Process
How We Work
We measure first, then execute. At every step we show what we do, what we need from you and how you can verify the result.
Seven steps
From a first audit to monthly reporting.
Intake and inputs
We take your website address, target topics and competitors. We choose together which languages to work in (Turkish, English or both).
Prompt universe
We prepare Turkish and English buyer questions by persona (buyer, investor, patient, importer) and intent (buying, comparison, reputation, problem-solving, risk validation).
Runs
We run the questions across several major AI assistants. For each answer we record whether the brand is named, how it is described, and which competitors and sources appear.
Analysis
We answer: who is winning, who is invisible, why AI recommends certain companies, and what can be fixed.
Action plan
We prioritize source and content gaps and produce a 30-day plan you can act on.
Execution
We carry out the plan with content, technical fixes and digital PR. You approve every piece.
Monthly reporting
Depending on scope, we show progress in the monthly report below.
Transparency
You should be able to see the results and check them yourself.
We give you a dashboard and the raw data: every prompt asked and every answer the AI gave. You can type the same questions into ChatGPT or Gemini and compare. AI answers change with every run and every engine, so we look at directional trends across many prompts and several engines, not at a single answer.
The pilot’s own deliverables and terms are set out in detail on the pilot page.
What does the monthly report contain?
Ten items each month in the ongoing program.
Executive summary
A short summary of the month and key findings.
AI answer share movement
The change in your share across tracked prompts.
Model-by-model visibility changes
The trend on each engine.
Top winning and top missing prompts
The main questions where you appear and where you do not.
Competitor movement
Changes in how competitors are named.
Citation source changes
Movement in the sources feeding the answers.
Reputation sentiment movement
The change in how your brand is described.
Content and PR assets published
The pieces published that month.
Technical improvements completed
The technical fixes made.
Next month action plan
The priorities for the coming month.
What do we need from you?
Three things are enough to start.
- Your website address.
- Your target topics and competitors.
- The working language: Turkish, English or both.
In the pilot we ask for about 30 to 35 minutes of your time in total (the kickoff call and approvals). The pilot’s first-payment term is on the pilot page.
What we do not promise
- We do not promise rankings. AI answers change with every run and every engine.
- No engine publishes a complete rank or share metric, so we report directional trends across many prompts and several engines.
- GEO lifts reported in research are not a promise.
- We publish nothing without your approval: you approve every piece we publish.
Frequently asked
How long does the process take?
How do I verify the results?
Will anything be published without my approval?
Where does your brand stand in AI answers?
Send us your website and topics and we will reply by email.