In the fall of 2014, Steve Kerr inherited a Golden State Warriors team that averaged 247 passes per game (the lowest in the NBA).
He set a target of 300, noting that previous champions ranked near the top in ball movement. Then he rebuilt the offense around constant motion: screens, cuts, ball reversals, positional fluidity. Where most teams ran isolation plays, the Warriors ran them at nearly the lowest rate in the league.
The results were great, of course.
Now, I’m a casual basketball fan (other than Knicks playoffs runs – and I’ve scheduled this post out a few weeks, so we’ll see how that plays out). But it seems like isolation basketball is transforming into integrated basketball. And isn’t this the plot of many sports movies? Selfish talent learns to be team players, team wins (but so do the players, emotionally and spiritually)?
Marketing, if you worked at a certain size company, ran the same isolation playbook for two decades. AI search is disrupting that from all angles (metrics, inputs, skills, technology, etc.)
We have to learn to work together if we want to drive performance.
“Who owns AI search visibility?”
This question threads many debates at the moment.
Good GEO is good SEO, so SEO owns it. Off page dominates citation sources, so PR/comms owns it. Brand owns it, because it’s multidisciplinary and AI answers are essentially digital mirrors of your brand perception.
There’s no simple answer. I don’t profess to have one. It’s probably variable across companies and industries and will evolve as often as AI models themselves do.
While customer acquisition channels still function independently (i.e. the value of a review on G2 still matters for human readers of G2 category pages), AI search has collapsed channels into a single interface that abstracts the source materials to end users.
The inputs to that interface are distributed across every function in marketing.
No single specialist truly has ownership over the output. And the org chart hasn’t caught up.
The Channel Era
For two decades (or more), digital marketers were largely defined by their channel. SEO specialist. Paid acquisition manager. Social media strategist. Email marketer. Each owned a platform, understood its algorithm, and optimized within its walls.
(Except for those growth marketing dilettantes who suffered benefited from a lack of focus and typically worked at startups with too much responsibility)
This never made a ton of sense from a bird’s eye view (dark social, WoM, interaction effects across channels, etc.), but click-based telemetry and attribution made it functionally easy to set goals and make progress on a channel-by-channel basis.
And of course, each platform had its own rules. Google’s algorithm rewarded different signals than Facebook’s. Your SEO traffic was your SEO traffic. Your social engagement was your social engagement.
This is still true, and understanding platform/channel dynamics still matters a ton.
The T-shaped marketer (deep expertise in one area, working fluency across others) was an early corrective that attempted to create some flexibility and holistic thinking in marketing career progression. Growth marketing was the operational version: someone who could analyze data, run experiments, and pull levers across channels (and product) to drive a composite and measurable outcome. But even growth marketers operated in a world where channels produced separable results, delineated by columns on a spreadsheet. You could trace which lever moved which number, granularly tweaking knobs like Facebook ad click-through-rates.
AI search is disrupting this model.
The Convergence
When a buyer asks ChatGPT, Claude, Perplexity, or Gemini a question, the answer draws from many places simultaneously, including its training data. Or rather, a probabilistic, ever-changing, multi-modal and multi-domain set of sources that can be impacted by paid, earned, owned, community, and more. One output. Many inputs.
Last year, Reddit was the talk of the town. If SEO owns AI search, does that mean they now own community management?
The past few months, LinkedIn has been the talk of the town. If SEO owns AI search, does that mean they now own organic social?
G2 reviews, Gartner Peer Reviews, and PeerSpot are cited by ChatGPT and Perplexity for product comparison queries. Review management is now search strategy.
I’m not arguing that SEO shouldn’t own AI search (it is, after all, still by and large a search problem). If anything, I’m biased to say that SEO should raise their hand and take on AI search, as well. But in practice, SEO teams are often disconnected from other teams’ efforts, or at the very least, it’s an uphill battle to work into their roadmaps. I know this from practice as a growth marketer at HubSpot trying to integrate SEO, affiliate, customer marketing, and paid efforts for the Surround Sound SEO program. It wasn’t easy.
AI search visibility is, in many ways, an emergent property of everything your organization does and also everything your competitors and customers do. Even the tactics many talk about on LinkedIn, such as chunking, FAQs, content updates, etc. are only about retrieval, which is one piece of a larger puzzle. But what does passage retrieval mean if your brand is absent from every other relevant page on the internet? “Questions as headers” may marginally increase your citation rate, but it’s a drop in the ocean in terms of brand visibility.
The Challenge
AI search is a conceptually similar problem as classic search: deliver information that users want through an intermediary platform.
The challenge is that the input sources a) differ from brand to brand and b) change often. So while aggregate citation data may tell you that Reddit is the most prominent source of influence, that may or may not be true for your brand individually. And if it is right now, it may not be in 3-6 months.
Thus, AI search requires a sort of orchestration or portfolio management approach that both retains a level of cohesive brand messaging and throughput (consistency is important for legibility, broadly) and also a tactical agility to run experiments and pivot resources when influence or consumer behavior patterns change (e.g. shifting from thinking about citations in ChatGPT to agentic web visits from something like Claude Cowork/Claude Code).
Again, it’s not simple or obvious how to structure a team right now.
We see a lot of client data. We also sit on calls with everyone from CMOs to SEO specialists to PR teams, analysts, growth marketers, and more. Our own engagements are increasingly broad and cross-functional. And what I can say is there’s no universal playbook or story to tell apart from the foundations you all know and that we’ve written about before.
While vendors who make content production functionally easy will talk about programmatic approaches to publishing or updating, for many of our clients, owned content is simply not the major problem to solve. So these solutions don’t map perfectly to AI search performance. In other cases, it does and is a great unlock.
In every case, however, I can say that no single channel delivers the full signal. The composite is what matters. This is true for a startup attempting to show up on AI visibility charts, and it’s also true of Salesforce who acquired a new company and wants to integrate the new functionality into their messaging that shows up in AI answers.
The Role That Emerges
So what does the marketer who owns this look like?
Not the T-shaped marketer, I don’t think. Or at least not the version from a decade ago.
The T was built for a world where channels were independent and you needed 1-2 deep skills to anchor your career (mine were SEO & CRO, by the way). The role that AI search demands is something different: a marketer whose primary obsession is the customer journey as both humans and AI platforms experience it, and whose core skill is allocating investment across disparate efforts that produce composite results.
Three traits define this person.
- Customer-obsessed, not channel-obsessed. They start with how buyers discover, evaluate, and choose, then work backward to which channels, platforms, and content types influence that journey in AI-mediated environments. They’re less interested in citation studies, and more interested in consumer behavior studies.
- Oriented toward perception and visibility metrics (and ultimately revenue outcomes). Not just traffic and conversions, or even citation share. But someone who can look at a constellation of metrics (oh how I miss the days of the OMTM), read patterns across several inputs, and prioritize the biggest levers against a brand’s goals. Sometimes this is brand mention frequency. Sometimes it’s sentiment or accuracy of outputs.
- Cross-functional allocation as core competency. The hard question isn’t “how do we optimize our blog for AI search.” It’s “given limited budget, how do we allocate across SEO, LinkedIn content, community management, creator partnerships, digital PR, review generation, and original research to maximize our composite AI visibility?” This is portfolio management, not channel management.
Heck, maybe this is a brand executive (like Eli Schwartz persuasively argues).
Maybe it’s a Director of Digital Visibility (combining SEO expertise with a broader remit and resources).
Maybe it’s a new role or a pod structure like how the growth role emerged to be able to work across product, engineering, marketing, and data.
What Changes in Practice
If the role changes, the operating model has to change with it.
- Shared visibility dashboards. Any shortcomings aside, tools like Profound, Peec, Meltwater’s GenAI Lens, Scrunch, or AirOps help to give teams a shared language and input metrics that cut across teams. It’s now easier, for example, to position your PR work as affecting AI search performance. Channel KPIs can of course still be a part of your dashboards, but the more we can share metrics, the better off we are.
- Cross-functional pods over siloed teams. An AI visibility pod might include an SEO strategist, a content creator, a community manager, a PR specialist, and a creator partnerships lead. They operate against shared citation and mention targets, not just channel-specific KPIs. Agents and AI systems help center context and also scale execution, but experts still own outcomes.
- Creator ecosystems as infrastructure. PartnerStack’s integration with AI visibility tools like Evertune lets brands identify which creators are already cited by AI systems and build partnerships with them. Reddit’s influence isn’t merely a marketing or SEO issue, but a brand reputation and customer experience issue. Social listening comes back with a vengeance as these data points are slurped up by the models, no matter how well optimized your product pages are.
- Conversion baked into the gestalt journey. Conversion rate optimization has often been looked at as an isolated lever applied on top of a given channel or tactic. You can optimize landing page conversions from paid campaigns. You can optimize a homepage or a pricing page. But conversion thinking needs to be integrated throughout the process, from the topics you target to the direct/non attributed traffic that is now hitting your homepage even though someone found you in an AI answer.
Candidly, I think organizations will have to experiment and try out a few different structures and land on one that works for their unique situation. And even at that point, there are still many open questions (such as revenue attribution and how input KPIs ladder up to those).
An open mind and growth mindset are also key nowadays.
It’s Your Website, But Not Just Your Website
When a buyer asks an AI system about your category, the answer isn’t built solely from your blog. It’s assembled from your reviews, your LinkedIn posts, your community presence, your PR coverage, your creator relationships, your original research, and yes, your blog. The blog is one input among many.
Deep channel expertise still matters. You still need people who understand the nuances of what works within SEO, within social, within PR. The algorithm-level knowledge is real and valuable. But the outcome is composite now. The channels feed a single mouth.
The marketing leader this era demands isn’t deeper in one channel.
They’re wider across all of them, with the judgment to know where investment compounds and where it dissipates.
Honestly, as I write this, I realize we’re harkening back to a tried-and-true archetype, someone who understands integrated marketing communications. Dare I say, a brand marketer?
Want more insights like this? Subscribe to Field Notes
