2010s
Search dominated ecommerce discovery
Brands competed on keywords, rankings, and organic traffic.
About
About Geoffy
The team building coherence infrastructure for AI commerce — so the right products become the reference answer in AI shopping conversations.
Why Geoffy exists
Search is changing.
For the last twenty years, ecommerce discovery has been dominated by search engines. Brands optimised for keywords, rankings, and clicks.
That model is ending. Customers increasingly ask AI assistants instead of typing queries into Google:
- "What's the best lightweight hiking backpack?"
- "Which protein powder is good for runners?"
- "What's the best standing desk under £500?"
These systems don't return pages of links. They return answers — and those answers recommend a small number of products. Often just one or two.
That one-or-two slot is continuously contested. Models update, competitors respond, answers drift. Winning it once isn't enough; staying in it requires infrastructure.
Geoffy is that infrastructure. We call the practice Reference Engineering — structuring product information so AI systems understand it, trust it, and recommend it.
Our background
Built by ecommerce operators and product builders
Geoffy was founded by Anthony, John, and Branko.
Between us, we've spent more than 60 years working in ecommerce, search, software, and digital growth.
- Built ecommerce platforms
- Launched and scaled online stores
- Run agencies and technology companies
- Worked with startups through global retailers
- Worked across SEO, product discovery, and ecommerce infrastructure
Like many people in ecommerce, we started noticing something: customers weren't just searching anymore. They were asking AI. And the way AI systems choose which products to recommend is fundamentally different from how search engines rank pages.
That gap — between how ecommerce data is structured today and how AI systems need to read it — is where Geoffy lives.
The Geoffy journey
How discovery shifted
Early 2020s
Structured product data matured
Search engines and marketplaces got better at reading product attributes.
2023 onwards
AI assistants entered the buying journey
Customers started asking conversational questions instead of typing keywords.
Today
AI recommendations shape what gets bought
Large language models surface specific products directly inside answers.
Geoffy was built to help ecommerce brands hold that recommendation slot, not just show up in it.
Building Geoffy
A focused team building coherence infrastructure for AI commerce
Geoffy isn't just the three founders. Behind the scenes we work with a small group of engineers, ecommerce specialists, and data experts designing and building the platform.
Together we focus on one thing: making ecommerce catalogues coherent enough for AI systems to confidently recommend them.
That means turning messy, inconsistent product data into structured meaning across every layer AI models read — product pages, reviews, supplier specs, and structured data — so the signal the AI sees is clean, consistent, and trustworthy.
What makes Geoffy different
Built for AI recommendations, not legacy SEO workflows
Most ecommerce tools still treat discovery like it's 2015.
They focus on keywords, rankings, and page optimisation.
Geoffy approaches the problem differently.
Built for AI discovery
Geoffy is designed specifically for AI assistants and recommendation engines, not traditional search engines.
Structured product understanding
Instead of just generating content, Geoffy builds structured product meaning using attributes, intent modelling, and Reference Pages engineered for how AI reads.
Transparent optimisation
- no hidden content
- no cloaking
- no hallucinated product attributes
- structured outputs that match visible pages
First-party discovery infrastructure
Everything Geoffy generates lives on the merchant's own domain. No marketplace dependency. No hidden optimisation layers. Just structured discovery pages AI systems can read and reference.
Geoffy principles
How we think about AI discovery
Geoffy is built around a few simple principles. These guide how we design the platform and how we believe AI discovery should work.
Machines should understand products
Product catalogues should be structured so AI systems can interpret attributes, context, and meaning — not just keywords.
Transparency matters
AI-optimised content should match what customers can actually see. No hidden content. No bot-only pages.
Discovery should belong to merchants
Brands should control their own discovery infrastructure rather than relying entirely on third-party platforms.
Optimisation should be explainable
If a system recommends a product, there should be clear structured signals explaining why.
The Intent Tiers
How buyers actually ask AI
Not every AI shopping question has the same depth. Geoffy structures your catalogue to win across all three:
Broad
Category only.
"I'm looking for a coat."
Mid
Category plus a few qualifiers.
"A waterproof and warm coat."
Ultra
Category, brand, specs, use case, thresholds.
"A Gore-Tex, fur-lined North Face jacket for the Himalayas, 20K waterproof rating."
Most discovery tools focus only on broad queries. Geoffy covers all three — so you're the reference answer whether the buyer is browsing or buying.
Our mission
Help great products get discovered in an AI-driven world.
We believe the next era of ecommerce discovery will be shaped by AI assistants, recommendation engines, and conversational search.
And when that happens, the brands that win will be the ones whose products are legible to machines.
Geoffy helps make that possible.
Get started
Start getting your products found by AI.
Turn your catalogue into AI-ready discovery pages. Set up in minutes.
Already have an account? Sign in