In this article
The honest answer to this question is: the right agency model is the one that matches the level at which your media is actually planned and bought. For locally traded media such as television, radio and out-of-home, the strongest local agency is usually the right starting point - selecting the best agency per market, an approach often called cherry picking. For media genuinely planned and operated at platform level, central arrangements are not a compromise, they are the appropriate buying level. Everything beyond that is a conscious trade-off between local quality and central benefits.
If you are building a multi-market media structure for the first time, or reviewing one that has grown by default rather than design, that distinction matters more than almost any other. The model you choose shapes your governance, your costs and your negotiating position for years.
What the ideal model actually looks like
Start with the buying level, not the organisational preference. An international media agency arrangement that does not match the level at which media decisions are actually made adds overhead without adding performance.
For most advertisers operating in multiple markets, the media mix contains both types of activity. Linear television in Germany, Japan or Brazil is negotiated locally against local rate cards and local broadcaster relationships. No network affiliation changes how a spot is traded in that market. For these channels, a strong local independent or a locally led network agency almost always produces better buying outcomes than a global mandate managed from the centre.
Digital platforms work differently. Google, Meta, programmatic inventory and connected television increasingly come with joint business plans and enterprise-level agreements that are negotiated centrally. Here, a genuinely central setup may be the buying level itself. Not a concession to organisational convenience, but the logical structure for how the medium works. The opacity of what those platform arrangements actually contain, and who benefits from them, is a separate question worth asking directly. The leakage that occurs between global deals and local execution is covered in depth in the article on where multi-market value leaks between global deals and local buying.
Five legitimate reasons to consolidate beyond the buying level
When organisations choose a network agency of record (AOR) across markets, it is rarely because they believe it will improve local buying on every channel. The real drivers are usually one or more of the following. Each is legitimate.
Reporting and governance. One contract, one KPI framework, one counterpart for the global marketing director. For a large, complex organisation this has genuine value, independent of whether it produces sharper media buying in any individual market. International headquarters often weight this heavily, and they are not wrong to do so.
Economies of scale on the relationship and the deal. One negotiation produces one set of commercial terms, and it enables participation in platform-level corporate arrangements. That value is real, but it is narrower than is sometimes presented. Local television in any market is still negotiated locally, regardless of network affiliation. The scale benefit applies to platform relationships and the negotiation process itself, not to the quality of local inventory access (exceptions apply).
Thin local marketing organisations. Where local teams lack the scale or expertise to manage an agency relationship with rigour, central governance compensates for that capacity gap. Less local decision-making then becomes a structural feature, not a loss. Where strong local marketing organisations exist, the same move frequently dismantles something that worked. Different markets within a single advertiser’s portfolio can sit in different categories here. A model that fits one region may be wrong for another.
Standardisation of strategy across markets. One strategy deployed consistently across many markets is a genuine operational benefit. But the honest question that follows is whether a network AOR is required to achieve it, or whether strong central direction combined with the best local agencies in each market delivers the same consistency. In practice, both approaches can work. The choice depends on the internal capacity to manage a distributed roster.
None of these four reasons makes a network agency locally better at planning and buying media in a specific market. They are legitimate needs, but they are needs of the organisation, not qualities of the agency. Making that distinction is precisely why the choice deserves to be made consciously, rather than by default. The Multi-market media agency selection: one review across markets framework is built around that principle.
The fifth option: governance without consolidation
There is a structural alternative that decouples the governance need from the agency choice. An independent vendor management program can deliver much of what advertisers typically seek from a network AOR - one reporting framework, one governance structure, one point of coordination across markets, and operational involvement where needed - while the roster underneath remains the best agency per market. The agencies plan and buy the media and own the strategy; the independent layer coordinates, reports and verifies. For organisations whose main driver is reporting and governance rather than the deal itself, this combination captures the central benefits without exchanging local quality for them. The verification and performance benefits that come with a vendor management program arrive alongside it.
Regional as the intelligent middle ground
A regional model, one agency covering a cluster of neighbouring markets, is often underused as a structural option. It is not simply a compromise between global and local. It can be the most deliberate choice of the three.
Choosing one agency for a small group of markets such as Southeast Asia or the Nordics means seeing exactly what is exchanged. This agency is somewhat stronger in one market, that one in another. The advertiser decides, with open eyes, that manageability and regional coherence can outweigh the performance difference in individual markets, while retaining the option to split the arrangement later without dismantling a global relationship. Media landscapes across neighbouring markets can vary significantly: the landscape for advertising in Southeast Asia looks quite different from what European markets present, and a regional agency that understands those nuances at cluster level often delivers better outcomes than a global mandate with local desks.
A regional model also allows different structural choices in different parts of the world. An advertiser might run a global AOR for digital platforms, a regional model in Asia-Pacific and local specialists in markets with particular buying complexity. That hybrid is not disorder. It is honest design.
The agency roster question
Consolidation and the roster model are often discussed as if they were the same decision. They are not. A global AOR collapses the roster to a single relationship by definition. A regional model typically involves a small number of relationships. A local model involves one relationship per market, which creates a roster management question of its own.
The agency roster consolidation decision, how many agencies to maintain, across which markets, and at what level of specialisation, deserves its own analysis, separate from the structural model. Similarly, the question of which functions belong inside the organisation versus in an external agency is covered in the in-housing versus agency article.
Making the choice explicit
The model that fits is the one that a well-informed procurement or marketing leadership team would choose again tomorrow, knowing what is exchanged. Global consolidation made by default because a previous global AOR exists, or because a network pitched well, is a different decision from global consolidation made because the organisation genuinely values governance and standardisation above local buying quality in most markets.
Write down the four reasons listed above. Assess honestly which apply to your organisation, in which markets and at which intensity. Where the organisational case is strong, consolidation is the right answer. Where it is weak, the best local independent or locally led agency in each market is very likely to outperform a network mandate on the metrics that matter most.
Frequently asked questions
When is a global media agency the right choice? When the organisation’s genuine needs, governance, standardisation, capacity at the centre, outweigh the local quality difference in individual markets, and the advertiser makes that exchange consciously. It is also the appropriate structure for the growing share of media that is genuinely planned and operated at platform level, where central is not a concession but simply the buying level itself.
Who are the big media agency networks? The holding groups historically referenced as the Big Six were WPP, Publicis Groupe, Omnicom, Interpublic Group (IPG), Dentsu and Havas. Following Omnicom’s acquisition of IPG, the market effectively counts five major groups. Each operates multiple network agency brands across markets, and structures continue to change through mergers.
Does a network agency always produce better buying terms than a local agency? Not automatically. For locally traded media such as linear television and radio, buying conditions are determined by local market relationships and local rate cards, regardless of network affiliation. For platform-based digital media, central network arrangements can carry scale benefits, though the precise value of those arrangements is rarely disclosed, which is a transparency question in its own right.
What does a regional media agency model mean in practice? A regional model places one agency in charge of a defined cluster of markets, Southeast Asia, the Nordics, DACH, for example. It offers fewer relationships to manage than a fully local roster while preserving more local proximity than a global AOR. It works particularly well when markets within the cluster share planning logic, buying season and some audience overlap, and when the advertiser wants the option to adjust the structure market by market without unwinding a global contract.
