Close Window

In Agency Holding Companies’ Drive To Stay Relevant, Will AI Help?

The largest global agency holding companies – Dentsu Aegis, IPG, Omnicom, Publicis, and WPP – are competing more fiercely than ever in the face of multiple challenges. Three of the biggest challenges include downward pricing pressure for services, clients investing in building in-house agencies to replace external partners (i.e. Unilever), and increasing competition from systems integrators and consultancies building and/or acquiring digital practices (i.e. Accenture, IBM iX, Deloitte Digital).

Because of their complex web of multiple operating entities with different operating models and cultures spread across geographies, the agency holding companies are particularly challenged in meeting client needs at the speed and scale they need ( for more detail on this issue, see my recent post on global agencies’ consistency problem). On paper, and certainly in their investor and analyst presentations, agency networks have all the resources and expertise they need for any project at their disposal. In reality, competition between operating entities, disjointed processes, and cultural differences combine to frequently slow down or prevent these entities from collaborating effectively.

It is against this backdrop that Arthur Sadoun, Publicis’ new CEO, dominated headlines of this year’s annual advertising awards fest in Cannes. Sadden announced that Publicis would not be investing in attending any awards shows or major industry events for the next year, in order to shift investments to build an intelligent assistant platform called Marcel. Sadoun noted that this would mean avoiding any investments in promoting the group until the right people and money were in the right place.

(Theoretical) Marcel debuts at Cannes . . .

Marcel, named after Publicis founder Marcel Bluestein-Blanchet, is described as a kind of personal assistant that will be able to help the holding company to better serve clients. As Chip Register, co-CEO of Publicis.Sapient and the architect of Marcel described to AdAge, “(it is) a transformation to go from a group to a platform,” he said. “A bunch of organizations to a flat leveling of capability that can be compiled in new creative ways that can solve new and creative problems for our clients. What Marcel does is create the mechanism for that happen (sic).”

According to the Publicis Groupe press release, “Marcel draws on the predictive nature of artificial intelligence (AI) to identify opportunities, anticipate clients’ needs, connect people, and unleash creativity while harnessing the power of the Groupe’s data spine to drive business solutions.” The press release reports that the features for Marcel are “inspired” by a global survey of Publicis employees.

That’s a ton of pressure to put on a platform that doesn’t exist yet using tools and technologies at various degrees of maturity (my colleague Connie Moore explains the hype and reality of AI well here), not to mention pressure on the employees and partners who now have to build the thing.

. . . but is he late to the party?

What press reports about Marcel and Publicis’ plan to sit out industry events for a year missed (as an aside, the reports were largely concentrated in publications focused on media and advertising like Adweek and Adage: tech-focused publications like TechCrunch, Recode, and others didn’t seem to register the event as newsworthy) was that Publicis’ competitors have been turning to both AI and other solutions to overcome their structural and historical limitations.

For example, Publicis former potential merger partner Omnicom somewhat more quietly told Adweek that Annalect, one of its operating companies, was launching an internal chatbot capable of analyzing disparate sources of data and providing results to answer employee questions. Annalect was demo’ed at Cannes, though oddly received scant other coverage despite being a tangible example of what Marcel is in theory going to do.

In comparison, look at WPP. In 2015, WPP announced that it would be taking steps to better incentivize its various agencies to work together. Calling this initiative “horizontality,” WPP appointed 46 global client leaders for the company’s largest clients, as well as experimenting with country managers in some markets. Rather than turning to technology as an answer to a problem, WPP is, rightly, is creating new organizational structures and processes to potentially solve a problem.

AI, like all tech, is a tool, not a solution

Arthur Sadoun is probably right: Cannes and other industry events may indeed be a massive money pit for Publicis funds that could earn a better return on investment elsewhere. At the same time, his recent announcements are problematic for several reasons:

  • Marcel is presented as a magic wand that will solve some of Publicis’ most intractable problems, which sets potentially unrealistic expectations for all the various Publicis operating entities that will supposedly benefit from it;
  • Sadoun and his team assume that a single platform can be built to be ready for prime-time demo quality in a year. Yet less ambitious technology projects for smaller and less complex organizations than Publicis regularly take multiple times that amount;
  • The decision seems to have been informed by incomplete stakeholder research: a global internal survey of Publicis employees apparently was done, but there is no sense that there was any similar research on what customers, partners, and suppliers (i.e. the tech vendors who will be asked for help to build the platform) want or even think is possible;
  • The plan assumes that the value Marcel will deliver will more than compensate for any potential lost business that Publicis may see if it does not participate in industry events. For all their extravagance and pomp, these events are some of the few opportunities that clients get to interact with agency partners in person, as humans. In other words, in the zeal to become more digital, Publicis may be weakening its ties to the physical world, which still very much matter for customer satisfaction.

My take is that Sadoun’s goal at Cannes this year was more of a personal one, to make his mark and show how he has a grasp of the importance of technology competency to the company’s ability to survive and thrive in the digital era. However, if this was the case, it was remarkably short-sighted on his part. Setting such lofty goals for his team and discussing technology as a cause, rather than an enabler, of organizational change, illustrates a lack of understanding of the role of technology in complex organizations. The most innovative AI platform in the world won’t help the company if it continues to incentivize its different operating entities to compete, rather than collaborate, or if it fails to hire people with the right expertise for roles where they are set up for success.

Which brings me to what this all means for customers of agency holding companies. The reality is that while these announcements all made a big splash at Cannes and in the industry, it’s more important to look to your potential digital partners – be they operating entities of Publicis, any other holding company, or something else – for how well they understand your business, your goals, and the challenges that will inevitably occur along the way to meeting those goals. They should also be willing to challenge your assumptions and correct your expectations of how long aspects of a project should take. Be wary of any partner that presents AI or any other technology tool as an answer to your problems. And make sure you choose a partner that demonstrates they will conduct the necessary due diligence in understanding all the stakeholders that matter in your organization before embarking on an engagement (we’ve published a stakeholder analysis report for you to reference, which you can download for free with registration here).

AI is cool, exciting, and in some cases really helping organizations to become more effective in creating great customer experiences, but it isn’t a panacea for every problem: not for agency holding companies or anyone else.



, , , ,

Meet us at: