Salesforce’s latest State of Marketing report is full of big numbers around AI adoption. As I read through it, a few things stood out — and not all of them in the way you might expect.

At first glance, the numbers feel encouraging. There’s real progress happening. But there are also some pretty big gaps between what marketers say they’re doing and what it actually takes to make AI work at scale.

Let’s start with this:

  • 72% of high-performing marking teams are using generative AI
  • Only 17% are using agentic AI

That gap? That’s where the real story is.

Generative AI Is Just the Starting Point

Remember when personalization meant “Dear <first name>”? That’s roughly where we are today with generative AI. Using ChatGPT to edit your emails, create your subject lines or speed up content creation is valuable, but it’s just the tip of the iceberg.

AI becomes powerful when it stops being a tool and starts becoming part of a system.

The Shift from Prompts to Systems

Generative AI drafts. Agentic AI connects, retrieves, analyzes and executes. That distinction matters. Agentic AI is what allows marketers to tap into unstructured data — documents, presentations, internal knowledge — and actually use it to inform decisions and workflows.

At Simantel we’ve seen this firsthand. We recently connected a custom GPT to our wiki. Instead of clicking through files, or using the native Boolean search, we can now ask our wiki questions in plain language. The system returns a straight answer and points us back to reference on the Intranet where we can explore more.

The custom GPT is our agent and the Intranet sources (PDF, PPTs) are our unstructured data. Together, they create something far more powerful than either could alone: a system that accelerates how our people find and use knowledge. That’s the difference between dabbling in AI and operationalizing it. And right now, only 17% of marketers are there.

The Optimism (and Reality) of “8 Hours Saved”

Another stat that stood out: marketers expect AI agents to save them up to eight hours per week. I love that optimism. But not to be an Abby Downer here, this may be an over estimation.

Yes, AI may give time back to individual marketers. But it creates new demands elsewhere. AI doesn’t run itself. It requires:

  • Governance
  • Integration
  • Data management
  • Ongoing optimization

In other words, AI will always need humans. As we lovingly say at Simantel, “We need more nerds!” When you move from prompts to systems, the work shifts. The hours reclaimed by practitioners are often reinvested in oversight, orchestration and refinement.

For marketing leaders, this is critical. If you’re building a case for AI investment, don’t just talk about efficiency. Talk about the infrastructure and talent required to sustain it.

While We’re Figuring It Out, Search Has Already Changed

While teams are still figuring out internal AI maturity, the market has already moved. AI-driven search is reshaping how buyers find answers.

According to the report:

  • 85% of marketers say they are reshaping their SEO strategy
  • 88% are in the process of optimizing for AI-driven search experiences like ChatGPT and Google’s AI Overview (AIO)

If you’re not part of the 88%, you are already behind. I know that sounds scary. But, take a look at your organic traffic. If you haven’t recently, now’s the time. Better yet use a platform like SEMrush that can show you AI Overview tracking. Because this shift is already happening.

SEO Isn’t Dead. But It Has Evolved.

The shift is as much technical as it is philosophical. Yes, schema, tagging and site structure still matter. But the majority of marketers don’t own the whole website ecosystem for their company. They own pieces of it — pages, hubs or content that contributes to the website.

But even if the only thing within your sphere of control is content, you can still make an impact. Focus on intent. AI search rewards:

  • Clear answers
  • Deep expertise
  • Content built around real user intent

Long gone are the days of stuffing content with keywords. Now, it’s about clarity and the depth of the answers you’re providing. Optimization means genuinely answering the questions your audience is asking and doing it better an anyone else. Think less about ranking for a phrase and more about being the best answer.

So, Where Do You Start?

If there’s one takeaway from this year’s report, it’s this: Most marketers aren’t behind on AI because they’re not using it. They’re behind because they haven’t connected it.

The next phase of marketing isn’t about generating more content faster. It’s about building systems that connect data, surface insights and help you make smarter decisions.

That’s how we close the gap between generative and agentic AI. And it’s where the real competitive advantage lives.

What were your biggest takeaways from this year’s report? How are you planning to integrate AI into your work this year? I’d love to hear from you! Connect with me on LinkedIn or reach out to start the conversation.