The surprise nobody priced in
Every marketing channel I have ever watched rewarded the people who arrived early. Google rewarded the SEO operators of 2008. Facebook rewarded the direct-response shops of 2014. TikTok rewarded the creators of 2020. Each time the pattern was the same: the channel was confusing, the experts loud, the budgets small, and the winners obscure.
What was never true — until now — is that a seed-stage company could outrank a public company on the strength of a Reddit thread and a clean schema block. SEO punished you for being new. Paid social punished you for being small. Every channel had a tax on early-stage life: domain authority, account history, pixel maturity, creative budget, agency relationships, or all five.
The answer engines do not charge that tax. They cite the clearest answer they can find. If yours happens to live on a domain that was registered last quarter — fine. The model does not care.
That is the arbitrage. And like every arbitrage, it has a clock attached.
Balfour's pattern: every channel has four acts
Brian Balfour has been writing about platform shifts since before most founders had heard the phrase. On Lenny's podcast in August 2025, he described the arc that every dominant channel follows — and he was unusually direct about which act we are in.
His four acts: early arbitrage, gold rush, saturation, professionalization. Early arbitrage is the phase where the channel is mispriced because most operators do not believe in it yet. Gold rush is when the believers pile in. Saturation is when CPAs creep upward and the easy wins evaporate. Professionalization is when the channel becomes a budget line item with a vendor, a benchmark, and a procurement process.
Balfour's claim — and I think he is right — is that AI search is in act one. Most marketing teams treat it as a curiosity. Most CMOs cannot tell you their citation rate in ChatGPT. Most agencies are still selling 2019-era backlink campaigns.
That asymmetry is the whole opportunity. The teams that move now will own category positions by the time the gold rush hits in 2026 and 2027.
Smith's mechanics: why this channel rewards the small
Ethan Smith of Graphite has the cleanest framework I have seen for how AI answer engines actually rank content. Four claims. Each one inverts a 15-year-old SEO assumption.
Mentions beat rank. Google ranked pages. The answer engines synthesize mentions. When ChatGPT decides which three companies to recommend for "AI marketing tools for early-stage SaaS," it is not pulling the top three blue links. It is weighting how often each name appears in context, how the surrounding sentences frame it, and whether the citations agree.
AI traffic converts 6× Google. A user who lands on your page from ChatGPT has already been pre-qualified by the model. They asked a specific question. They got a specific answer. They clicked because they want exactly what you sell.
Reddit is kingmaker. The model training data skews heavily toward Reddit. A detailed, genuine, high-upvote thread recommendation carries more weight than most branded content you've ever published.
There is no domain authority gate. A company mentioned once on a high-trust source can outrank a company that owns the keyword on its own domain.
The structural advantage of being small
The incumbents have a problem. They have twenty years of positioning built around what Google rewards. Changing that is a procurement process, a brand committee, and six months of approvals.
You have a morning.
The fastest thing you can do today: open your three highest-traffic pages and add a definition block (what exactly is this, in one sentence), a FAQ section (the five questions your customers actually ask), and a statistics section (three numbers about your category with sources cited). Those three elements are the signals that answer engines extract when deciding who to mention.
Run a free AEO audit → to see exactly where you stand across ChatGPT, Claude, Gemini, and Perplexity. The report takes 90 seconds.
Frequently asked
How long does the AEO arbitrage window last?
Brian Balfour's framework suggests 12-18 months before incumbents professionalize the channel. Most CMOs still can't tell you their citation rate in ChatGPT — that's the window.
Does domain authority matter for AEO?
No. AI answer engines cite the clearest answer they can find regardless of domain age. A 30-day-old domain can be cited next to a 30-year-old one.
What's the fastest way to improve AEO standing?
Structure your highest-traffic pages for extraction: definition blocks, FAQ schema, statistics with sources cited. These are the signals that get you mentioned.