A little more than a year ago, I wrote that the Big Five would become four. A few months later, I tried to imagine how the global shock caused by the merger of Omnicom and Interpublic would be felt in the markets I know best, from Sarajevo through Belgrade to Zagreb. Both texts contained many assumptions because, at the time, it was mostly possible only to speculate about what life in the new company would look like once the presentations about the merger ended and the actual integration of people, clients, technology, budgets and different agency cultures began.
Today, we no longer have to speculate quite as much.
Omnicom completed its acquisition of IPG on November 26, 2025, and very quickly showed that one of the biggest mergers in the industry’s recent history would not end with two enormous collections of agencies simply being placed under the same roof. The elimination of overlaps began. DDB, FCB and MullenLowe, names that marked different eras of advertising, were retired as standalone global networks. FCB was integrated into BBDO, while DDB and MullenLowe entered the TBWA structure. In the new architecture of Omnicom Advertising, BBDO, McCann and TBWA became the three key global creative networks.
Jack Morton took a different path. It was not merged with another Omnicom agency, but sold to Impact XM, with the backing of investment company Riverside. The new company continued operating under the name Jack Morton precisely because of its global recognition. It is an interesting detail because it shows that leaving one system does not necessarily mean that value disappears. Sometimes a company simply no longer fits the architecture of one owner, while another combination opens new room for growth.
At the same time comes the much more painful part of every major restructuring. Omnicom had approximately 120,000 employees at the end of 2025, and CFO Phil Angelastro said in early September that around 105,000 employees was a good estimate of the number with which the company could end 2026. The difference is significant, but those 15,000 should not automatically be read as 15,000 conventional layoffs. The figure includes the elimination of duplicate corporate and regional functions, outsourcing, offshoring, as well as people leaving Omnicom together with companies the group is selling.
That, of course, does not make the change any less real for someone whose job has disappeared. Words such as synergy, integration and efficiency sound different in an investor presentation than they do to someone leaving the company. But it would be equally wrong to reduce the entire transformation to the number of layoffs.
Because Omnicom is growing at the same time.
In the second quarter of 2026, its core operations recorded organic revenue growth of 6.1 percent. The cost synergy target following the merger was doubled from the initial $750 million to $1.5 billion, of which $900 million is expected during this year. At the same time, management is talking about clients consolidating an increasing share of their business within the group.
This creates an unusual picture of a company that is becoming larger as a business while, as an organization, trying to become smaller and simpler.
When viewed only through Omnicom, this can seem contradictory. When we look at what is happening around it, it begins to look like part of a much broader change.
From a collection of agencies to a connected system
Perhaps the biggest mistake is to view Omnicom in isolation. Almost all large communications systems today are trying to solve a similar problem: how to turn an organization made up of dozens of agencies, hundreds of offices, different P&Ls, technologies and internal interests into something that is simple for the client to use.
Publicis began that transition much earlier. Since 2016, its Power of One model has been trying to open the entire group to the client through one lead partner, instead of forcing the client to find their own way between creative, media, data and technology silos. Epsilon, Sapient, Marcel and CoreAI have over time become shared infrastructure connecting different areas of expertise, while individual agency brands can retain their own identity.
And the results show why that model matters to the rest of the industry. Publicis ended 2025 with organic growth of 5.6 percent and an operating margin of 18.2 percent, and after a strong first half of this year raised its organic growth forecast for 2026 to 4.5 to 5 percent. What looked like a radical reorganization about a decade ago now looks like a model competitors are studying carefully.
WPP is perhaps an even more striking example. The company that for decades was one of the symbols of the holding structure is now simplifying that very structure. The Elevate28 strategy envisages a shift toward a more integrated company with four major operating units: Creative, Media, Production and Enterprise Solutions, organized across four regions and connected by the WPP Open platform. Management itself cites excessive organizational complexity as one of the reasons for its earlier problems.
Havas reached a similar goal by a different route. The Village model has for years been bringing creative, media and other disciplines closer together, while Converged.AI is trying to do the same at the technological level, connecting data, AI, media, production and creativity. At the same time, Havas continues to acquire specialized independent agencies. During the first half of 2026 alone, the group acquired majority stakes in a number of such companies.
That is an important detail. Integration clearly does not necessarily mean that everyone has to become the same. A system can become more connected while new knowledge, new cultures and specialist capabilities enter it at the same time.
Dentsu is not standing still either. During 2026, it reduced layers of global management, while in the EMEA region it moved to a simpler cluster model whose stated goals are greater connectivity, less complexity, faster decision-making and more room for local innovation.
Different companies, different problems and different speeds of transformation, but the direction is quite clear. A holding company that once proved its value by the number of agencies it owned is now trying to prove how well it can connect them.
The agency is no longer an island
That is why the disappearance of names such as DDB is so emotional for the industry. Advertising has a special relationship with its agencies. We often talk about them almost like football clubs. They have their great eras, heroes, rivalries, creative schools, a specific language and humor. People do not just say where they work. They often feel that they belong to something.
Excel logic therefore cannot explain everything that is lost when a name that has existed for decades disappears. But nostalgia in itself is not a business model either. The problems clients bring today can no longer be neatly divided according to old agency disciplines. They generally do not care where creative strategy ends and commerce begins, which P&L influencer marketing sits in, or which network production belongs to. They want someone to understand their business problem and assemble the best people who can solve it. That is precisely where the potential of all these changes lies.
If a shared platform can eliminate hours lost to internal administration, if the same data no longer has to be bought and analyzed five times in five different agencies, and if a strategist from Sarajevo can work without major organizational obstacles with a creative from Belgrade, a production team from Zagreb, data experts in London and a commerce expert in a fifth market, integration is not just a way for the finance director to reduce costs. It can become a way for good people to spend less time on the structure of the system and more on the work for which the system exists. Of course, that is the better of two possible versions of the future.
If technology, data, administration, tools and knowledge are centralized while different creative cultures are preserved, a large system can give people more opportunities than any individual office ever could. If, however, the way of thinking is centralized along with administration, if efficiency becomes more important than diversity, the result could be a perfectly organized company in which everyone works in a similar way.
Perhaps this is precisely where it will be decided who will make the best use of the next decade of transformation. Not between the large and the small, but between systems that know what should be connected and those that will connect too much.
AI is not the beginning of the change, but it is strongly accelerating it
It is easy to attribute the entire transformation to artificial intelligence. AI really is increasing pressure on organizational structures, particularly in production, analytics, adaptations, media operations and jobs in which value was for a long time measured by a large number of repetitive hours.
But the transformation of holding companies began long before generative AI. AI has simply made the existing complexity even harder to justify.
When Publicis talks about CoreAI, WPP about WPP Open, Havas about Converged.AI, and Omnicom about the Omni platform and Acxiom, they are all, in different ways, trying to do something similar: create a shared intelligent layer through which people across the system can access the data, technology, tools and expertise of the entire group. This is also changing the very idea of the value of a global network.
It was once enough to have an office almost everywhere. Today, it is necessary to prove that all those offices, people and data can truly work as one connected organization when needed, while remaining different enough when it is precisely that difference that produces better work.
Sarajevo, Belgrade and Zagreb: when global change reaches our market
In the previous text, I tried to imagine what the Omnicom-IPG merger might look like when it comes down from global organizational charts to Sarajevo, Belgrade and Zagreb. These are markets I know much better than the major global centers, and precisely for that reason their response to this change is more interesting to me than yet another organizational diagram from New York or London.
In our markets, almost no one has the luxury of deep specialization. Teams are smaller, people often work across broader roles, the boundaries between disciplines are more porous, and solutions have to be found faster and with fewer resources. What could long be seen as a limitation of a small market could become a very useful kind of experience in the new model.
Because there is not much business logic in Sarajevo, Belgrade and Zagreb each duplicating every capability that exists in the other two cities. Technology, data, certain production and specialist functions, and even parts of management can increasingly be shared. Some positions we know today therefore probably will not exist in the same form tomorrow.
But that does not automatically mean that a smaller city or smaller market will become less important. If networks really begin to organize themselves around capabilities rather than geography alone, Sarajevo no longer has to be only an office servicing Bosnia and Herzegovina, just as Belgrade or Zagreb do not have to prove their value exclusively through the revenue they can generate in the domestic market. Each of those cities can develop expertise that will be useful to a much broader system.
One team can be particularly good at production, another at media or data, a third at creative work, strategy, events or content. People are then connected not because the organizational chart places them in the same region, but because that particular combination is the best one for the work that needs to be done. That changes the logic of a small market.
Sarajevo does not need to become Belgrade. Belgrade does not need to be Zagreb. Nor does Zagreb have to copy some larger European center. The more important question is what each of them can give a broader system that is not easy to find elsewhere. A traditional network often viewed a local office through the size of the market, local revenue and the number of clients it could service in that country. A more connected system has the opportunity to value it according to how good it is at something that others also need.
For talent from Sarajevo, this could mean that international work no longer has to wait for a major global pitch to happen to come to Bosnia and Herzegovina. A creative in Belgrade does not have to physically relocate to a major Western European center to regularly be part of an international team. A specialist from Zagreb does not have to be limited by the size of a Croatian budget if their knowledge is relevant to clients in other markets.
And here we come to another part of the story that is easy to overlook. In the new architecture, it will not only be the local offices of large networks that matter. Independent local agencies may become more important precisely because large systems are becoming more selective about what they truly need to own and what they can find more effectively outside their own organization.
Sarajevo, Belgrade and Zagreb have independent agencies that for years have had to build their businesses without global infrastructure behind them. They have learned to make decisions quickly, connect disciplines, work with more compact teams and read local culture, consumers and business context in a way that is difficult to produce from a headquarters thousands of kilometers away.
Such agencies do not have to be opponents of large systems. They can become a very valuable part of them without needing to be owned by them. One can be the best specialist partner for a particular discipline. Another a regional production or creative center. Another a partner for a market where it no longer makes economic sense for a network to maintain its entire infrastructure. And another a place where a global strategy finally stops looking like a global strategy and begins to make sense to someone who lives in Sarajevo, Belgrade or Zagreb.
Havas’s acquisitions of specialist independent agencies show that large systems are already actively looking for this kind of knowledge. Omnicom’s sale of individual companies shows the other side of the same logic: it is no longer necessarily desirable to own every capability simply because you once bought it. The boundary between the network and the independent could therefore become much more interesting than it is today. Both will have to answer the same question: what are we so good at that the system wants to include us?
Not just good enough for our own market, but relevant enough for someone assembling an international team to say: for this part of the work, I want precisely the people from Sarajevo, Belgrade or Zagreb. If that happens, the competition will no longer be only for local market share. It will also be for a place in a much larger value chain.
After the cuts
It would be easy to end this story with an elegy for DDB, FCB, MullenLowe and the thousands of jobs that have disappeared or will disappear in the consolidation process. And part of that sadness is completely justified. But that would still be only half the story.
For most of the past forty years, the advertising industry grew by adding. New agencies, new disciplines, new specializations, new digital units, new regional levels and new boxes on the organizational chart. Almost every new business problem produced another organizational layer. Now the industry is trying to learn how to subtract as well.
Publicis started earlier and showed that a more integrated model can grow at the same time. Havas is trying to combine proximity between people and the diversity of agencies with shared technological infrastructure. WPP is dismantling the complexity it built over decades in order to once again find a simpler path to the client. Dentsu is reducing management layers and giving more room to connectivity and local initiative. Omnicom is doing all of this on an accelerated timeline because, through its acquisition of IPG, it gained the largest possible laboratory for this new phase of the industry.
That is why the figure of approximately 105,000 employees, if Angelastro’s estimate is realized by the end of the year, will not in itself tell us whether Omnicom’s transformation has succeeded. What will matter much more is what those 105,000 people can do together that approximately 120,000 could not. If the answer is only “the same, but cheaper,” then the industry really is in trouble.
If the answer is that a strategist in Sarajevo can use the data and technology of the entire network, a creative in Belgrade can regularly work on a global brief, a specialist team from Zagreb can become part of international delivery, and an excellent independent agency can be included because it possesses knowledge the large system lacks, then perhaps we are not looking at the end of the agency era. Perhaps we are looking only at the end of one of its organizational forms.
When we looked at Omnicom and IPG a year ago, the question was how large a holding company could become. Today, it seems to me that the more interesting question is how simple it can become without losing precisely the human complexity from which culture, ideas and creativity emerge. If Omnicom, Publicis, WPP, Havas and Dentsu find that balance, the biggest change may not be that several famous names disappear from the front doors.
It will be that the network stops being merely a closed system of companies owned by someone and becomes something more useful: a system that knows how to connect the best people, offices, technology and local partners when they are needed, regardless of whose logo is on the door.
And in such a system, Sarajevo, Belgrade and Zagreb could become more interesting to the global industry than the size of any of those three markets suggests today.
