AI Agents. The Agency Business Model for the Next 40 Years (it’s right there in the name).

Picture of Judy Shapiro

Judy Shapiro

Editor-in-Chief at The Trust Web Times
Picture of Judy Shapiro

Judy Shapiro

Editor-in-Chief at The Trust Web Times

In 2009, I had lunch with a longtime friend, we worked at the same agency years ago. Now he was EVP of Strategy at “mega-agency.” Our conversation centered on marketing technology and that agencies were leaving marketing technology to be developed by adtech engineers who did not even like marketing.

As I expressed my frustration to him for what must have seemed an endless 15 minutes, my ever-patient friend smiled gently and said, “But Judy, clients hire us for our strategic abilities and creative. Digital is about tactics and media channels and that is not where clients value us the most. Digital is bleeding edge. That is what you are but that is not the agency business model.”

That stopped me cold because it never occurred to me that digital tech was bleeding edge. More than that, I was stunned; “Digital tech is not bleeding edge – it is the future of marketing.” Again, his gentle smile came across his face and he said, “Marketing technology is bleeding edge and that’s not what clients pay agencies to do.”

I continued. “Perhaps it is bleeding edge but agencies must change their business model to embrace technology and expand beyond fees based on labor.”

In an exasperated moment, I blurted out: “I really believe that digital marketing budgets will overtake TV advertising budgets. When that happens agency business models will be toast.”

I didn’t mean to be so dramatic but I believed every word of it. Now it was his turn to be stunned. “That’s not possible. TV advertising accounts for 95% of all spend. There is no way digital marketing will ever overtake TV advertising – ever.”

I doubled down. “Digital marketing will overtake TV advertising in about 5 years’ time.”

Again, his grin told me he thought I was wrong – very very wrong.

It turns out I was wrong about timing by four years but I was right about digital marketing’s dominance. It took a little longer than I thought (probably because agencies were active saboteurs) but by 2018, digital marketing eclipsed TV advertising. The rest is history.

So let’s take a stroll back in time to see how agencies reacted to every marketing technological innovation over the decades. The story is a consistent story of missed opportunities.

1990s – Internet, email – oh my

This was the age when the killer applications of online connectivity were coming into view.   

Prodigy did the agency rounds to introduce the concept of the Internet to agencies. Agencies were amused but unimpressed. I remember being in the main conference room where they did their presentation. I was sitting next to my friend and I nudged him saying this was the future. I knew that bemused look which revealed that, as interesting as this was, it had nothing to do with that he had to do for clients tomorrow. Agencies missed how transformative Internet would be in the end.

At the same time, by 1994, cable television had achieved mainstream adoption, with over 64% of U.S. households subscribing instead of relying on traditional over-the-air broadcasting. Agencies’ response was to dig in and avoid cable arguing that media fragmentation due to cable was a side show compared to reach achieved with major TV networks ad buys. In other words, cable was not worth advertisers’ time.  No doubt agency folks believed this but it was also commonly understood that planning/ buying cable TV was more labor intensive with no commensurate fee increase to cover these added labor costs. During this decade, the push against cable by agencies was driven by self-interest. This sparked a battle of attribution where brands would want to see cable TV as part of the ad buy but agencies dragged their feet. Eventually agencies caved.   

The last transformative technologies of this decade was email, Hard to imagine business today without it. Yet even with email, agencies were slow to adopt it. I was at a large agency when AT&T wanted to start using email to communicate with us. I asked the management supervisor for a PC (no laptops yet) so I could communicate with the client. My boss, looked rather dismayed and took me into his office and closed the door. “You worked so hard to get to account supervisor [I started in print traffic], if you get a PC people will think you are a secretary.” I knew he was looking out for me and I also knew that there was some truth to his point. Account people were the brain trust of the organization – high hourly rates. Word processing was a low-level task with low fees. Despite the potential internal backlash, I insisted. It turned out that I was the only person on the AT&T account team who had a direct line into the client for quite a while.

2000s – So websites are not just digital collateral?

At first, agencies did not focus on creating websites. After all, the pervasive thinking about websites at agencies in those days is they fell into “Digital Collateral” category. This was actually not an unreasonable position as online commerce was limited. Amazon was barely five years old selling books and SSL’s would not become standardized until about 2005.

Simply, agencies did not take online activities too seriously. Just after the dot com bust, agencies felt vindicated in their skepticism of digital because everyone’s worst fears came to life. As agencies were strutting smugly, they missed the big picture. They didn’t pick up on the fact that dot com bust opened up everyone’s eyes to online commerce’s potential to vastly expand the potential addressable market. No brand was going to pass that up.

Whether agencies liked it or not, this is when website creation became a core business function, and unsurprisingly agencies struggled. Those agencies with no technical competencies, hired outside folks to create these sites. A few agencies, notably Organic, did embrace digital transformation during these early years and were able to charge hefty six figures for website creation.

As the decade proceeded, new platforms were launched like WordPress which made website creation much much easier and much much cheaper. Unfortunately, too often agencies took advantage of brand’s ignorance about technology and still charged brand hefty six figures to build these templated sites. Soon enough brands realized that agencies were charging so much for a website they hadn’t even built themselves. They felt exploited. This set the stage for brands mistrust of agencies and their application of technology.  

2010’s – Push a few buttons and billions of impressions start falling from the sky.

This was the decade of programmatic media. It sounded magical. Lots of impressions can be bought cheaply, targeting “exactly” the people the brand wanted to target.

Pulling this off required a sophisticated adtech stack which agencies definitely did not have. Clients were not interested in the details. They just wanted the agency to use all this cool (over-hyped) technology. Agencies scrambled and did deals that relied on adtech firms to execute the buys.

This meant tapping into a new ecosystem; DSPs who tapped in ad inventory from publishers using another tech platform called SSPs. In the middle were exchanges who acted much like the stock market enabling all the buying and selling.

The kicker was that all this tech came with a tech tax that was often invisible to brands. There were many hands taking small percentages of the media buys; DSPs, data providers, SSPs, exchanges. Often, when adding up all the tax, about 20 – 25% of the media buy never even made it into active media. Fraud, privacy and brand safety “management” tech sliced off another 15% or more from active media dollars. By the time agencies made their media commissions, it was common for a media budget to be reduced by 50% or more. Not an easy story to explain to clients.

In addition to the issues around programmatic media buying, this was also the decade of everything SaaS – social publishing, sentiment analysis, influencer marketing, blog creation. Clients were being pitched directly by adtech firms and brands were signing up. No agency needed.

The pressure to evolve agency’s business model only grew more urgent. Between programmatic media with all its trust issues that came at a steep price to brands and increasing brand confidence in executing marketing tasks themselves, agencies fee-based business model was on life support. Brands were evolving through the pains of digital transformation but agencies did not evolve in ways that clients needed.

Toward the second half of the decade, agencies finally “got it” and the larger agencies just started buying up digital transformation/ adtech companies as fast as they could. This acquisition spree though came at a high price, since valuations was at their highest during this time. Worse, the new platforms were just awkwardly “bolted” onto existing agency systems and functions. Lots of round tech pegs trying to figure out how to fit into square agency process holes.

All these process issues dwarfed the larger issue that programmatic was viewed with increasing suspicion by brands. Rumblings from clients about brand safety, transparency and outcomes in digital media became a rolling drumbeat – getting louder with every passing day. Agencies had invested heavily in the very technologies that were becoming stranded assets, as nextgen solutions disrupted earlier tech. ROI for these acquisitions seemed to recede further and further into the background. This expensive miscalculation is what set-up the struggles of larger agencies in the next decade.  

2020s – Data here. Data there. Too much data everywhere.

This is when agencies finally, finally realized their tech blind spot and they raced to play catchup.

They learned to manage profile (a.k.a. surveillance) targeting data.

They learned how to buy programmatic media more efficiently despite ongoing issues of fraud, brand safety and transparency.  

Agencies started o develop deep data credentials and we see an explosion of terms from “zero party data” to “third party data.” Data became the foundation upon which almost all strategic and tactical decisions were based covering a wide range of critical functions:

  1. Data management
  2. Social media management and analytics
  3. Personalization and profile targeting
  4. Media buying and ROI optimization  
  5. Creative optimization
  6. Data management platforms and Customer Management Platforms

This was a huge undertaking and agencies were beginning to get their sea legs wrestling with all this fragmented data. The road to revenue redemption seemed clear.

That is, until AI.

2024 – AI tsunami drowns out the human spirit.

As AI proliferated in marketing, agencies first response was, predictably, resistant. Loud cries arose that AI was trained on the back of creative works. True enough but the objections didn’t stop there. Agencies primary objection to AI creative was that it was inferior to human-created work. Also true but that misses the real point. Yet again agencies were too busy fighting off perceived fee losses and missed the key point that clients started to play with AI without agencies. AI offers advertisers a cheaper alternative for tasks traditionally handled by agencies, including creative work, content production, and business operations.

Clients are going to utilize AI whether agencies like it or not.

The Future.

Agencies never missed an opportunity to miss a tech opportunity. It is sobering to consider that about 50% of advertisers will be doing an agency review this year. That is a lot of stress and a lot of gyrations to stem a potential loss. One can’t help but wonder if one of the drivers of all these agency reviews is trying to find agency partners who can address their whole business; marketing and marketing technology within a business model centered on total sales and business outcomes.

Brands are right to be skeptical of agencies’ ability to drive business. While agencies might be media agencies – they are not in the media business. Other agencies might be creative agencies but they are also not in the creative business. Agencies are in the business to build businesses – in its entirety – not just the marketing aspect of a company.

Agencies consistent rejection of tech over the decades, had serious consequences causing clients to stop trusting agencies to know how to activate tech to drive their business.  SaaS models, from analytics to social, meant self-service and clients were comfortable buying this tech. No agency labor fees required.  To add insult to injury, clients asked agencies to learn to use the technology brands acquired within their marketing tech stack. Again, agencies’ fees took another hit – learning new tech with no increase in fees.

Agencies lost their swagger. We see that today.

What about tomorrow? This time around this story can have a happy ending.

AI Agents – The Business Model That Can Go The Distance For Agencies.

Technology can be the game changer on the road to win new business because clients want to trust an agency’s tech credentials. Creative, strategy and media management will always be important but they are not the star of the show anymore. Technology takes center stage.

I dont mean any technology because not any technology is robust enough to be able to change the agency business model. The only technology that fits the bill is AI agents – not to be confused with AI tools.

AI agents in particular, is not just a shiny new object but a rich new category opportunity as dramatic as Internet was in its day. This is the chance for agencies to reclaim the mantle of tech/ trust leadership.

Before we proceed, we need to understand what AI agents are especially in the context of AI as a category. AI is as dissimilar to AI agents as bicycles are to motorcycles. Both are transportation vehicles with two wheels but that is where the similarity ends.  

AI tools can do a very specific tasks; research or writing an email about a budget. More like an assistant, AI tools require manual inputs and ongoing input from the user for each step. They can’t proactively work towards a broader goal on their own.

AI Agents can complete multi-step processes for instance, create a media plan and then do the media buys on the plan. AI agents can operate independently after an initial prompt or goal is set. They can evaluate objectives, break down tasks into sub-tasks, develop their own workflows, make decisions, and take actions without constant human supervision. AI agents are more like a department capable to executing a full function. The innovation breakthrough with AI agents is that AI agents can maintain context throughout a process, remembering previous steps and information to make more informed decisions.

Now we can imagine a new agency business model. Instead of focusing on a fee structure that is hourly based, and obsolete in the age of AI, it’s  time to rebuild the agency model from the ground up with an emphasis on agility, measurability, and efficiency.

It’s a rare moment to shift to a productized service model that allows clients to buy services plus tech bundles together.

In this vision, there are three competency areas making up a core team:

Progress planners. This is where the strategy and campaign planning responsibilities are. Within this is team are account planners; creative and technical campaign planning; experiential designers — translating experiential design into workable campaigns — and social engagement planners.

Performance planners. This new expertise will plan the performance of all marketing programs with a new set of tools and competencies; media planning (all platforms), campaign proforma modeling, fraud management and customer topic journey mapping.

Platform planners. This is where agencies connect the dots between platforms, programs and business results. This team owns predictive modeling, audience data, privacy compliance, transparency and platform auditing. This is also where clients get support with their technology challenges such as data integration.

This new structure allows agencies to operationalize this revenue-rich, productized service vision through the power of AI agents. With AI agents, agencies can again become guardians of clients’ budgets protecting campaigns against fraud and inefficiencies by applying AI agents to manage media buying, content syndication, social, etc. For agencies that invested heavily in data, AI agents will help clients maximize their data assets to the maximum.

For almost 40 years, agencies have been playing tech catchup. With AI agents, agencies can get ahead of the curve because the development of AI agents is just beginning. Yes – developing AI agents is a heavy tech lift but they offer agencies the chance to integrate AI agents into their core business, giving agencies an opening to reclaim their role as trusted advisors.

A chance like this may not present itself to agencies for another 40 years. “Opportunity is missed by most people because it is dressed in overalls and looks like work,” Thomas A. Edison.

Time to get to work!   

————————————————————————————————-

Author’s note: I started the company, The Trust Web, in 2012 to create a data and analytics platform that uses topics to target audiences because topics are the most reliable indication of intent.  The technology is also meant to offer a trust adtech stack. I explain more here. Beyond the Click: A Trust-Based Model for AdTech –  https://trustwebtimes.com/beyond-the-click-a-trust-based-model-for-adtech/ .

Share: