For years, marketing technology sold the same promise: presence is expensive, and you can win without it. SEO puts a firm in one city ahead of the shop next door to the customer. Programmatic bought attention in markets nobody had visited. AEO, the newest version, makes a page legible to an answer engine that will resolve a stranger’s question without ever introducing you to them. Every one of them traded proximity for reach, and the trade paid – for a little while.
AI hit the mainstream in 2022, and has changed the way we strategize and do business. In their mid-2026 report, Graphite found the number of 50% of articles on the internet are AI-generated, which is now equal to the 50% written by humans. That trajectory has reversed, and the reversal is measurable.
When the machine can produce competent copy about any topic instantly and answer the question without sending anyone anywhere, the ability to compete at a distance stops being an advantage. Everyone has it. It costs nothing.
Competence at a distance was the product these tools delivered, and that same competence is now free and infinite. What that leaves behind is commodity content: pages a model could have produced without access to anything the publisher knows.
What remains scarce is the thing the tools were built to make unnecessary. Relational proximity, the absence of layers between a business and the problem it solves, is now the only input that cannot be synthesized.
Search Engine Optimization – SEO
SEO is the foundation everything else is built on; earning visibility in traditional ranked results. AEO, GEO, and VEO are surfaces that sit on top of it — one connected stack, not four separate jobs.
Answer Engine Optimization — AEO
AEO optimizes direct answers, such as featured snippets, People Also Ask, voice search, and engine-generated answer boxes. This system is built on concise, structured, question-based content and schema.
A Virtual Connection Can Still Be a Real One
A virtual connection can still be a real one, provided what sits underneath it is authentic. The point was never to reject technology. The point is that technology should serve real relationships rather than substitute for them.
AI and VR work over the long run when they are built on something grounded and true. Strip the grounding away and what remains is an increasingly convincing simulation of connection, which is a worse product than the real thing rather than a better one.
The grounding is usually a place. A shop, a route, a territory, a room where the work happens. That is what a search engine cannot fabricate on your behalf, and what a competitor two states away cannot borrow.
Proximity Marketing: Physical Closeness as a Targeting Parameter
Marketers already have a term that sounds like this and means nearly the reverse. Proximity marketing is beacons, geofencing, and a coupon pushed to your phone as you walk past a storefront.
Own Your Backyard – OYB
Building dominance in the community, network, and market directly in front of you — hiring locally, showing up physically, reinvesting where you operate, compounding relational trust — before extending into distant or virtual growth channels.
The inversion is the whole point. Closeness is not the trigger for a message; it is the source of the knowledge that makes the message worth reading. One approach treats the neighborhood as an audience segment to be intercepted. The other treats it as a research advantage that cannot be bought, licensed, or generated.
Relational Proximity: The Distance That Actually Matters
This is what saves the argument from collapsing into a claim that only local businesses can win.
Relational Proximity
The distance between what a brand claims and what a person has actually experienced of you. Distinct from location-tech “proximity marketing” (beacons/geofencing).
A firm in Newport Beach serving accounting practices in twelve states has no physical proximity to most of its clients and enormous relational proximity to their problems. Relational proximity sits in the quarterly calls, sees the same reconciliation failure across forty firms, and knows which software migration goes badly in which month. That is a backyard, not a neighborhood.
The reason geography keeps appearing in these examples is that physical closeness is the cheapest and most reliable way to accumulate relational proximity. Show up to enough job sites in one county and the knowledge accrues whether you plan for it or not.
Remote operators can reach the same position, but they have to construct deliberately what a local operator gets as a byproduct: recurring contact, unmediated exposure to failure, a client roster narrow enough to produce patterns rather than anecdotes.
Distance in the relational sense is what commodity content actually measures. A page scores high when the writer was far from the problem, working from summaries of summaries. It scores low when the writer was close enough to see something the summaries missed. Miles are a decent proxy for that. They are not variables.
What the Algorithms Are Quietly Reaching For
The question keyword tools cannot answer is the one that now decides visibility: how much of a page could a model have generated on its own, versus how much represents something only this business knows?
SewerTV came to Flying V Group through a direct referral from Google’s Local Service Ads team. Dave had already built a strong local presence and wanted to extend it with content. The obvious move was a blog program. We went a different direction, because the thing Dave knew could not be written down without losing most of it.
So we filmed him instead. Dave set up his hydro jetting machine against a length of clear pipe and walked through what the machine does and why it does it, with the audience able to see the blockage give way and hear what that sounds like.
That video now has more than 100,000 views on YouTube and continues to send leads and authority back to the SewerTV site.
Video Engine Optimization – VEO
Optimizing video (YouTube and on-site) to be discovered, ranked, and cited across video search and the AI answers that increasingly pull from video: transcripts, chapters, video schema, thumbnails, channel authority.
Nothing about that was expensive or clever. It was a recognition that Dave had access to something no competitor could write from a desk, and that the format had to match the access. The blog version of that video would have been a thousand words on the benefits of hydro jetting, which is a passage a model produces from the prompt alone.
That distinction holds at the sentence level too. Here is the version that gets published:
Hydro jetting clears stubborn blockages and is more effective than snaking for long-term results. Most plumbers recommend it for recurring drain issues.
Put a competitor’s name on that and nothing breaks, which is the tell. No answer engine has reason to retrieve it over the ten thousand near-identical passages already indexed, because it holds nothing the other ten thousand lack. Here is the version written from the work:
On clay laterals in older neighborhoods, root intrusion comes back on a cycle rather than getting solved once. We jet them and put a camera through afterward, and the roots are usually back inside two years unless the joint itself gets addressed.
Same topic. The second names a pipe material, identifies a failure pattern, contradicts the implied promise of permanence in the first, and reports something the writer saw through a camera. Put a competitor’s name on it and it becomes a claim they cannot support.
What separates them is not effort or word count. One writer had a camera in a lateral and the other had the general idea of drain cleaning. Everything downstream, whether the page ranks and whether an answer engine quotes it, follows from that gap.
The 50/14 Split: What’s Published and What Gets Cited
Publishing has reached parity. Citation has not moved with it.
Graphite’s analysis found that 86% of articles ranking in Google Search were written by humans, along with 82% of articles cited by ChatGPT and Perplexity, and AI-generated pages that do surface in Google tend to rank lower. Machine-written work is half of what gets published and 14% of what gets ranked.
The study measured authorship rather than causation, and the distinction matters.
Human-written pages skew older, better linked, and hosted on more established domains, so the gap has explanations that have nothing to do with humanity. But the weaker reading is still damaging to the volume strategy. Whatever the mechanism, flooding the index has not produced visibility.
The irony underneath it is that the machines need what they cannot make. A retrieval system’s entire value depends on finding something real: a lived experience, a novel insight, a number from someone who was in the trenches.
Synthetic content trained on synthetic content collapses into noise, and there is no way out of that loop except by locating a human who was there. Even as AI mediates more of how we find each other, it keeps routing back to somebody who was close enough to the problem to know something about it.
Fishkin and Schaefer Disagree, and the Disagreement Is the Useful Part
Two of the sharpest voices in marketing have looked at this and reached incompatible conclusions.
Rand Fishkin, who founded Moz and now runs SparkToro, published an article in May 2026, Inimitable Product is the New “Make Great Content,” arguing that as AI compresses and summarizes the content published online, the only durable advantage left is what he calls “inimitable product:” the things that cannot survive being summarized because their value lives in the doing, not the describing.
His blunt assessment of where the platforms are headed: “We’re going to ruin the Internet.” The artifact gets absorbed; the trust, and the human relationship behind it, does not.
Mark Schaefer went the other way. The subtitle of his book Marketing Rebellion is “The Most Human Company Wins,” and his newer book, Audacious: How Humans Win in an AI Marketing World, builds the case from interviews with working creatives.
His response to the flood is not to abandon publishing but to make authorship unmistakable, to the point of badging his own posts as human-written. Where Fishkin sees a channel not worth defending, Schaefer sees a channel where the human signal became the differentiator due to its scarcity.
Fishkin describes businesses whose product is the content: publishers, affiliates, course sellers, newsletters. Those businesses have no relational proximity to anything, instead, their product is always placeless, reproducible, and delivered through a channel they do not own, which is why the channel could take it. Schaefer describes businesses whose product exists independently of the content, where publishing was never the revenue engine but the evidence that the engine works.
The reconciliation is that they are arguing about the channel while agreeing about the asset. Both locate durable advantage in something a system optimizing for scale can’t replicate.
Fishkin puts the responsibility on the product; whereas Schaefer puts it on the person. For an operator close to the work the two collapse into each other, because the person doing the work is the product. Content there is not the thing under threat. It is proof of an inimitable product you already own.
Where I fall: Fishkin is right that the traffic loop is dead and wrong to generalize the prescription past media businesses. Schaefer is right that the human signal is a scarce input and light on how it survives compression. What both leave out is the simplest form of inimitability there is: proximity to a problem, sustained long enough to produce a pattern.
“We aim to solve the problem by creating patterns and interlinking content that addresses the stories that are happening in our office. One way we do this is using our Fathom note recordings to repurpose into LinkedIn posts or webinar material. It is a simple and effective way to produce non-commoditized content at scale,” – Robb Fahrion, CEO of Flying V Group
Owning Your Backyard
Stop asking what a topic requires and start asking what you know that nobody at a distance could write. The corroded coil at year seven. The permit desk in your city that moves faster on Thursdays. The reconciliation failure you have now seen at forty firms. The pricing conversation that runs differently in your market than in the case study you read.
Proximity is doing real work here, not decoration. An operator with a defined territory or a narrow client roster holds a sample nobody at scale can access: every job in that population, over years, under one set of conditions. That sample produces claims simultaneously specific enough to be cited and impossible to generate remotely. A national content team writing about drain maintenance reasons from general principles. Dave has a camera, a clear pipe, and twenty years of laterals in the same neighborhoods.
The same material serves both channels, which is the part most teams miss. It earns a citation in an AI Overview and a referral at a chamber breakfast, because in both cases what transmits is that somebody knows something firsthand. Revenue-model framing clarifies the stakes. If your LTV:CAC ratio depends on a hundred low-intent visitors, the zero-click web is a problem. If it depends on twelve qualified conversations a month, first-hand proof is the cheapest acquisition input available.
Which leaves a question with no amount of optimization answers on your behalf. Competing on volume against systems that generate volume for free is not a production problem to be solved with more production.
It is a question of whether the business has been close enough to something, long enough, to know what the machine cannot say for it. That knowledge was always in the backyard. What changed is that it is now the only thing worth publishing.
