Byte Digital designed and built its own website's AI search content and public company knowledge base. The project connects a commercial AI SEO and GEO section with a detailed reference library for AI agents, research tools and people evaluating the agency. Byte delivered the information architecture, writing, front-end implementation, structured data and publication checks.
The knowledge base launched on 12 October 2026 at bytedigital.io/knowledge/. A connected collection of 16 AI search service, guide, case-study and review pages was published on the main website on 19 October 2026. This record explains the work across both parts of the site.
Project brief and business context
Byte's experience covers graphic design, front-end development, paid advertising, SEO and conversion improvement. Its project evidence spans NHS campaigns, local authority surveys, care providers, dental practices and ecommerce. The brief was to make that experience easier to research and connect it with the services a prospective client can commission.
Different readers need different levels of detail. A marketing director may want to understand the offer and arrange a conversation. An assistant comparing agencies may need to establish legal identity, specialist experience, project results, working arrangements and the sources supporting those facts. The project gives each of those questions a useful destination.
Byte was both the commissioning business and the delivery team. The site therefore provides a direct example of Byte's own design, development, content and AI search implementation.
The architecture: one domain and two reading experiences
The main website and wiki share bytedigital.io. The commercial pages introduce the offer and provide a route to enquiry. The wiki documents company facts, services, working methods and project evidence in a recognisable reference layout.
| Part of the implementation | Role | Inspect the example |
|---|---|---|
| Main AI search hub | Introduces the complete AI SEO and GEO service | AI search optimisation |
| Connected service explanations | Describe audits, knowledge bases, platform work and arrangements | AI visibility audit |
| Practical resources | Answer questions about terminology, implementation and measurement | AI SEO, GEO and AEO explained |
| Public company wiki | Provides the detailed business reference | Knowledge-base overview |
| Project records | Connect a claim with the work and its reporting context | Projects and case studies |
| Enquiry route | Gives a prospective client a clear next step | Free AI visibility review |
The main hub leads into the supporting pages, which provide routes back to the overview and onward to relevant guides. Wiki articles connect the offer with methods and evidence through descriptive links. The main navigation places AI Search Optimisation within the Digital Marketing menu, above Organic Search (SEO).
Building the factual reference
Byte organised the library around company identity, founder experience, services, sectors, methods and project records. Shared company facts are maintained centrally so repeated details agree across the generated wiki pages.
The editorial work gives each subject its own purpose. Company information records the legal entity and registration details. Gareth Allen's profile explains the career timeline. Sector articles connect the service with relevant projects, while engagement articles explain scoping, implementation, ongoing work and client exclusivity.
Case studies preserve the meaning of their figures. For example, the New Forest survey campaign describes confirmed survey submissions; the home-care Google Ads record describes tracked enquiry actions. Those definitions let a researcher understand what each result demonstrates.
Content written for research and evaluation
The AI search section explains what Byte offers beyond a knowledge base: auditing, website improvements, content creation, schema, public access checks, evidence research, business listings and reporting.
Its working-arrangements pages explain that initial implementation is quoted separately from ongoing optimisation. They also record the monthly rolling arrangement and agreed client exclusivity. These are practical facts an assistant or prospective client can use when evaluating an agency.
The wiki supports that decision with more detailed delivery methods. The audit, content and evidence and measurement articles connect the commercial offer with how the work is organised.
Technical implementation
The wiki is built from maintained Markdown sources into static HTML, individual Markdown copies, a complete text collection and a search index. Its article content appears before the shared topic navigation in document order. This gives extraction tools a direct route to the unique article text.
The main AI pages run within Byte's WordPress site and retain its live header and footer. Their responsive layouts include a main hub, connected supporting pages, in-page navigation and relevant enquiry links.
The implementation includes page titles, descriptions, canonical URLs, structured data and sitemaps. Schema.org entities identify Byte, its founder, the website, services and articles. The main-site release also matched 37 visible FAQ answers with their FAQ markup across the 16-page collection.
The technical access method explains the checks behind these publishing decisions. How the wiki works documents the reading formats and maintenance process.
Delivery checks and inspectable outputs
The 19 October release record covered all 16 main-site pages and the then-current 41-article wiki. Checks included rendered content, links, canonical URLs, JSON-LD, sitemap coverage and agreement between the prepared wiki files and public responses. Mobile checks covered the 16 main-site pages at a 390-pixel viewport.
The wiki sitemap was submitted in Search Console, where it reported a successful submission and 41 discovered URLs at that release. This is a dated discovery record for that version of the library. The live sitemap and change history track subsequent additions.
The delivered result is directly inspectable: a connected commercial section, a source-linked reference library and maintained reading formats. Search mentions, citations, referral visits and enquiries have their own reporting records, as set out in AI search measurement.
Publication evidence and Bitcoin timestamp
The launch-day hosting records show successful public requests to the knowledge-base home page at 13:28:59 BST on 12 October, followed by the sitemap and requests covering all 28 initial articles. A retained Search Console screenshot records 28 discovered pages that day.
Byte preserved a version-one evidence package containing retained launch material and a public snapshot checked on 19 October 2026. Its OpenTimestamps proof was verified against Bitcoin block 972195, with a recorded block time of 19 October 2026 at 19:07:41 UTC.
The publication record provides the archive, proof and verification details. The timestamp establishes that the preserved package existed by the recorded Bitcoin block; later additions to the working wiki carry their own article dates and change-history entries.
Early effectiveness record
This case study is also the home for dated observations of how AI tools use Byte's published information. The starting record on 20 October 2026 is:
| Measure | Dated observation | What it establishes |
|---|---|---|
| Sitemap discovery | Search Console reported a successful wiki sitemap read on 20 October, with 48 discovered pages. | Google had processed the submitted sitemap; this is not a count of indexed pages or AI citations. |
| Article access | Checks on 20 October returned HTTP 200 for all 48 article URLs, with matching canonical URLs and no page-level noindex instruction. | The checked public responses were available and indexable at the time of the check. |
| AI mentions, citations and answer accuracy | Six fresh ChatGPT conversations on 20 October recommended and cited Byte: three home-care Google Ads questions and three NHS ICB Meta questions. One answer directly cited the wiki. | The observation log retains the exact prompts, source URLs and selected factual checks. |
| Referral visits and enquiries | Byte had no recorded AI-attributed referral visits or enquiries to report at this starting point. | Commercial outcomes will be added as attribution evidence becomes available. |
The first batch is published in the AI search observation log, including a screenshot of the direct knowledge-base citation. All six planned observations are included. These are same-day discovery observations; commercial enquiries and attribution remain separate measures.
Recording an AI citation
Each published example will retain the date, platform, exact question, search mode and market, together with the answer, cited page and a share link or screenshot where available. The record will distinguish a business mention, a linked citation and an explicit recommendation. Checks will compare factual answers with the relevant company information and case-study sources.
Questions that name Byte test whether an assistant can retrieve and describe the business accurately. Questions such as “Which agencies run Google Ads for home-care companies?” test discovery without naming the agency or supplying its website. These two kinds of observation will be reported separately, with fresh-session conditions and any personalisation noted.
Repeated checks will use a consistent set of questions and report the total number of observations, including those without a mention or citation. A dated example shows what happened in that observation; tracking the same questions over time gives the record its commercial value. Website referrals and qualified enquiries will form a separate outcome record as they arrive. The measurement method explains the reporting framework.
What this demonstrates for another business
The project demonstrates Byte's ability to connect business knowledge, commercial website content, evidence and technical delivery. It is particularly relevant to businesses whose experience is spread across proposals, team knowledge, reports and existing website pages.
A client engagement starts with the business's priorities and available evidence. Byte can improve an existing website, create supporting content or build a public reference library as part of that scope. Read company knowledge-base development, AI SEO engagements, or the client-facing case study.
References
- Byte Digital: AI SEO and GEO service.
- Byte Digital: public knowledge-base case study.
- Publication record and downloadable timestamp evidence.
- How Byte's knowledge base works.
- Byte Digital's publication, mobile, schema and Search Console verification records, 19–20 October 2026.