 case study. The log distinguishes that navigation link from the source supporting the claim.
What the answers understood
The home-care answers identified Byte's Google Search work for an anonymised London provider and preserved the reporting distinction between tracked contact actions and unique enquiries or acquired care clients. The selected checks covered campaign scope, reporting period, the £55.71 cost per tracked enquiry action and its definition.
The NHS answers identified Byte's NHS 111 campaign and described relevant Facebook and Instagram delivery, creative testing and comparisons between areas with different deprivation levels. The third observation also recognised NHS App and childhood-vaccination examples. These selected factual descriptions agree with the NHS 111 project and NHS service page. The check covers these Byte text facts and sources, rather than every statement or visual in the generated answers.
This is useful commercial evidence: Byte appeared in answers to prospective buyers' questions, with relevant published work supporting the recommendations. Referral visits, enquiries and won work will be recorded separately when evidence is available. Links opened during these checks are verification traffic and are excluded from customer-attribution claims.
Evidence and continuing the log
The machine-readable observation record contains the prompts, UTC and BST capture times, per-answer results, citation URLs, factual checks and SHA-256 hashes of the retained original files. Byte retains the full response text, browser accessibility captures and screenshots for each observation. Conversation URLs are retained as session references; they are not presented as permanent public share links.
Future batches will append dated observations using the same core questions, including answers where Byte is absent. New questions and other platforms will have separate identifiers. Each batch will state its platform, test conditions and total observation count. Corrections will be dated, leaving the original capture intact. The measurement method explains the wider reporting approach.
These captures were made on 20 October 2026. They are new evidence and are outside the original 19 October Bitcoin timestamp.