Home SEO & AEO
SEO & AEO
Your customers still search Google, but more of them now ask ChatGPT, Gemini, or Perplexity what to buy. We fix the technical foundation, rewrite content so it gets quoted, build the outside signals AI tools trust, and track both every week.
Two jobs, one foundation
A brand can sit on page one of Google and still never come up when someone asks an AI tool for a recommendation. We work on both, and we measure them separately so you can see which one is moving.
The classic work, done properly: a site Google can crawl and trust, pages that match what people search, and fixes ranked by how much traffic they unblock.
Answer engine optimization (also called GEO) is about the answer, not the link. AI tools pull from pages they can parse cleanly and from what the rest of the web says about you.
How it works
A crawl of up to 500 pages, checked in parallel for technical health, content, schema, sitemaps, speed, mobile, images, and AI readiness. Speed is measured in a real browser on a mid-range phone, alongside real-user data from Google.
Every finding goes into an action plan ranked Critical to Low, with a realistic date for when you should see it in the data. Server errors and crawl blockers go first, because they cost rankings fastest.
Pages rewritten answer-first, with direct answers under each heading, FAQs, tables, and sourced numbers. For large blogs we run AI-assisted updates through a human review queue, so nothing publishes unchecked.
AI tools trust what others say about you more than what you say about yourself. We map where AI answers get their sources, find the Reddit threads and press they cite, and build your presence there.
Rankings every Monday in your own market. AI answers checked weekly, plus live tests across five AI tools once or twice a month. Search Console, analytics, and Business Profile data pulled straight from Google.
Search, traffic, rankings, Business Profiles, AI visibility, and the work done with evidence. Each report passes about 25 automated checks and a human review before you see it.
Two scores, never blended
Most "AI SEO" scores mix up whether a site is set up for AI with whether AI actually recommends it. We report them side by side, because the gap between the two is usually the most useful finding.
We ask AI tools the buying questions your customers type, never "what is [your brand]". Each answer earns full credit for a top pick, partial credit for an option, a little for a brief mention, and nothing if you're missing. Missing answers stay in the score, so it can't be padded.
A realistic six-month target for a brand new to this work is above 50.
A checklist score for the technical groundwork. A high readiness score doesn't guarantee citations, but a low one usually means the mentions you do get are fragile.
Results
Client names are left out on purpose. Figures are raw totals from each client's own Search Console, analytics, and tracking data, as of October 2026.
Monthly Google clicks, April vs September 2026 (5,864 to 8,806).
Monthly Google clicks, April vs September 2026 (251 to 3,987).
Monthly Google clicks, April vs September 2026 (5,084 to 6,928).
Page-one keywords in a single quarter, July to September 2026.
Estimated AI visibility score, June to October 2026.
Pages with correct canonical tags, out of 121, across four audit rounds.
Mobile load time (LCP) for a multi-brand retailer after new image guidelines.
Time to optimize a blog post, down from 6 to 8 hours by hand, using AI agents and a human review queue.
Views on 38 community posts and replies placed over six weeks, August to September 2026.
What audits usually turn up
Real findings from recent audits, with the names taken out. Most of them are invisible until someone goes looking.
One rebuilt site lost about 86% of its tracked keywords in seven months: no canonical tags, and both www and non-www versions live. Another left 62 of 77 ranking keywords pointing at pages that now return 404.
Over 3,000 branches showed up to search engines as "Loading branches..." because the list only existed in JavaScript. Another locator served 85 words of HTML against 33,100 monthly "near me" searches.
Price searches for one restaurant chain ran 22,200 a month, but the menu was an image. The only text menu AI tools could quote was an old orphaned page with outdated prices.
Homepages carrying 30 to 69 MB of images, including 4 MB logo files and 10 MB images preloaded in the page head. Mobile load times of 10 to 14 seconds.
Malformed schema on 239 pages, two conflicting Product records for every product, and one business record copied onto 100 pages. Search engines and AI tools ignore or distrust markup like this.
96 to 98% of search visits coming from people who already knew the brand. One site had 172 non-brand page-one keywords against competitors with 4,000 to 5,000.
A lender named in 0 of 67 loan answers that cite Reddit. A restaurant chain named in 0 of 309. An insurer with no official reply anywhere in 188 comments about it.
A tag manager whose only analytics tag was the old Universal Analytics, which stopped collecting data in July 2023. Two booking buttons sending people to two different calendars.
What you get
Honest timelines
Google typically recrawls an existing page about every 30 days, so we don't call a fix a failure until its window has passed. These are the windows we plan around.
Who we've worked with
Questions
SEO gets your pages ranking in Google's results. AEO (answer engine optimization, also called GEO) gets your brand named and linked when people ask AI tools like ChatGPT, Gemini, Perplexity, or Google's AI Overviews for a recommendation. They share a technical foundation, but AI tools lean much more on clear, quotable answers and on what other sites, forums, and the press say about you.
We ask AI tools the buying questions your customers actually type, never "what is [your brand]". Each answer is scored on whether you are the top pick, one of the options, mentioned briefly, or missing, and the score is the credit earned out of the total possible. Questions where you are missing stay in the score, so it can't be padded.
It depends on the fix. Title and snippet changes usually show in 1 to 2 days, canonical fixes in 1 to 3 weeks, and structured data in hours to 2 weeks. Google typically recrawls an existing page about every 30 days, site moves take 1 to 3 months, and recovering from a core update usually takes 3 to 6 months.
It doesn't hurt, but it isn't required, and Google's own guidance treats it as optional. We add one when it's cheap to do, and we weight it lightly. Crawler access, structured data, and clear answer-first content matter far more.
Current and recent work covers the Philippines, the United States, Australia, and New Zealand. Keyword and ranking data always comes from your own market, not a US default.
Let's talk
Book a 20-minute call. Tell us your site and your top competitors, and we'll tell you honestly what's holding you back and whether we can help.
me[at]wesleetrout[dot]com