01
Entity consistency
The same identity, description and details across your site, schema, directories and profiles.
Grow · 03.3
Become the source the answer is built from.
When an engine cites its sources, your site should be one of them. That takes content that is readable, trustworthy and easy to extract without losing its meaning.
What's the best invoicing software for freelancers?
[1] yourcompany.com[2] review-site.com
Cited in 7 of 20 prompts
01 / What it involves
Four areas, two technical, two editorial.
Pick the areas you want to work on.
02
Machine-readable files
Page-aligned schema, current sitemaps and an llms.txt generated from your content.
03
Quotable structure
Self-contained passages, clear headings and one hierarchy for people and parsers.
04
Citation tracking
A fixed prompt set checked across several engines on an agreed schedule.
02 / Included
What's included.
Most of the gain comes from reorganizing the content you already have.
Real questions
The prompt set is based on how your customers ask, not on a keyword tool.
Citation baseline
A starting point on each engine to measure against.
Competitor citations
Which rivals appear instead of you, and how their pages differ.
Restructured content
Existing content reshaped into self-contained answers before anything new is written.
Off-site consistency
Directories and profiles corrected, because engines read beyond your website.
Monthly report
Which questions name you, which don't and what changed, in plain language.
03 / What to watch
What to be skeptical of.
Each risk, and how we avoid it.
×Citation guarantees
No one controls the model's decision. We remove the reasons not to cite you and measure the results.
×Tactics sold as secrets
Nearly everything that works is public and structural.
×Advice that ignores the technical foundation
We review the technical foundation first.
×Tools that measure one engine
We measure several engines with the same prompt set.
×Claims without a baseline
Everything, including our own claims, is compared to the starting point.
04 / How it works
How it works.
Our usual way of working. The quote sets the real dates before work starts.
Baseline
Define the prompt set, test it across engines and review your entity wherever it appears. Deliverables: citation baseline and entity findings.
Foundations
Schema, llms.txt, sitemaps and fixes for the technical blockers that prevent a clean read. Deliverables: files published and schema in production.
Restructure
Existing content reorganized into self-contained answers, merging or splitting pages around real questions. Deliverables: restructured pages and a content model.
Measure
Rerun the prompt set, compare against the baseline and decide next month's work. Deliverables: monthly report and next month's plan.
05 / Questions
Questions we get asked
They share a technical foundation, but the content structure, the measurement and the goal are different.
No. We remove the reasons not to cite you and we measure. If nothing changes after six months, we report that as a finding.
ChatGPT, Perplexity, Gemini, Copilot and Google's AI answers, adjustable with you.
Three to six months on a sound foundation; longer if the site needs a rebuild first.
Not necessarily at first. Publishing matters more once the foundation is in place.
Find out where you stand.
Send us your URL. We'll tell you which searches you should own, which answers leave you out and what would change that.