Because the first question to answer is "what's actually being said?"
You can't fix AI visibility without knowing what's currently broken. Most engagements pitched as "AI SEO" start by recommending content — without first auditing what models are saying. That's backwards. Half the time, the model is already saying something close to right; you just need to fill specific gaps. The other half, it's confidently fabricating things you'd never have known about.
The Quickstart spends week one finding out which scenario you're in. Week two delivers the first piece of the fix — and installs the instrument that makes the fix measurable.
"Every gap in your articulated content is an invitation for the model to invent something on your behalf, and you don't get to vet what it invents before it propagates."
The whole thing, two weeks, fixed fee.
Multi-provider audit (week 1)
We query OpenAI, Google's Gemini, and Google's AI Overview surface (via SerpApi) for the questions a real customer would ask about your business. We capture the response text, the cited sources, and three position metrics: where your brand appears in the answer text, where it appears in any recommendation list, and where it appears in the citation list. Each provider is a different game; you'll see all three.
Content-gap analysis
From the audit, we identify the specific places where the model is improvising — fabricating facts, omitting context, or anchoring to outdated/incorrect sources. We rank them by impact (how often they show up in queries that matter to your business) and produce a priority-ordered backlog of content interventions.
The first article, written in your voice
We write the highest-priority piece. Long-form, articulated, specific. Not generic SEO content — the kind of content that fills the exact gap the audit identified, in the language your model would quote. Drafted in your voice, reviewed with you, published on your site or sent for you to publish.
Monitoring installed
We set up continuous multi-provider scanning so you can watch the article land. Indexing latency for new long-form content is typically 48–72 hours from publish to AI-search surface — the monitoring shows you the before/after curve, in real time, week over week. Monitoring continues at £149/mo from week 3 onward (cancellable any time).
Two weeks, two milestones.
Audit + analysis
Multi-provider scanning across the three surfaces. Ranked content-gap report delivered end of week. Article topic and angle agreed.
Article + monitoring
First long-form article drafted, reviewed, finalised. Monitoring detector installed and running. Handoff with credentials and dashboard access.
£950, fixed. Then £149/mo to keep watching.
If you decide to commission a full WordPress rebuild after the Quickstart, the £950 audit fee is credited against the rebuild quote. Quickstart clients also get the monitoring at the lower starting tier regardless of whether you ever rebuild.
Be honest about whether this fits.
- → Your business has been around long enough that customers ask AI about you.
- → You don't yet know what the AI summary says.
- → You suspect inaccuracies but can't quantify them.
- → You can spare 2–4 hours of input over two weeks.
- → You want a measurable first win before committing further.
- · You launched last week and have no online footprint yet — wait six months.
- · You're hoping for "rank #1 in ChatGPT in 7 days" results — not how this works.
- · Your target audience doesn't use AI search (some niches still don't).
- · You aren't able to publish content on your own site in any form.