OpenAI Needs Two CMOs Because It Has a Problem Marketing Can’t Solve

Salesforce under Marc Benioff was the master class in B2B brand-building: clear positioning, massive share of voice, distinctive brand assets, relentless customer success storytelling. Fleming absorbed all of it, so much so that ServiceNow followed the same playbook, and might even have applied it better. He knows exactly what enterprise buyers need to hear, and how often they need to hear it.

But Fleming’s Salesforce instincts will now collide directly with the OpenAI reality. 

Salesforce and ServiceNow sold trust as their core proposition. Their customers willingly passed on sensitive data, critical workflows, revenue operations. The brand promise was: we will never screw you. 

That promise was credible because neither company faced a structural conflict between its consumer product and its enterprise product. OpenAI does. The free ChatGPT that runs ads is the same model enterprise customers are being asked to trust with their most sensitive internal data. That dissonance is not trivial.

Then there is the competitive context. Anthropic is suddenly the beast in the room. It is growing faster, smarter, cleaner, and it’s winning in the regulated industries—financial services, healthcare, legal—where enterprise deals are biggest and trust matters most. 

Meanwhile Google’s Gemini is slashing prices and going after developer mindshare. Fleming is not entering a market where OpenAI can coast on brand halo. He is entering a knife fight with a smaller-than-expected pocket knife, and facing a guy with the machete and a woman with an unlimited arsenal of throwing daggers.

The dual-CMO structure is also a potential issue. OpenAI has split marketing into two roles: one for consumer, one for business. Presented as sophistication, it’s actually an admission of conflict. The consumer brand and the enterprise brand have diverged to the point where one person cannot credibly manage both. That is a strategy problem, and not one Fleming can easily solve.

None of this should be read as a prediction that Fleming will fail. He is exactly what OpenAI needs. If anyone can build a credible B2B brand on top of a messy underlying reality, it is someone with his track record. 

But good hires do not always fix declining fortunes. They work within them, around them, and sometimes despite them. 

The real question for OpenAI is not whether Fleming is talented enough for this job. The question is whether the company is willing to make the harder strategic decisions that Fleming will inevitably ask. About product separation, about the advertising model, about what OpenAI for Business actually stands for. 

Hiring a marketer like Fleming is just the start of the problem being sorted.