AI Search Should Be Run as Marketing, Not SEO, Industry Strategist Argues

ChatGPT sends under 1% of Google's traffic, but recommended brands see 2.5x higher visit rates within a week. A new industry framework argues AI search demands marketing ownership, not SEO tactics.
By Tom Whitfield
3 min readAt a glance
- ChatGPT sends less than 1% of the traffic that Google sends
- Similarweb found ChatGPT brand recommendations made users 2.5x more likely to visit a brand's site within seven days
- Most large publishers block AI crawlers under copyright protection, reshaping their discoverability strategies
- Jan-Willem Bobbink reported B2B wins came mostly from fixing missing or hard-to-find information
ChatGPT sends less than 1% of the traffic Google does, yet Similarweb research found users who received a brand recommendation in ChatGPT were 2.5 times more likely to visit that brand's website within seven days than users recommended a competitor. Those two figures anchor a widely shared industry argument this week: AI search is a marketing problem, not an SEO one.
The author, a longtime SEO practitioner who has spent recent months working on AI search, information retrieval and discoverability strategy, argues the industry has overcomplicated the shift. In their view, practitioners chasing citation tactics and "cheap visibility tricks" are missing that AI systems synthesize a brand's positioning, reputation, expertise and presence across the whole internet — not just its website.
What changes for practitioners?
The recommended approach starts with audience research rather than prompts. The author uses SparkToro to map where target audiences spend time, which sites they visit and who influences them, then combines that with traditional search research and brand positioning audits. For their own audience, they state plainly: "LinkedIn is a non-negotiable."
Technical foundations still matter. The framework calls for:
- Bot management and crawler access reviews, recognizing that training crawlers, search crawlers and user-triggered retrieval systems serve different purposes
- Entity management: consistent identity, explicit positioning, evidence such as credentials and case studies, then corroboration from third-party sources
- Monitoring identity evolution over time, using tools such as Alertmouse
The author distinguishes sharply between what a model has learned in training, what a system retrieves, and what it ultimately says about a brand. Being crawled does not guarantee inclusion in training data; being retrieved does not guarantee a citation; a citation does not guarantee commercial value.
Publishers face a different equation
Most large publishers block AI crawlers, sometimes indiscriminately, citing copyright protection. The author frames this as a legitimate commercial trade-off — comparable to a paywall — but one that requires understanding what visibility is being surrendered and where value might exist instead, through content licensing, data partnerships or direct audience relationships.
Practical tactics get support from named testers. Jan-Willem Bobbink, cited after three weeks of SEO and GEO work for B2B clients, said: "The changes that made the biggest difference were mostly about information that was missing or hard to find." Adding pricing pages and visible case studies, the argument runs, gives both buyers and models a clear reason for inclusion. Statements such as "Client Y saw a 30% improvement in deliverability metrics" can noticeably improve citation rates.
Measurement and ownership
AI visibility scores are described as "hugely flawed" given personalization and absent prompt data, but better than nothing. The recommended method tracks a representative prompt set across four areas — branded queries, existing positioning, desired positioning, and the wider competitive market — while treating visibility, referrals and citations as supporting indicators rather than success metrics.
The primary measure should be incremental business impact: baseline demand, branded searches, conversions and commercial outcomes before and after campaigns.
Organizationally, the author argues responsibility should not sit solely with SEO teams. For larger businesses, the recommendation is a Head of Marketing or Head of Discoverability with authority across technical infrastructure, audience research, positioning, PR, content and measurement. SEO professionals have three viable directions: technical crawling and retrieval work, building discoverability tooling, or full marketing.
The author's own appearance in AI recommendations for news SEO expertise — achieved without AI-specific optimization, through years of publishing and relationship-building — serves as the closing evidence that durable brand work outlasts any single algorithm or interface.
Source: searchenginejournal.com
Tom Whitfield
Show full bio
News editor covering regulation, policy and their impact on Social platform news and policy at Reach Brief.
Related articles
6.9% of ChatGPT Conversations Show Purchase Intent, SparkToro Panel Finds
ChatGPT holds 81.4% of global chatbot traffic as growth returns
ChatGPT Ads Earn $0.89 Per Session, a Third of Google Search's $2.45
OpenAI Launches Business Hub: ChatGPT Presence Management Goes Live
AI Overviews Now Hit 48% of Google Queries — and Last-Click Is Blind to It