The Marketing Productivity Paradox and Evolving Cost Pressures. A Guide for the C-Suite

AI capabilities have already had a significant impact on marketing, and that impact will continue to accelerate. The prevailing view among business leaders is that AI is likely to help reduce the cost of marketing, an expectation shared by many marketing practitioners. But this assumption deserves closer examination.

Understanding how AI may ultimately affect marketing expenditure requires a more holistic view of the underlying economics of marketing execution. Predicting whether AI will drive an increase or decrease in marketing expenditure should begin with an analysis of how marketing cost dynamics are structured.

The fundamental cost structure of marketing execution

The overall cost of marketing execution is determined by two primary drivers:

P) Cost of Production: The cost of producing the artifacts that construct and communicate a marketing message.

M) Cost of Media: The cost of carrying that marketing message to its intended audience.

The total cost of production (P) can be divided into two broad categories: labor costs (skills and human resources) and tool costs (technology and materials).

The cost of media, or visibility (M), encompasses media in all its various forms, online and offline, from advertising, sponsorship, endorsements, pay-to-publish and any other forms of amplification. For the purposes of this analysis, we can include “organic” (or earned) media in the cost of media. While organic media is not paid in the usual sense, its availability and reach is impacted by the direct or indirect provisioning of paid media, as discussed in the article on the scarcity and volatility of organic media.

The impact of decreasing marketing production costs and its compounding effects

The cost of a unit of marketing production has been gradually but consistently declining for some time, driven predominantly by three trends.

First, the widespread offshoring of certain marketing production services to lower-cost economies is reducing labor costs. Second, the introduction of software applications has reduced marketing’s reliance on expensive, IT-heavy production. Third, the significant expansion of training and development opportunities for people entering the marketing production workforce has created a healthy skills supply and, therefore, competition among qualified marketing production specialists across a wide range of disciplines.

However, hard evidence shows that greater production efficiency over a prolonged period has not automatically translated into lower overall marketing expenditure for businesses. This exposes a fundamental misconception in marketing economics: reducing the unit cost of production does not necessarily reduce the total cost of marketing but instead increases production complexity.

An example of this dynamic is the rising cost of the martech stack, the technologies and tools used to support marketing production and execution.

Martech was originally positioned as a cost-containment strategy, enabling organizations to replace expensive, bespoke IT systems with more flexible and economical SaaS solutions.

In the immediate term, martech tools increased the productivity of marketing teams and helped reduce the unit cost of production. Over time, however, this contributed to increased competitiveness and the need for complex architecture solutions. As a result the economics of martech shifted and marketing software tools now account for an estimated 25–30% of marketing budgets, compared with less than 10% less than fifteen years ago. The number of available solutions grew from roughly 150 in the early 2010s to more than 12,000 by 2025, pushing marketing teams to adopt increasingly complex technology stacks and forcing organizations to reassess how martech is evaluated and implemented.

The AI cost causal loop: Marketing efficiency rate-limiter and primary upward cost drivers

At its most fundamental level, increasing marketing productivity means increasing the rate at which organizations can produce marketing artifacts. That sounds inherently beneficial. However, the ability of those artifacts to achieve cut-through is constrained by a hard rate limiter: the finite availability of media inventory and audience attention.

An organization’s marketing production efficiency may increase dramatically, but its receiver capacity has a hard ceiling that it cannot change or bypass. This ceiling is determined by the availability and accessibility of media inventory and ultimately by the audience’s available share of attention and interest.

A steep increase in marketing production output is likely to determine a saturation of media inventory and relative price increase, and at the same time a decline in the addressable audience's rate of response. Because in today's performance-driven media buying ecosystem, the price of media exposure is ultimately determined by the propensity of an addressable audience to respond to or engage with a message, as the share of interest declines, the effective availability of media inventory also declines.

As effective media reach becomes more expensive—either because unit costs increase or because a higher frequency of exposure is required to generate the desired level of response—organizations will need greater reach, higher frequency, broader distribution or more sophisticated targeting to achieve the same level of impact. This creates a reinforcing causal loop.