The Real Promise of AI in Marketing is About Expansion, Not Time Savings.

I think it's time we stop promoting generative AI primarily as a time-saving tool for marketers. For most marketers, it's not. At least not yet.

That's not an argument against AI. I'm incredibly optimistic about how it’s already transforming marketing. I'm experimenting with it constantly, building workflows and tools, testing use cases, connecting platforms, and learning where AI can and can't meaningfully improve my work.

But that experience has also made something increasingly clear to me:

As marketers, many of us were sold on an oversimplified value proposition that AI would save us time.

Granted, in earlier iterations, time-savings might have been the primary value proposition for generative AI in our marketing workflows, most notably for content creation. But today, AI in marketing holds so much more potential. And as long as we cling to the “time-savings” narrative, we’ve created unrealistic expectations for ourselves and for many of our stakeholders while still figuring out how to effectively operationalize an extraordinarily powerful and rapidly changing technology.And, as long as we focus on “savings” as a primary reason-to-believe in AI in marketing, we’re vastly limiting its potential, and creating a dynamic for marketers that is confusing and stressful.  

The AI implementation paradox

I recently asked a marketing leader how she was experimenting with AI.

Her answer was telling, and expected.

"I don't have enough time to do it. And the output isn't very good when I do, so I just end up doing the work myself anyway."

This person is in an organization facing substantial pressure from board leadership to not only implement AI, but to begin showing measurable ROI from this implementation.

And time is the biggest constraint.

Here’s the reality: effective implementation of generative AI tools requires an investment of time.

That's the paradox.

Marketing leaders are being asked to use AI to drive business impact, while simultaneously doing all the work required to figure out how AI can drive business impact. It’s as though there’s a perception that the implementation of AI into complex marketing workstreams that rely on multiple tech tools is as easy as waving a magic wand. 

Most importantly, the immediate needs that the business has of marketing haven't changed. We tell ourselves and our teams to "make time for what's important." But the urgency and demands of day-to-day priorities don't necessarily take a back seat because AI has arrived.

This isn't an isolated experience

The more conversations I have with marketers, the more frequently I hear some version of this story. So I asked ChatGPT to source credible data for me, which I verified. (Note: I acknowledge that asking AI to source the data did indeed save me time on this tactical research activity).

• 98% of CMOs are piloting or using AI, yet one in three senior marketing leaders say they aren't seeing the returns they expected. — Gartner, 2026

• 66% of marketers say learning new technologies takes significant time away from their day-to-day work. — Gartner, 2026

• And Deloitte found that more than two-thirds of organizations expected 30% or fewer of their GenAI experiments to fully scale within three to six months.

That last statistic is compelling, particularly as it relates to the paradox. Effective implementation is increasingly looking like a process measured in months, not days or weeks.

Access isn't implementation

Giving a marketing team access to ChatGPT, Claude, or another AI platform doesn't automatically create productivity. It creates access. There's a lot of work between access and impact.

Examining what effective implementation could look like helps clarify the amount of time required. 

• Teams could ideate and prioritize what problems should be solved by adding or replacing existing workflows with AI. 

• They could run experiments, perhaps dividing and conquering multiple experiments simultaneously within a matrixed team. (Some marketing teams are doing AI hackathons, which is such a clever way to make it fun and effective.)  

• They’ll develop workflows that could impact other business functions, which may require modification of the project. 

• They’ll recognize the need for quality controls that may surface moral and ethical conversations, which should lead to decisions regarding where human judgment and decision-making belong. 

• And there will be experiments that take a considerable investment of time that will ultimately be abandoned, and being okay with this is important. 

The reframe

I increasingly believe time savings undersells the larger opportunity for AI in marketing.

Instead, I think we should be asking:  What can we do with AI that we couldn't do before (or couldn’t do easily)? This is far more exciting, and here are a few examples:

• To strengthen website copy and improve SEO & AEO, use AI to analyze thousands of customer comments, including those posted in third-party channels such as Reddit. 

• Perhaps create an automation that continuously analyzes competitors’ positioning and performance in key geographies, that also dynamically updates sales enablement accordingly.

• Build a campaign effectiveness agent that dynamically tracks and reports campaign impact on metrics like earned Share of Voice or new user login.

This is a mere handful of ideas, but the opportunities are literally endless. And these aren’t examples of just getting our time back. They're examples of massively expanding the capability and potential of an existing marketing team, with direct line of sight to meaningful ROI.

I think this is a much more compelling value proposition.

AI can make the work we already do better

I originally shared these thoughts on a LinkedIn post, and one commenter described AI as being like having "more perspectives on a team."

This resonates with me.

One of my favorite ways to use both ChatGPT and Claude isn't automation at all. It's conversation.

I use them to pressure-test an idea with a specific ICP in mind. Challenge an assumption or find holes in an argument. Help me see around corners that I might not otherwise have access to.

The better I've become at working with AI, the more useful it has become, not simply because it completes tasks faster, but because it expands the way I think.

Idea: Treat AI implementation like building a product

If companies want to realize the full potential of AI in marketing, implementation should be approached with purpose, framing, and some flexibility.

1. Start with the problem and be specific

Start with the most important problems, like workstream barriers or inefficiencies, or use the expansion lens to ideate unrealized opportunities that have potential for measurable ROI. As an example, I might start with a focus on the OKRs or outcomes the marketing team is most accountable for, and consider whether the application of AI can help me remove barriers or increase our likelihood to hit and exceed performance. 

2. Scope

Prioritize the use cases most likely to create meaningful business value, scope the specifics (I use a variation on the who, what, why, where, when, how framework, which can set the stage for really solid prompting). Also, include a success measurement in your scope.

3. Allocate resources, identify constraints

Think: Budget, time, people. Having a marketing roadmap can come in handy.

4. Invest the time and build

Time is a fixed quantity. If it’s important, dedicate the time and shift other priorities accordingly. Stakeholder alignment will be important.

5. Measure results & iterate as needed

Make sure to document baseline data before beginning, and track what makes sense. Depending on results, iterate and improve (or abandon).

6. Scale what works (celebrate!)

Meanwhile, a little support goes a long way.

The potential of generative AI in marketing is as transformative as the launch of the Internet was. Our function has not experienced such a radical mindshift and demand that we change how we do almost everything in decades. If organizations want marketers to simultaneously perform their existing jobs, learn an entirely new technology, redesign workflows around it, and immediately demonstrate measurable ROI, we shouldn't be surprised if people struggle a little. 

If implementing and leveraging AI is an important priority, I encourage leaders to provide direction and collaborate on the priorities with reasonable expectations of the time it may take to show results. AI is evolving incredibly quickly. I fully expect that as the tech evolves and integrations become easier, the use cases will become clearer and the ROI will become increasingly obvious.

This is a fun time to be a marketer, especially if we can reframe our perception of this new technology in terms of what it can give or expand, instead of what it saves.