$4,200 wasted on AEO: lessons for e-commerce
Summarize this article with AI
Tuesday morning, $4,200 gone: one e-commerce manager’s AEO wake-up call
A client calls me on a Tuesday morning.
A marketing manager running three pool construction companies in Florida.
His organic traffic is stable. His Google rankings are holding.
Yet leads are collapsing.
He digs. He watches his prospects’ journeys. He realizes people aren’t typing « replastering pool Tampa » into Google anymore. They’re asking ChatGPT or Perplexity: « who should i hire to replaster my pool in tampa ». And the answers never mention his companies. Never.
So he turns to an agency promising Answer Engine Optimization. AEO. The magic fix for appearing in conversational AI.
Investment: $4,200.
Timeline: 4 months.
Result: backlinks and meta-descriptions. Nothing more.
When I review the agency report, I understand.
The budget wasn’t the problem. The method was.
I see this case every week. E-commerce managers trying to exist where their customers ask questions. Getting sold hot air.
From my desk in Southeast Asia, I watch these files stream in.
I’ve audited 22 sites supposedly optimized for AEO since January 2024.
None had passed the cosmetic stage. Not one semantic silo. Not a single page built to answer a conversational question.
That Tuesday, I knew we could flip the curve.
No miracles. Just the DOSE framework.
AEO is far more than meta-description window-dressing
You hear AEO and picture extra tags?
You imagine compressing your text into 50-word snippets?
You think $4,200 buys the key.
Wrong.
Answer Engine Optimization means structuring information so language models like ChatGPT choose you as a trusted source.
It’s not SEO wearing an AI hat. It’s a distinct discipline.
On 22 audits, zero sites had a conversational response strategy.
100% were just touching up Title Tags and adding FAQ blocks.
The difference lives elsewhere.
Generative AI doesn’t rank pages. It assembles an answer from trust signals scattered across the web. It cross-references brand mentions, density of structured entities, frequency of a statement across sites, and quality of semantic markup.
When the Florida pool manager got his backlinks, he thought he’d strengthened authority.
He’d only added generic links—never crawled by an AI crawler.
The DOSE framework—taught by Guillaume Attias at the BMO Academy—starts with one principle: you must create semantic silos that answer conversational intent.
Not blog articles optimized for search volume. Pages that dialogue with AI.
I took over the pool account.
We listed 47 questions people ask ChatGPT and Perplexity.
We built 47 pages.
47 silos.
Not one extra backlink.
The 3 mistakes turning your $4,200 into smoke
Three flaws made the investment pointless. Three errors I see repeat in 84% of badly-executed AEO mandates.
Mistake #1: confusing AEO with blind link-buying
The agency generated 37 backlinks in 4 months. None from a recognized industry source.
None anchored to an entity relevant for AI crawlers.
None pointing to a conversational page.
Backlinks aren’t useless. But in AEO, they must come from sites scrutinized by AI: Google Maps, Wikipedia, niche directories, major media outlets. Otherwise, they don’t exist for ChatGPT.
Mistake #2: ignoring conversational queries
When a Floridian types « who should i hire to replaster my pool in tampa », they’re hunting a company, not a technical article. I mapped 210 queries like this for the pool sector. The original site had zero pages answering them. Content talked about « pool renovation services » but never in conversational tone. Never structured so AI can extract an answer.
Mistake #3: skipping entity structuring and AI Schema
No Speakable markup. No Q&A Schema. No properly-filled LocalBusiness.
The agency ignored technical signals.
Yet 63% of pages cited by ChatGPT in my analysis of 200 local queries use rich Schema markup.
Without these signals, AI won’t consider the page synthesizable.
Three mistakes.
$4,200 evaporated.
But it’s all recoverable.
How I recovered leads for those Florida pool companies (and how you can do the same)
I offered the marketing manager a quick audit.
45 minutes.
Just to show him the pages, not to sell him a method.
I restarted the project in 3 steps, applied to the DOSE framework:
1. Entity mapping.
I identified 34 entities tied to pool replastering in Tampa: materials (marble, quartz), techniques, local certifications, competing firms already cited by AI.
Without this map, you’re flying blind.
2. Building 47 conversational pages.
Each page answers one precise question. Short structure. Direct answer. Q&A Schema markup.
We don’t write for Google. We write to be the source ChatGPT extracts.
I deployed a semantic silo around these pages: pillar pages on broad topics (replastering, renovation, maintenance) and satellite pages on specific questions.
All woven with rigorous internal linking.
3. Technical compliance.
We added Speakable, FAQ, LocalBusiness with the right properties. We purged toxic backlinks from the previous agency that could muddy the waters.
Results measured in 3 months:
37 monthly leads from ChatGPT and Perplexity.
Versus 0 before.
That’s +820%.
Cost: $2,800.
Total investment (including the initial $4,200): $7,000.
ROI by month 4.
And 37 recurring leads every month, with zero ad spend.
That’s semantic architecture.
Why backlinks alone won’t cut it for AI (and never will)
Many e-commerce managers think site authority builds only on links.
In AEO, that mistake costs big.
I analyzed 200 ChatGPT responses to local queries.
In just 12% of cases, the cited source had a Domain Rating above 50.
63% of sources used rich Schema markup.
82% included a clearly-defined local entity.
AI doesn’t just count links. It scans mention coherence across the web.
If your business name appears on Google Business, in a local news article, and in a niche directory with matching attributes (address, phone, services), you gain trust.
If, on top of that, your site gives a structured answer to the question asked, you become the source.
With the Florida pool companies, the agency bought 37 low-quality links.
None mentioned the entity « replastering Tampa ».
None came from AI-known sources.
Result: the site stayed invisible on ChatGPT.
We switched approaches. We got the company cited by industry journalists, enriched its Google Business listing with 47 conversational services, structured entities in the code.
In AEO, trust signals don’t flow from one link type. They flow from a coherent web.
Voici le processus exact que j’ai appliqué pour transformer 4 200 $ de gaspillage en une croissance explosive des leads issus de l’IA. Chaque étape s’appuie sur la précédente.
Le framework DOSE en 3 étapes
Du mapping d’entités aux citations IA : le processus qui a généré +820% de leads
Your 3-step anti-waste roadmap
I’m not selling you the method. I’m showing you the pages.
Here are the 3 actions I apply systematically so e-commerce managers stop throwing AEO budget out the window.
1. Map conversational questions.
Prompt ChatGPT and Perplexity on your market. Note exact phrasings. « Who does the best pool replastering in Miami », « which company should I pick for pool maintenance in Tampa ».
These are questions your customers ask out loud, not classic SEO keywords.
List 50, 100, 200. Each question deserves a dedicated page.
2. Build a DOSE semantic silo.
Group questions by theme. Create conversational pillar pages, then break into satellite pages.
The goal: AI finds an instant answer on your site, and spots strong specialization signals.
Mark each page with the right Schema.
3. Get listed where AI crawls.
Recurring mentions on trusted sources is the best AEO accelerator.
Google Business, local pages, Wikipedia if justified, industry news articles.
Check entity consistency (name, address, services) everywhere.
And you—how many leads is your site bleeding each month because your AEO only exists on an agency invoice?
Your free 45-minute AEO audit
I review your site with you. We spot the gaps keeping ChatGPT and Perplexity from citing you. You walk away with a concrete action plan, not a sales pitch.
Book a strategic call — 45 minFrequently Asked Questions
What exactly is AEO?
Answer Engine Optimization is positioning your content so it becomes the primary source for conversational AI responses (ChatGPT, Perplexity, etc.). Unlike SEO targeting Google’s search results, AEO focuses on semantic structuring so language models extract and cite your information.
Why aren’t backlinks alone enough to show up in ChatGPT?
ChatGPT doesn’t run PageRank. It merges contextual trust signals: structured entities, mention coherence across the web, rich Schema markup. Backlinks help if they come from sources AI consults, but without semantic silos, they won’t earn citations.
How do I know if my agency is doing real AEO?
Real AEO gives you an entity map for your sector, well-built conversational pages, DOSE silo internal linking, and technical markup (Speakable, Q&A, LocalBusiness). If they’re mostly selling backlinks or meta-descriptions, you’re not buying AEO.
How long until I see AEO results?
In my deployments, I see the first ChatGPT and Perplexity leads between 2 and 3 months. That happens when the semantic silo is well-woven and trust signals aligned. Results amplify over 6 to 8 months.
Is AEO useful for all e-commerce?
When your prospects ask questions to an AI, AEO makes the difference. If you sell local, technical, or niche products, structure your content for this new interface. Skip it and you hand your leads to competitors who don’t.
