MediaMedia BuyingThe Practical Guide to Programmatic Planning in China: Part 3

The Practical Guide to Programmatic Planning in China: Part 3

How to evaluate demand side platforms for a programmatic strategy in China.

There are three integral parts to a complete programmatic implementation: demand side platforms (DSP), data management platforms (DMP) and the trading desk. Or in layman terms the buying platform, the data platform, and operations team.

As I have emphasized in my previous articles on this subject, it is important to promote a conflict free technical stack by using different companies across the three key disciplines. You can read about programmatic conflicts of interest in China here, and the finer points to planning a programmatic strategy for China, here.

In today’s piece, I will focus on how to setup the strategy and evaluate a buying platform (DSP). First let’s compare the typical programmatic setup between China and the West:


Setting Up the Programmatic Stack

In the West, private inventory and biddable exchange inventory are bought together within a single DSP. This way the advertiser can better control frequency and creative rotation between the two types of inventory. Even with the overall prevalence of programmatic in the West, private inventory is still going to take up at least half of a client’s digital budget. Private inventory can be managed via a unique deal ID managed under the same DSP that buys biddable inventory. This consolidated approach gives trading desk operations teams an easy to manage interface all through a single DSP. All the while the DSP connects to both first and third-party data sources and integrates tags across multiple DMP’s for targeting. From both an operations and conceptual perspective this way is the most efficient way, for both the advertiser and agency to manage everything under a single platform.

In China however, I recommend advertisers set it up differently.The primary difference being: separating the DSP that buys private inventory from the DSP that buys exchange inventory.

The reason for this is very much related to conflict of interest issues within China’s programmatic ecosystem. Most of the western DSPs make their margins purely based on service fees, so it matters little to them what type of inventory the advertiser buys (private or exchange). But this is precisely what’s different about DSPs in China, as most of the local DSPs evolved out of traditional ad networks, they still insist on making a mark-up on exchange inventory, while they can only make a service fee on private inventory because the price is negotiated directly between the publisher and advertiser.

Hence, if a DSP makes a higher profit on one type of inventory versus another, then the innate tendency for the DSP is to shift the advertiser’s budget from private to remnant inventory.

An example of this is many local DSPs like to tell the story of programmatic direct buying (PDB) and real-time bidding (RTB), using a combination of private and remnant inventory to increase overall media buying efficiency. But it is in the DSP’s best interest to continuously demonstrate to the advertisers that remnant is better than private inventory, because they make a mark-up on remnant. But is the efficiency on remnant inventory really worth the brand safety and media quality problems? I certainly have my reservations, so my recommendation to advertisers is to absolutely refrain from using a single DSP that does both PDB and RTB, and instead use two separate buying platforms:

1. The trading desk platform – this platform only buys you private inventory – it is in the advertiser’s best interest if the supplier is not even connected to the open exchanges. In effect, taking away their innate bias to shift towards exchange inventory.

2. The open exchange DSP – this DSP only buys open exchange inventory, but has to be integrated with the trading desk platform for it to centrally control frequency and creative rotation. Since the majority of the local DSPs make a mark-up on exchange inventory and run managed services, then I recommend that we treat them all like regular ad networks.

Trading Desk Platform Evaluation

The trading platform will be managing private inventory bought in a fixed cost per thousand (CPM), and a pre-negotiated fixed pass-back ratio with the publisher. In this case, the platform is responsible to pick out the most on-target/relevant impressions that the publishers sends, then return the non-relevant impressions. So when evaluating this supplier, the only metric that matters, is the “cookie recognition rate”.

The higher the rate, the more the platform can identify the same cookie between the publisher and DMP, and only after identification can a pass-back decision be made. This rate is highly dependent on cookie mapping volumes between the platform with both publishers and DMPs.


Open Exchange DSP Evaluation

Since we’re treating the DSP that buys open exchange like an ad network, then we should also only focus on one metric: CPM. Because all the DSPs are connected to the same exchanges, so there isn’t any difference between inventories. During the pre-buy, evaluate them purely based on the CPM, and post-buy focus on efficiency metrics such as cost per click (CPC) / cost per action (CPA).

The next and final article of this series I will be covering how to evaluate the right data provider (DMP) for your programmatic stack.

To learn more about programmatic in China, join Charlie at ClickZ Live Shanghai on September 9&10.

*Homepage image via Shutterstock


US Mobile Streaming Behavior

Whitepaper | Mobile US Mobile Streaming Behavior


US Mobile Streaming Behavior

Streaming has become a staple of US media-viewing habits. Streaming video, however, still comes with a variety of pesky frustrations that viewers are ...

View resource
Winning the Data Game: Digital Analytics Tactics for Media Groups

Whitepaper | Actionable Analysis Winning the Data Game: Digital Analytics Tactics for Media Groups


Winning the Data Game: Digital Analytics Tactics f...

Data is the lifeblood of so many companies today. You need more of it, all of which at higher quality, and all the meanwhile being compliant with data...

View resource
Giving a Voice to Your Brand

Whitepaper | AI & Automation Giving a Voice to Your Brand


Giving a Voice to Your Brand

Voice commerce, or conversational commerce, is a major disruptive opportunity in US retail. It opens up the possibility of building deeper relationshi...

View resource
Mobile Messaging Masters

Whitepaper | Mobile Mobile Messaging Masters


Mobile Messaging Masters

Every year the mobile market continues to grow and develop. Cyber Monday 2018 saw $2.2 billion in sales stemming just from smartphones in the United S...

View resource