Mat Baxter reckons agencies face an existential threat from AI because it kills the billable hours model by performing tasks much quicker – reducing workloads by a fifth.

The global boss IPG’s Huge, now back in Australia, told Mi3 a combination of AI-led automation and “margin killing” procurement teams will hurt agency profitability – which is a main reason for Huge lifting consulting’s playbook and ditching time-based billing to instead sell products and templates.

“The business model we had was broken. Every year, we saw it decline consistently,” Baxter said. “For the last eight to 10 years, it's been on the decline because we were selling people on hours. Every year procurement comes to you and says ‘I'll pay you less’ so it's not a model that works.”

AI risks breaking the model, per Baxter – and bosses at holdcos and independents agree that billing on hours and headcount has long created perverse outcomes.

Per Publicis Groupe ANZ CEO Michael Rebelo, traditional remuneration models risk commoditising agency services and can lead agencies and clients to “act against the interests of growth, unintentionally or otherwise”. He claimed by taking a different approach, Publicis is managing to increase fees, provided it hits targets.

This is Flow CEO Jimmy Hyett agreed hourly-based models are old hat and his agency is “transitioning” to value-based pricing. But he said clients must evolve too – and incentivise over-performance or miss growth.

Thinkerbell’s Jaime Morgan said it is already working on multiple fee structures, including experimental models, and that agencies need both service and product models – but some are struggling to evolve.

Atomic 212 CEO Claire Fenner disagreed that AI will ease workloads – and may actually increase hourly rates – but agreed standard remuneration models are outdated.

Media.Monks co-founder Wesley ter Haar thinks the future pay cheque will need to properly value “people and pipeline” – or just be a percentage of total spend.

In response to Baxter’s ever-punchy point of view, Mi3 asked five agency bosses whether the standard remuneration model must evolve, how – and whether AI disruption is to blame. Here are their lightly edited answers:

Michael Rebelo, CEO, Publicis Groupe ANZ

This is a question that has been at the forefront of the industry for some time. The answer is ‘yes’, the traditional FTE/hourly fee-based model is outdated and inhibitive for both clients and for agencies. 

Ultimately, both a client and their agency partners should be in the same business – value creation. The traditional model which views agencies as a cost may be easy to balance from a procurement and operations perspective, but ultimately it drives both sides to act against the interests of growth, unintentionally or otherwise. 

Creativity – which can be unique to each business, brief or consumer challenge – is one of the last bastions of competitive advantage that is generative, yet we insist on handicapping it. I don’t think this is lost on anyone in the industry, but we have been slow to evolve. 

Transformation takes time and at Publicis we have been evolving remuneration models with our clients that put value ahead of costs. An example being that we regularly have large percentages of our remuneration tied to the same growth/business metrics our clients measure themselves by. This could be anywhere from productivity ratios, through to sales or long-term customer retention, with the remainder sitting in a traditional structure. Similarly, by continuously proving value and evolving our operational models, we have seen positive fee structures (not declining) emerge, provided targets are met. 

This evolution will require the entire industry to be onboard … Personally, I welcome the role that AI will bring in sharpening our understanding of the commercial value our human-generated ideas create. 

Claire Fenner, CEO, Atomic 212

I don’t think AI is the reason that agency remuneration models need to adapt. As a media agency we are certainly very aware of AI advancing and its automation capability.  Its not new, the Google Adwords platform has made year on year advances with automation and recently Google's Performance Max product is a game changer in terms of significantly reducing the manual ad buying and optimisation process for that network.

Despite this, the Parkinson's Law paradox is that our workload doesn’t ease. Currently the proliferation of platforms has replaced any automation gains and when this consolidates something more valuable emerges, which we expect will be of higher human value such as analysis and insight, therefore the hourly rates for this work increased in relation to the complexity.

Having said this, we would welcome a shift in the way media agencies are remunerated to better reflect the important value and impact effective media can drive for a business. That value is even greater with the benefit of AI and machine learning helping to provide fast insights into marketing performance against business results.

Jimmy Hyett, CEO, This is Flow

Now more than ever, effectiveness measurement and modelling continue to take the forefront of all our client conversations – and the advancements in data, technology and AI are making answers even more available than they have been in the past. This deeper understanding of effectiveness is already changing the narrative in the agency landscape to more value-based pricing rather than hourly-based models, and something we are currently transitioning to.

AI has been a huge driver of this change, but not because it is replacing people, but because it has the ability to enhance what we do within an agency offering. It’s this combination of human and tech where the powerful advantage comes in for brands, and the justification of why value means more than head hour remuneration.

Continual improvements to tech, like the launch of Chat GPT Enterprise offering further protections for first party data within AI framework, mean there will be even more ability for an agency to create bespoke solutions for clients. And if an agency invests more into people and tech, which means costs double – but delivers a 5x better advantage for clients – surely that shifts the objective for procurement teams away from head hour fees to more outcome based remuneration models. And while they’re at it, they would be wise to include scaling bonus incentives to encourage the pursuit of discovering even more improvements through tech for a win-win for all – that’s just smart business. “ 

Jaime Morgan, Managing Director, South, Thinkerbell

We believe that agencies benefit from having both services and products, with different remuneration models for both ­– which is what we do at Thinkerbell. However, I think many agencies are stuck in a services mindset, and are finding it hard to create the space and discipline needed to create products to sell, thus leaving many with an antiquated revenue model.

Our operating model is already different. It allows us to have an open conversation with our clients about the different remuneration options available and what each delivers. This transparency and partnership with our clients has allowed us to lean into either performance-based, value-based, output-based, risk and reward, and other more experimental models. This is currently working for both our clients and our business. 

Wesley ter Haar, Co-Founder, Media.Monks

Yes, instead of traditional time and material models, it will move into talent and machine, a mixture of people and pipelines, with value mapped to how these combinations solve for business goals. This can be related to speed and scale, cost savings, revenue generation, or it will become a managed model, where broader agency services are delivered as a percentage of media or total marketing spend.