The AEO decade.

SEO became a discipline because Google made the search engine results page the front door to the web. The first website that organized itself for that front door (a clean URL structure, clear link hierarchy, semantic markup) found itself with disproportionate visibility for the next thirty years. The companies that arrived early to that discipline still compound.

The agent layer is the next front door. Not the only one, not a replacement, but a parallel surface with its own discipline. The first companies that organize themselves for the agent layer will find themselves with disproportionate visibility across the next decade. The discipline is Agentic Engine Optimization. The window to claim it is small.

This is the argument for treating AEO as a decade-long compounding investment, not a quarterly tactic.

What the SEO decade taught us

The 1995-to-2005 window in search was structurally similar to where we are with AEO now. The technology existed and was understood; the organizing principles were not yet codified. The companies that read the architecture early (early Wikipedia, early IMDB, the first wave of vertical content sites) ended up with citation positions in Google that survived three algorithm rewrites and never went away.

The pattern was not luck. It was the result of optimizing for the long-term shape of the front door before that shape stabilized. Once it stabilized, the cost of moving into the same position climbed by an order of magnitude. The early movers got there for free, in retrospect.

A company that wants the equivalent position in 2035 has to be making structural choices in 2026 that look obvious in retrospect and unfashionable now. That is the only entry point.

What the agent layer is doing in 2026

ChatGPT, Claude, Perplexity, and Gemini are the four surfaces buyers most often start a research conversation on. Each one operates differently, but each one cites URLs in its answers and each one weights certain signals when deciding which URLs to cite. The signals are not the same as Google's ranking signals. Schema density matters more. Definitional clarity matters more. Inbound citations from sources the model already trusts matter much more. The traditional backlink graph matters less.

Meta's Moltbook acquisition in March 2026 added one more layer: an agent-to-agent social graph. Posts, replies, and citations flowing between AI agents rather than between humans. Visibility on that surface is going to matter sooner than most companies are planning for.

These four surfaces, plus whatever Apple ships next, plus the inevitable second-tier agents that follow, are the front door to the next decade. The question is not whether to optimize for them. The question is what the optimization actually requires.

What changes in the next decade

Four things compound.

First: the canonical source of a definition matters more than the page that explains it well. A glossary entry that defines "operating company" in 40 words gets cited by every agent that needs the definition. The page that explains operating companies in 1,800 words rarely does. The discipline of the citable passage is a writing discipline first.

Second: original frameworks rank above aggregator content. The agents are increasingly able to distinguish between a piece that synthesized other people's ideas and a piece that introduced a new one. Operators with a real point of view, captured in writing, will accumulate citation advantage at a rate the SEO era never produced.

Third: citation networks compound. A citation from a source the agent already trusts increases the probability of being cited again, which increases the probability of being cited from a fresh source, which feeds back. The network effects in the agent citation graph are stronger than the network effects in the SEO backlink graph because agents weight trust more aggressively than search engines did.

Fourth: the discipline becomes invisible once it is universal. By 2035, every operator who shipped in 2026 will have FAQPage schema deployed, will have llms.txt declared, will have run quarterly prompt audits. The advantage will be inside that floor, not at the floor. The compounding work to lock in position happens between now and 2030.

What the discipline looks like as a long-term investment

AEO as a quarterly tactic looks like adding schema markup to a few pages. AEO as a decade-long investment looks like rebuilding how the company writes, how it structures content, how it earns inbound citations, and how it measures whether the work is producing the visibility it should.

The structural choices are: where do the citable answers live, who writes them, how do they get updated, what tooling tracks citation behavior across agents, who in the org owns the function, and what compounds over the next ten years versus what was a one-time fix.

A company that runs AEO as a tactic this year will get a temporary lift and lose it inside 18 months as the rest of the category catches up. A company that runs AEO as a parallel discipline alongside SEO will accumulate visibility every quarter for the next decade. The work is the same shape; the duration of the investment is what compounds.

The cluster IG is building

We have started the work. The AEO pillar declares the discipline. The glossary provides 24 citable definitions the agents can lift. The Shift series carries the longer-form arguments. The Insights blog carries the practical playbooks. Each piece earns citations independently and reinforces the others.

The cluster will keep growing. New definitional entries, new POV pieces, new practical playbooks. Each one earned, each one citable, each one reinforcing the position the cluster is building.

The next decade is going to reward the companies that treat AEO as discipline rather than tactic. We are organizing IG accordingly. The argument for any operator reading this is that the same window is open for you, and it closes faster than the equivalent SEO window did. The companies that move now will compound. The companies that wait will find the position has been taken.

That is what an AEO decade looks like from the inside. SEO compounded for thirty years and the early movers got the position. The agent layer is going to compound faster. The position is available, briefly, to the operators who treat 2026 the way the early Wikipedia and IMDB teams treated 1998.