Measured in Ahrefs against a fixed October 2025 baseline. Where the starting figure was a genuine zero it is stated as such rather than converted into a percentage.
Problem
HeyOz builds an AI platform for creating image and video ads. The product worked and people liked it. Everything around discovery, however, was missing rather than broken, which is a different and in some ways easier problem.
There were no landing pages beyond the homepage, so there was nothing for a buyer-intent query to land on. There was no blog infrastructure at all. There was no structured data anywhere on the site, which meant Google and the LLM crawlers had to infer what the product did from marketing copy rather than being told. Site architecture, internal linking and indexation had never been designed, because the site had never needed them.
The result was a company with real product-market fit that was, to a search engine or an answer engine, close to invisible. In October 2025 it ranked for nothing, and across all seven major AI answer engines it was cited exactly zero times. For a category where marketers routinely ask an AI assistant which ad tool to use, that absence was the entire commercial problem.
Baseline
The October 2025 snapshot is unusually easy to read, because almost every number in it is zero. That is worth stating plainly rather than dressing up. It also means every figure later in this document is measured from a genuine standing start.
Engines tracked: ChatGPT, Perplexity, Gemini, Google AI Overviews, Google AI Mode, Copilot and Grok. None of the seven cited the domain. Because the baseline is zero on three of these four measures, growth is reported as absolute figures throughout rather than as percentages, which would be meaningless.
Strategy
The instinct with a new domain is to start publishing immediately, because content feels like progress and foundations do not. We inverted the order deliberately. Three principles.
No content sprint in month one. A startup with no architecture, no internal linking and no schema that publishes forty posts ends up with forty pages nothing can properly crawl, and a founder convinced that SEO does not work.
Execution
SearchAxe built the whole discovery layer from nothing. Because the site had no history, almost every item was a first-time build rather than a fix, which made sequencing straightforward and mistakes cheap to avoid.
-
Designed and built site architecture and internal linking from scratch, with an indexation strategy defining what should and should not be crawled.
-
Implemented structured data across page types so Google and LLM crawlers could parse the product without inferring it.
-
Built the landing page set against buyer-intent queries in the ad creation category.
-
Stood up blog infrastructure and a content plan shaped for AI Overviews, AI Mode and LLM extraction.
-
Set and applied the on-page format standard: question-led headings, direct answers, scannable specifications.
-
Ran a backlink programme targeting topical and relevant placements rather than volume.
-
Ran Reddit marketing in the communities where marketers and creators evaluate ad tools.
Results
Ten months in, HeyOz ranks on page one for thirty queries, is cited in six of the seven major answer engines and has a link profile ten times the size it started with. The absolute numbers are still those of an early-stage company, which is what it is.
Source: Ahrefs. July 2026 used as the last complete month; AI citation counts captured August 2026. Domain Rating peaked at 61 in April 2026 and has since settled in the low fifties, which is normal volatility for a domain acquiring links at this rate.
Lessons for early-stage products
Order of operations is the whole strategy
Ten months of content published onto an unparseable site would have produced almost nothing. For a new domain, the sequence in which you build is not a detail of the plan. It is the plan.
Schema is how you talk to LLM crawlers
Structured data stopped being an SEO nicety the moment answer engines needed to state what a product does and who it is for. Say it explicitly, or leave a model to guess from your headline copy.
Zero is a genuinely useful starting line
No legacy index, no cannibalisation, no historic bad links. A new domain can be built correctly the first time, and that advantage is gone within about a year of unmanaged growth.
ChatGPT is not the only engine that matters
HeyOz earns more citations in Grok than in ChatGPT, and more in Gemini than in Google AI Overviews. Optimising for one assistant and assuming the rest follow leaves a large share of the answers on the table.
Ready to turn search into a growth channel?
Let's build a stronger organic growth engine.