Salesforce CMO: Your Marketing Playbook Has a One-Week Shelf Life

Salesforce CMO Erin Oles told an Advertising Week New York audience that calling anything a "marketing playbook" is almost an oxymoron now, because customer behavior now shifts faster than the funnels marketers built around it.

Oles spoke with Jenny Rooney on the Marketing Vanguard podcast, recorded live at Advertising Week New York. Her central argument: buyers changed the route, and the traditional funnel assumptions no longer hold.

What does this change for the funnel?

The old model assumed marketers could get buyers to identify themselves, land on an owned property and move neatly through nurture. Oles said buyer behavior increasingly does not look like that. Buyers research through LLMs, talk to peers and move between sources long before a brand ever sees them.

Her replacement is what she calls "surround marketing": build enough relevance and credibility around the customer that the brand matters wherever the buyer chooses to look, instead of obsessing over pulling people onto owned properties and into nurture sequences.

Your own website may be working against you

Oles shared a finding from an event with 30 CMOs: every participant, including herself, discovered owned properties actively working against them in LLM results. The uncomfortable lesson is that a brand's AI-search problem may originate in content the company already owns.

She also flagged a wider gap between CMOs who say AI visibility matters and those who actually use what they learn to change their marketing. Answer engine optimization, on this view, is not simply SEO with a new acronym — the mechanics of discovery inside LLMs reward different work.

Give AI a problem, not a use case

Oles' clearest rule for AI adoption is to start with the problem, not the tool. Salesforce applied this with an SEO agent that found 10,000 broken links and pages across the company's owned web properties — work Oles estimates would have taken roughly 5,000 human hours. The AI handled it within hours.

The win was not "using AI," she argued. The win was returning thousands of hours to people for work that genuinely needs human judgment. Her broader stance on AI is learning by doing rather than mastering technology in theory, and handing repetitive work to machines.

At Salesforce, that approach covers AI search visibility, customer discovery, conversational agents and real-time campaign optimization.

Attribution and the value of context

Oles said yesterday's attribution models may not explain tomorrow's customer journey. Traditional attribution assumptions are starting to crack precisely because discovery now happens in places — LLMs, peer conversations — that legacy measurement never counted.

She added that context becomes more valuable as access to intelligence becomes universal: when everyone has the same AI capability, the advantage shifts to whoever understands the customer situation around it. AI-driven discovery, in her framing, could push marketers beyond the traditional boundaries of marketing.

Oles began her career in sales, an experience she credits with teaching her how quickly you can tell whether a message actually resonates with a real person. That test — does this work on an actual buyer, in the place the buyer actually is — is the standard she now applies to AI, attribution and the funnel itself.