Adapting Pricing Models: The Impact of Generative AI on Marketing Agencies

As marketing agencies integrate generative AI into their operations, they’re reevaluating how they charge clients to reflect the costs associated with AI tokens.

Navigating the Cost of Generative AI in Marketing: Pricing Models Evolve

As marketing agencies increasingly integrate generative AI into their workflows, a crucial issue has emerged: how to adapt pricing models in light of the associated costs of AI tokens. With the unique structure of generative AI usage—where each prompt and response is measured in tokens—agencies are finding themselves at a crossroads, weighing whether to absorb these costs or pass them onto clients.

The Token Economy: More Than Just a Cost

Generative AI does not operate without expense. Companies like OpenAI utilize tokens to meter AI compute, charging based on the volume of text processed. While each token may seem inexpensive, costs can quickly accumulate. A striking example illustrated by recent campaigns reveals that Coca-Cola’s ad project alone required 70,000 prompts, leading to millions of tokens being used.

Johnny Rohrbach, co-founder and director of partnerships at Silverside AI, notes, “There’s no one model that’s the one silver bullet… All of those have different token economics.” As agencies navigate this landscape, they must decide how to manage these costs and align them with their service offerings.

Diverse Pricing Strategies Emerge

At full-service agency Merge, Kyle Smith, the CTO, maintains a flexible approach by passing token costs onto clients based on specific project needs. This case-by-case model allows for precise billing aligned with the actual usage of AI. In contrast, Big Spaceship adopts a slightly different approach—considering AI compute costs as a standard budget line item, much like catering or equipment rentals.

Silverside, known for its innovative campaigns like the Super Bowl ad for Svedka, utilizes a subscription model with “seat” pricing akin to systems used by Salesforce, recognizing the potential for scalability and predictability in costs.

However, not every agency is ready to shift these costs onto clients. RPA has opted to absorb token costs, with Lisa Herdman indicating, “We can’t charge our client for something we don’t know is actually going to work for them.” Such caution reflects a broader sentiment among agencies reluctant to treat AI expenses as a new revenue stream, as echoed by Chris Neff of Anomaly, who views it as a potential “money grab.”

Economies of Scale: Leveraging Commitments

Companies like Pencil, part of Brandtech, are adept at negotiating better rates with LLM developers by establishing bulk agreements. This allows them to offer “generation credits” to clients based on their commitments, fostering an environment where volume leads to reduced rates. Will Hanschell, CEO and co-founder of Pencil, highlights that this structure supports both simplicity for clients and alignment of incentives.

Ensuring Accessibility: Minimal Fees and Initial Cost Recovery

For media agencies, a different strategy is emerging. Horizon Media, for instance, introduced its AI-powered media planning tool, Blu, while carefully managing client costs through nominal fees that prioritize cost recovery rather than profit. As Krish Kuruppath noted, understanding that client onboarding is often where expenses amass helps manage expectations.

Kepler embraces a similar philosophy, embedding AI tools within their retainer models without charging directly for token usage. As Peter Rice emphasizes, the focus remains on delivering impact rather than getting caught up in the minutiae of token expenses.

The Debate on Transparency and Fair Pricing

The ongoing discourse around token costs parallels debates in the advertising industry about transparency and fair pricing. With customers likely to expect accountability, especially as AI becomes more central to marketing strategies, agencies may soon find themselves needing to track and audit token usage, as suggested by Ebiquity CEO Ruben Schreurs.

Emphasizing Value Over Cost

Ultimately, the consensus among industry leaders is that agencies should focus on demonstrating value to clients rather than merely billing for inputs. As Hanschell states, “You want to compete on merit, not on lowest price.”

The evolving token economy presents both challenges and opportunities for marketing agencies. As they continue to integrate generative AI into their strategies, the importance of adaptable pricing models and a focus on value delivery will be paramount. The future lies not just in understanding costs but in harnessing AI to drive meaningful business results for clients.