The answer engine optimization agency for growth-stage brands

How an AEO engagement runs

The work follows the same seven steps for every client. What changes is where the effort goes, because the audit decides that.

1. 1. Build the real prompt set

   We start from the questions your buyers ask AI, seeded from the questions your sales team hears on real deals. Then we expand across the buying journey: best in category, best for a use case, competitor alternatives, is this brand worth it. Most clients end up with 80 to 150 tracked prompts.

2. 2. Baseline every prompt across engines

   We run the full set through ChatGPT, Perplexity, Gemini and Google AI Overviews and log four things per prompt: whether you are cited, at what position, which competitors are cited, and the exact source URLs the engine pulled from. That last column drives everything after it.

3. 3. Diagnose why you are not cited

   It is almost always one of four causes. You are missing from the third-party sources the engine cites for your category. The engine is not sure what your brand is. Your content exists but is not in a form a model can retrieve and quote. Or AI crawlers are blocked at robots.txt or the CDN and nobody noticed.

4. 4. Fix the site

   We build or rework the pages models quote: direct-answer comparison pages, best-for-segment pages and clear definitional pages, each written so one passage can be lifted cleanly. We add Organization, Product and FAQ schema and make the brand description consistent across the site and your third-party profiles.

5. 5. Fix what is off the site

   Using the source list from the baseline, we go after the specific domains the engines already trust in your category: digital PR, contributed expert content, inclusion in third-party best-of roundups, and original data that other writers reference.

6. 6. Weight the levers per engine

   The engines retrieve differently. Google AI Overviews leans on what already ranks in classic search and on strong entity signals, so technical and on-site work counts most there. ChatGPT and Perplexity lean on retrieved third-party sources, so off-site presence counts most there.

7. 7. Re-run and compound

   We re-run the prompt set on a schedule, track citation share and position per engine, and put more into whatever moved.

Why most of the work happens off your website

We track client brands in AI engines every day. For a study we took 90 days of citations in three unrelated verticals and sorted the cited domains by type. The brand's own website was 2 to 6 percent of what the engines cited. Third-party editorial, community and reference sites made up 26 to 59 percent.

So an agency that only edits your own pages is working on a small slice of the answer. Our engagements put most of the effort where the citations come from, and the audit tells us which third-party sources those are for your category.

The method and tables are in our AI citation sources study, linked below.

Getting you onto the lists AI reads

The biggest single lever off your site is the best-of list. Ahrefs found that about 43.8 percent of the pages ChatGPT cites are listicles, and when we asked three engines to recommend agencies in our own category, the sources were mostly other companies' roundups.

So listicle placement is a standing part of every engagement. We map the lists the engines cite for your prompts, pitch the publishers, tell you where a placement is paid and what it costs, and publish fair comparison pages on your own site. Reddit and digital PR run beside it, because the engines read community threads and news coverage too.

Each of these has its own page: listicle placement, Reddit marketing, and digital PR and citation building, linked below.

What the first 30 days cover

Every engagement starts with a 30-day sprint scoped to your brand, market and goals. It ends with a baseline, a plan and the first fixes shipped. From there the work continues as a retainer, or your own team takes the plan forward.

• AI visibility audit and strategy
• Entity optimization and structured data
• Content production and prompt targeting
• An AI rank tracking dashboard for your prompt set
• Weekly reporting and strategy calls

How we report

Reporting is prompt-level. For each tracked prompt you see whether you were cited, where, who else was, and which sources the engine used, per engine and over time. Where a result can be shown publicly we capture it as an engine-hosted share link, so anyone can open the original answer on the engine's own domain and check it.

We also say what we cannot promise. AI answers are rebuilt on every query and the wording of a prompt changes the result. Nobody can guarantee a number one position, and an agency that does is selling something else.

Who we work with

Growth-stage brands with a real product and buyers who research before they purchase. Our client work covers DTC and ecommerce, B2B SaaS, healthcare, and multi-location services. If your category is one where buyers ask an AI assistant what to buy or which vendor to shortlist, this is the work.

If you are pre-product, or you need a guaranteed ranking by a fixed date, we are the wrong agency and will tell you so on the first call.