SEO Lived in Marketing. GEO Lives in Community.

Overview

Community managers spent a decade proving their value with page views and visitors. Now more of the web is crawled by bots than browsed by people, and the member who once found their answer in your forum is getting it from an assistant instead.

This is the pilot episode of CommunityAssistant.ai, so Chris Heuer and Marius Ciortea spend part of it laying out why the show exists at all before getting to the argument that ends up carrying the episode.

Why this show exists

Chris and Marius have been circling the same question for years, going back to Marius’s time as Chief Community Officer at IBM. It sharpened once Marius started building and deploying agents against real communities. What they kept finding is that practitioners are broadly aligned on the shape of the answer: the busy work and the business case scaffolding can move to AI, and the relationships, the trust building, and the irreducible human work stay with people. Marius frames the show’s premise as a question rather than a warning. Not what AI takes, but what an AI-enabled community lets community managers do that they could never do before.

They also name the older problem underneath all of this. Community has been underfunded in most organizations for a long time, and the business case has been hard to make. Part of what Chris wants this show to do is help practitioners connect the community to product feedback, customer insight, and organizational value in terms the people holding the budget already understand.

GEO, and Marius’s series

The conversation’s center of gravity is generative engine optimization, and the anchor for it is Marius’s ongoing GEO series on LinkedIn, which is where he has been working this argument out in public. It goes deeper on several threads the conversation only touches.

His claim, in short: SEO lived in marketing teams, and GEO will live in or next to the community team. Not as a consolation prize, but because the community is usually the only user-generated content a brand actually owns and can influence. Social is user-generated too, but from sources nobody can vouch for. Marketing copy is the thing models discount by default.

What makes it practical is that the signals are already in your hands. Whether the person who answered is recognized by the community as an authority. Whether they are employees. Whether they are specifically from support. Whether other members endorsed the answer. Whether it was marked correct. Whether it is current, or five years stale against a product that has changed twice.

Chris raises the emerging technical layer, llms.txt files and proposed ontologies for describing what a badge or an employee flag means to a crawler. Marius’s prediction is that most of that will not decide anything, because the models are already fast enough to read your pages and are not confused by your HTML. Platforms including his own will ship the backend files regardless, but the real leverage stays in the community work.

What replaces traffic

Chris brings a live example. While setting up this very recording, he and Marius hit a problem. Rather than searching Riverside’s community, Chris asked Claude and had a working answer in about thirty seconds. Riverside never saw the question, never saw the resolution, and never counted him as a visitor.

That is the zero-click problem stated concretely, and it is why Chris asks the harder version: how does a community manager justify their budget when traffic was the primary metric and traffic is going away? Marius’s answer is that if traffic is the only justification, the trouble started earlier. The more useful part is what replaces it. Citation data and bot-versus-human traffic are already becoming visible in platform back ends. Higher Logic is building GEO dashboards against exactly that data. The lawsuits over AI training turned out to have a side effect, which is that crawler behavior is now tracked in ways it was not before.

They also look ahead to attribution and advertising. Marius expects citation-linked advertising to arrive in AI answers eventually, on the grounds that no channel with that many eyeballs stays unmonetized forever.

Chris pushes on where GEO actually sits in an organization. His read is that larger companies will keep a technical specialist for the science of it, ICP definition and the parts of SEO that still apply, while the community manager gains credibility by raising their hand early and owning the part only they can own.

Only time, social capital, and the courage of each community manager will resolve this question.

The Work Map, and its gap

Chris walks through the Work Map, the open inventory of community management jobs mapped across the stages of the community journey and across community types, with each job marked as Human, Agent, or Hybrid. Two admissions come with it. There is still some AI slop in the map because the volume of work exceeded what he could do by hand. And GEO, the discipline this episode spends most of its time on, is not in the map yet. Both will get fixed in public, with contributions credited.

Why Marius is doing this

Chris asks the question directly, including how a vendor-employed practitioner gets a platform-agnostic show approved. Marius’s answer is that he owes his career to this industry and sees an obligation to move it forward rather than to only promote a specific product. He wants the long investment so many organizations have made in community management to pay off, and he thinks the age of AI is a genuinely good moment for a human-driven story. Otherwise, as he puts it, it is just agents talking to agents.

The waves before this one

Marius describes himself as outcome-driven. He does not care how the work gets done as long as it gets done in a way that frees him for more. That instinct predates AI by decades. When he started, there were no moderation features and every thread had to be read by hand. He asked for profanity filtering, then for a view of unanswered questions, and eventually had a moderation system. None of it replaced him. It moved work he did not want off his plate.

Translation is his best example. He once hand-translated a thread through Google Translate for a German member whose English was struggling, and posted it just for that person. The member felt seen. It was also, in Marius’s words, completely unscalable.

That lands on a phrase Chris has been carrying since the social media era: humans don’t scale. He explains where it came from, back when social media managers were expected to reply to everything, and how he has reframed it since. The question is no longer how to reply to everything. It is which things require a reply from a person and which can be handled by the technology instead. Which is, as Chris points out, the show’s name.

AI slop, and the awkward part

LinkedIn recently added a way to report AI slop, and both hosts expect similar features to land in community platforms. Marius flags the hard case, and it is not the obvious one. Much of the slop is well intended, written by people who are not fully in control of the language of the forum and who use AI to be understood. A report button in that context risks becoming a way to say that you do not like someone.

Marius closes on the uncomfortable version of the argument. When Chris got his answer from Claude instead of from Riverside’s community, the brand’s interest was still served. The user had a question and now has a working product. Insisting that every good answer must come from the community is the wrong hill.

What’s ahead

Guests are coming, including practitioners Chris and Marius have worked alongside for years. Richard Millington and Mark Birch both come up, and Marius wants ROI measurement to be a recurring thread across the season. The invitation is open on both sides. If you are doing something with agents in your community that the two of them have not thought of, they want to hear it, including the experiments that did not work.

Chapters

Time	Chapter
00:00	Cold open: GEO belongs to community
00:57	Welcome, and why this show exists
03:01	Amplification, not replacement
04:32	Underfunding and the business case problem
05:48	Bots now outnumber humans on the web
07:14	The support question that never reached the forum
08:12	How AI reads a community thread
09:51	Authority signals: badges, staff attribution, upvotes
12:00	Temporality, llms.txt, and ontologies
13:17	The prediction: do the community work, not the file format
15:23	What replaces traffic as the metric
16:26	“If traffic is your only ROI, you’re already in trouble”
16:49	Citations, bot-versus-human data, and GEO dashboards
18:11	Why GEO belongs to the community team
20:28	Attribution, and the ad model that’s coming
21:34	The Work Map, and owning its gaps
23:51	Why Marius is doing this
26:47	“I’ve always been lazy”: outcome-driven automation
28:54	Translation, humans don’t scale, and the reframe
33:09	AI slop, and why the brand doesn’t care
35:15	Close

Key takeaways

GEO is a credibility opening for community teams. Community is usually the only user-generated content a brand owns and can influence, which makes the community team the natural home for it.
The signals that make an answer legible to a model are things community managers already control: who answered, what standing they have, whether the community endorsed it, whether it was marked correct, whether it’s current.
The highest-leverage GEO work is not a technical file. It is recognizing authority, marking correct answers consistently, and retiring stale content.
More of the web is now crawled by bots than browsed by humans, which breaks the page-views story community managers have been telling executives.
Citation and bot-traffic data is starting to appear in platform back ends, which gives community managers a replacement metric for the traffic they’re losing.
Automation in community is not new. Profanity filtering, unanswered-question queues, and translation each moved work off the plate. AI is the next turn of that wheel, not a break from it.
The brand’s interest is that the member got a correct answer and can use the product. Insisting the answer come from the community is the wrong hill.
AI slop moderation is arriving in communities, and the hard cases are well-intentioned: non-native speakers using AI to be understood.

A correction, and what it cost

In this episode Marius cites a figure: 65% of people say they trust AI answers over human answers. The number is wrong, and the claim underneath it is wrong too.

The research he was reaching for is a Clear Spark Digital survey of 2,127 US adults from April 2026. What it found is that about 68% of respondents trust AI-powered search results as much as or more than traditional Google search results. “As much as” is doing most of the work in that sentence. Broken out by age, the share who trust AI more than Google is closer to a third: 33% of adults 61 and older, 23% of Gen Z. And it’s a comparison against Google, not against people. Other research points the other direction entirely. YouGov’s 19-market survey puts the US last for AI search trust, with search engines ahead by more than forty points.

So: a misremembered number, a misstated comparison, and a stronger claim than the source supports. That’s slop. Human slop, which is the kind that doesn’t come with a report button.

Here’s the part worth sitting with. I heard it, I had a feeling about it, and I let it go by because we were rolling and I didn’t want to break the take. I meant to check it with Marius before we published, and I didn’t get to it until the episode was already cut. Every step of that was me handing my own judgment over to the momentum of the recording.

That’s the exact thing I keep telling community managers not to do with AI. Don’t give up your agency to the tool, the workflow, or the schedule. Turns out the rule applies to podcasts too.

What we’re doing about it: the claim stays in the episode and in the transcript, footnoted, because quietly removing it would be its own kind of dishonesty. It no longer leads the show. And the argument this episode actually makes, that GEO is a community job, never needed the number in the first place. It stands on what community managers already control.

Sources

Clear Spark Digital, AI search study, April 2026 — https://clearsparkdigital.com/ai-search-study/

About the show

CommunityAssistant.ai is the podcast and open resource library exploring the human work of community management: what AI can take off your plate, what only you can do, and how to grow into the role that’s coming. Hosted by Chris Heuer and Marius Ciortea. Presented by Higher Logic.

Our sponsor. Community Assistant is made possible by Higher Logic, supporting the practitioners who build the communities that matter.

Our independence. Sponsors make this show possible. They don’t make this show. Marius works at Higher Logic; he appears here as a practitioner speaking for himself, not for his employer.

Overview

Why this show exists

GEO, and Marius’s series

What replaces traffic

The Work Map, and its gap

Why Marius is doing this

The waves before this one

AI slop, and the awkward part

What’s ahead

Chapters

Key takeaways

A correction, and what it cost

Links

About the show

The Purpose of Our Show

AI Trusts Humans.

Value of Human Content

Balancing Tech and Human Touch

Humans Don't Scale

The Potential for AI in Community Management

Community Managers Opportunity as GEO Managers

AI Can't Replicate The Human Voice in Community