Stage 01
Representation
Premium media relationships, inventory access and outbound partnerships generate revenue and establish commercial entry points on both sides of the border.
Creates the commercial beachhead
How it works
Six steps turn a requirement that currently cannot be met into a campaign that is contracted, billed, localised, served, verified and reconciled across the China border.
Last reviewed
Cross-border China outbound media buying works when one operator holds both sides of the transaction. Mammoth contracts with the Chinese advertiser or agency and with the global media owner separately, then resolves billing, creative localisation, third-party ad serving and reconciliation internally — so neither party needs an entity or capability in the other's market.
The reason this requires an operator rather than a marketplace is that each of the four connective functions is a judgement-bearing operational discipline, not a piece of software. Settlement structures depend on entity type and jurisdiction. Verification depends on which standard the buyer's internal reporting recognises. Localisation depends on the destination market's cultural and compliance context. None of those resolve themselves at the point of a bid request.
The sequence below is the standard path. Individual engagements compress or extend particular steps, but the order is fixed, because each step produces the input the next one requires.
01
The engagement starts with a specific requirement that currently cannot be met: inventory an advertiser wanted, a market an agency could not reach, or a demand pool a media owner cannot access. Identifying the binding constraint first prevents scoping work that does not remove it.
02
Mammoth contracts directly with the relevant counterparty — a media owner on the supply side, or an advertiser, agency or DSP on the demand side. Contracting sits with Mammoth in each party's own market, so neither side requires an entity in the other's jurisdiction.
03
Cross-border billing is structured so each party transacts in a currency and process its finance function already operates. RMB settlement, invoicing requirements and credit exposure are consolidated into Mammoth as a single counterparty rather than distributed across many entities.
04
Creative arriving from Chinese advertisers is adapted for the destination market, covering format, pacing, cultural reference and compliance requirements. Where adaptation is insufficient, market-native assets are produced instead.
05
Campaigns are trafficked and served through the ReachMax third-party ad-serving stack, producing an independent delivery record. This gives the Chinese buyer measurement its reporting stack can consume, and the media owner documentation of what actually ran.
06
Delivery, invoicing and verification records are reconciled and reported into both parties' formats. Because the same operator holds both relationships, discrepancies are resolved internally rather than negotiated between counterparties who cannot see each other's data.
Mammoth develops through three stages: representation, which generates revenue from media relationships; enablement, which embeds that inventory into the systems agencies already use; and infrastructure, which converts accumulated access into a proprietary marketplace and activation layer.
Stage 01
Premium media relationships, inventory access and outbound partnerships generate revenue and establish commercial entry points on both sides of the border.
Creates the commercial beachhead
Stage 02
Inventory is embedded into the buying systems agencies already use, through DSP connectivity, localisation, creative adaptation and cross-border execution.
Embeds into agency workflows
Stage 03
Aggregated supply, a scalable activation layer and a proprietary marketplace convert accumulated access into durable platform leverage.
Owns the rails and the defensibility
The model is staged because each stage funds the next and de-risks it. Representation produces revenue from existing relationships immediately. Enablement uses that revenue to embed into agency workflows. Only then is there enough aggregated supply and demand history to justify building marketplace infrastructure.
Building the marketplace first would require predicting which supply Chinese demand actually converts on. The staged approach observes that behaviour first and builds the rails around what has already proven to move, which is a materially cheaper way to arrive at the same destination.
The timing is favourable in one specific respect: programmatic is becoming the default transaction mode. dentsu forecasts that more than 80% of global digital investment will transact programmatically in 2026, which means the infrastructure stage is being built into a market that is standardising rather than fragmenting.
Source: dentsu, 3 December 2025
Next step
Step one is naming the requirement that currently cannot be met. If you can describe that, we can tell you quickly whether we are able to remove it.