CMOs Are Building AI Scorecards (But They’re Tracking The Wrong Score)

This article was originally published on Forbes Technology Council.

There is a number circulating in boardrooms right now, and it is making CMOs nervous: 20%.

That is the cost-reduction target that CEOs and CFOs at some of the largest companies in America are asking marketing leaders to hit, through AI, efficiency gains, or both, over the next two years.

A Spencer Stuart survey of roughly 90 senior marketing leaders, conducted last November and reported in The Wall Street Journal, put real data behind what many CMOs had already been hearing in budget conversations.

The Numbers Are Striking

More than a third of CMOs say they expect to reduce headcount over the next 12 to 24 months. The drivers: the growing use of AI tools and the elimination of overlapping roles.

Larger companies with annual revenue over $20 billion are even more aggressive. Half say they are anticipating staff reductions, and more than 30% say cuts have already occurred.

Despite the increasing reliance on AI tools over human staffers, CEOs have not yet seen anticipated returns from their AI investments, according to a separate Teneo survey also covered in The Wall Street Journal.

These tools hold promise. The problem is how that promise is being measured.

The Scoreboard Problem

The use of AI tools by marketing departments is saving time, reducing vendor costs, and increasing output exponentially. Automating ad voice-overs, synthesizing customer polling data, spinning up creative variants at scale: these have real dollar values, and getting credit for them matters in a budget conversation.

But this is a dangerously incomplete picture. CMOs focused exclusively on an efficiency scorecard are optimizing for the wrong half of the equation. Cost savings keep the lights on. Customer understanding is what grows the business.

A BCG analysis of AI adoption across industries shows that companies using AI to drive deeper customer intelligence through predictive insights and personalization are more likely to see above-average revenue growth than those that do not.

The Real Edge: Customer Intelligence

Gaining an edge through AI-powered marketing does not come from access to tools. Competitors have the same access. It comes from superior customer intelligence: how precisely you personalize at scale, how quickly you run experiments, and how accurately you predict what a customer segment wants before they ask for it.

The challenges marketing organizations face with their AI implementation are not about talent or budget. They are about data quality and technology integration. AI tools are only as good as the data they are built or trained on.

When Companies Get It Right

One consumer goods client, under pressure to rapidly deploy AI-generated content, took time to clean and consolidate customer data that had been siloed across three different platforms. Despite the four-month delay, this proved to be a good call. Their due diligence resulted in higher email engagement rates, shorter cycle times, and the identification of a high-value customer segment previously overlooked.

Investments in data quality upstream make every AI tool deployed afterward measurably more accurate and actionable.

The right question is never only "How much will this save us?" It is "What will we learn from this, and how fast?" That measurement discipline is what separates organizations building durable competitive advantage from those running in place.

What Scorecards Fail To Measure: Brand Trust

Brands that over-rely on AI technology or fail to manage its use thoroughly risk brand damage. Some companies have already learned this the hard way:

• A Chevrolet dealership's AI chatbot recommended a competitor's truck to potential buyers.

• A Vogue ad for Guess included AI-generated models with proportions that defied reality and were roundly mocked.

• When Taco Bell turned to AI to power drive-thrus, customers struggled to place even simple orders, and in one case the AI accepted an order for 18,000 cups of water.

Trust takes years to build and seconds to erode. Using AI to mass-produce bland, undifferentiated content to cut costs is a reliable way to destroy it.

Build Your Intelligence Engine First

The efficiency scorecard is a table-stakes requirement in 2026. CMOs who cannot show measurable cost reductions from AI investments are going to have difficult conversations with their CFOs. That work needs to happen.

But the organizations that will look back on this moment as a turning point rather than a cost-cutting exercise are the ones building the customer intelligence infrastructure now. That means:

1. Investing in data quality, not just AI tooling.

1. Designing measurement systems that capture revenue lift from personalization and faster experimentation, not just labor savings.

1. Treating the talent redeployment question as a capability-building opportunity, not a headcount management problem.

The score that matters most is not how much you saved. It is how much better you understand your customers than you did a year ago, and how much faster you can act on what you know.

Read the full article on Forbes Technology Council.