Named processes (not generic clichés)
The processes below are not generic "best practice" phrases; they are defined, repeated methods within TYS, supported by our codebase.
| Process | What it does | Evidence / mechanics |
|---|---|---|
| AI Authority Score | Measures a brand’s AI visibility with a 100-point, repeatable anchor. | Three fixed-weight categories: Entity 35, Citability 30, AI Visibility 35. Four engines: ChatGPT, Gemini, Claude, DeepSeek. Live queries archived with screen evidence. Every report states the score is a point-in-time, methodology-dependent measurement. |
| SEO-safe release and fast rollback | Tightens SEO checks at every release; prevents critical errors from staying live. | A pre-release checklist plus fast return to a stable version on error. The goal: bots see the correct version quickly, and critical SEO errors are rolled back before they spread. |
| BrandLock canon | Defines the brand’s verified, machine-readable truth and corrects misrepresentation. | Canonical signals (Organization/Person, sameAs, schema.org, llms.txt) are aligned with the brand’s truth. Boundary honesty: AI output is not edited directly, the signals it feeds on are aligned. |
| Change log and human approval | Logs every change; publishes no report without human approval. | A record of what changed, why and when. Even when report text is AI-drafted, it passes human approval; there are no auto-published metrics. |
| Monthly reporting framework | Delivers measured results in a standard structure every month. | Score, E-E-A-T, AI Share-of-Voice, sentiment, recommendations, and at least one concrete fix plus one implemented improvement. Classic search context when GSC/GA4 are connected. |
| 90-day proof window | Produces measured results before any long commitment. | A Day 0 baseline, a Day 90 re-measurement with the same methodology, and a sourced, screen-evidenced delta. The continuation decision is the client’s. |
Evidence and case-study methodology
Every number in our case studies was measured for that client and is labeled with an evidence type: research, experience or hybrid. We do not turn one client’s outcome into another client’s promise.
The basis for numbers is the baseline-then-measure principle: a baseline first, then a re-measurement with the same methodology, and the delta reported in the client’s own data. No growth figure is quoted before a baseline exists.
As the published case list grows, the principle does not change; where a situation calls for verifiability, reference requests are welcome. For current cases, see the case studies page.
The measurement and reporting framework is open
The AI Authority Score weighting (Entity 35, Citability 30, AI Visibility 35) and the four-engine query set are documented openly; this is a deliberate choice to close the "no standard measurement framework" perception.
It is stated plainly that the score is stochastic, meaning AI answers can vary by nature. That is why repeated measurement, not a single query, is the basis, and every report notes this condition.
The reporting rhythm is fixed: a monthly cadence, standard sections, human verification. To reduce expectation uncertainty, we state clearly what we measure and what we do not (below).
Data, security, GDPR and SLA
Policy and legal texts are published on separate, verifiable pages; this page is their reference hub.
Timeline and honest expectation
SEO and AI visibility impact is not instant but medium-term; the first measurable movement usually appears within weeks, and the matured result within the 90-day window. We document this as a design choice, not a flaw.
Short-lived spikes are easy to produce but do not last. The model targets sustainable visibility over short spikes: value compounds through a concrete fix plus an implemented improvement every month.
That is why success is defined by the measured difference between Day 0 and Day 90, not by an invented target like "so many points in 90 days". For teams that want speed without sacrificing durability, this is the right frame.
We also document what we do not measure or promise
- No ranking or score guarantees; we report the measured delta, whatever it is.
- We cannot directly change AI output; we align the canonical signals AI feeds on with the verified truth of the brand.
- Core Web Vitals field data requires a separate measurement pipeline; no performance commitment is made before that pipeline exists.
- This is not a self-serve software subscription; reports pass human verification and billing is a manual process.
- Our capacity is deliberately limited; we do not sell when capacity is unavailable.
Frequently asked questions
Is there public documentation for TYS?
Yes, this page is it. Processes, measurement weights, the reporting framework and the data approach are documented openly here; legal and policy texts are published on separate verifiable pages (privacy, terms, legal notice, how we work) and linked from here.
How is the AI Authority Score calculated?
Three categories are summed at fixed weights: Entity 35, Citability 30, AI Visibility 35, for 100 points. Live queries run across four engines (ChatGPT, Gemini, Claude, DeepSeek) and are archived with screen evidence. Every report states the score is a point-in-time, methodology-dependent measurement and that repeated measurement is the basis.
Why do you not publish ready, striking case numbers?
Because a number is only meaningful if it was measured. Our case numbers are measured for the client in question and labeled by evidence type; we do not convert one client’s outcome into a general promise. Where verifiability is needed, we accommodate reference requests.
Are your processes specific, or generic best practice?
Specific. The AI Authority Score, SEO-safe release and fast rollback, the BrandLock canon, change logging and human approval, the monthly reporting framework and the 90-day proof window are named, repeated processes; they are listed with their mechanics in the table above.
How do you access my data and analytics accounts?
If Google Search Console and GA4 are connected, only read access is used. Hosting is in Germany/EU (Hetzner). What data we process and why is documented on the privacy page; commercial terms on the terms page.
Why are results not instant?
SEO and AI visibility are a medium-term compounding effect; the first movement is measured within weeks and the mature result within the 90-day window. The target is sustainable visibility over a short spike; success is defined by the measured Day 0 to Day 90 difference, not an invented target.