Audit, diagnose, optimize, build, monitor.
Every engagement runs the same five steps. We measure where you stand across search and AI, and work out what is holding you back. Then we fix it, build what is missing, and track both channels every week: search, Google AI Overviews and Google AI Mode, and ChatGPT, Perplexity, Gemini, Copilot and Claude.
How it works
Every engagement runs the same five steps, in the same order.
- 01
Audit
We measure your current SEO and AI search visibility, and record what each system says about you today.
- 02
Diagnose
We identify the technical, content and authority gaps holding you back, and rank them by what they cost you.
- 03
Optimize
We improve your website, content, entities and technical foundation so both channels can read you correctly.
- 04
Build
We create the content and earn the external sources required to close the gaps the diagnosis found.
- 05
Monitor
We track your visibility across search and AI every week, and tell you what changed and what to do about it.
An illustrative example
What the work looks like.
No client has agreed to publish figures. This is the sequence an engagement runs, written out.
Before
Which firms should we shortlist for this?
Three named. Not you.
What the measurement found
- Competitors held stronger category pages
- Third-party publications cited them and not the client
- The client's entity information was ambiguous
- Several buying questions had no answer anywhere the client owned
- Structured data was incomplete
What was changed
- Rebuilt the information architecture
- Completed the structured data and entity definitions
- Created the missing category and comparison pages
- Built the supporting material an answer can cite
- Kept the same question set under weekly measurement
The client begins appearing in the answer, and the record shows when it started and which source moved first.
Illustration — not client data.
One pipeline, two channels, run every week.
Step 05, monitor, runs on this pipeline for as long as the engagement continues.
The system
Visibility you can watch, not a report you wait for.
- Prompt bank
- A custom bank of questions, written specifically for your engagement: the ones a prospect would actually type into a search bar, or ask an assistant, before deciding.
- Scheduled runs
- We put your questions to search and the assistants live, every week, and keep every result.
- Analysis
- Raw results from both channels become share of voice, sentiment, and cited sources, the most actionable of the three.
- The record
- A record that accumulates, not a document you wait for.
Layer 01
Layer 02
Layer 03
Layer 04
Questions
- Why measure directly instead of using an off-the-shelf AI visibility tool?
- We write a question bank for your specific category, then run it live against search and the assistants every week, keeping every result. That record reflects what is actually said about your company, to the questions your own buyers actually ask.
- How often is it measured, and why every week?
- Every week, across search and AI. Rankings move and model answers change as providers update them and as the sources behind an answer shift, so a single measurement only captures one moment. A weekly cadence catches a change early enough to act on it, and builds a trend line that means something.
- What is actually reported each week?
- Whether you appear at all, in Google's results and in the assistants' answers. Beyond that: how you are described, who appears alongside you, and which sources the answer draws from. Those are what can be changed; a position number alone is not.
- Do the five steps always run in this order?
- Yes, on a new engagement. Audit and diagnose set the priorities, optimize and build carry them out, and monitor tracks the result every week afterward. A client who only wants monitoring can start there instead.
See what Google and AI already say about you.
We check your visibility across search and the assistants we cover, and show you where you are found, where you are missing, and which sources the answers are built from.