In our last article we drew a quick picture of yield management today, showing that publishers monetization and yield strategies moved from a “hierarchical” system to a multichannel/multiproduct universe. As a result of this move, we see publishers make common mistakes when it comes to finding a coherent pricing strategy for different channels.
In the next three articles, we will lay out 3 ways to rethink and increase yield within this new context:
- Work with real data
- Set up a global monetization management tool
- Adapt your decision-making process
#1: Work with Real Data
It will sound basic, but the first thing to do to work on yield is to know what we’re talking about, which means relying on actual data. Your database doesn’t have to be super complicated, if some dimensions do not matter to you (bidder, device?), you don’t need to spend time collecting, cleaning and analyzing them. Actually, we even advise to start from a few dimensions and metrics that are key to your business, and build a simple but accurate dataset to work with.
In such a complex and fragmented digital advertising ecosystem, there are 3 steps to reaching this “clean” database.
1. Make sure you have the CPMs in your AdServer
We sometimes hear publishers saying that their data within DFP cannot be relied upon, because some line items have “fake” CPMs, or even because CPMs are not even input in the adserver.
However, to have a clean database, the absolute first step is to make sure your adserver CPMs are accurate. If the adserver data is corrupt, publishers are unable to track performance, or measure the impact of one change, and end up losing control over their yield.
Source: Adomik CrossView - An example of corrupt CPM data in a Publisher’s adserver - March 2018
2. Track the right volume
Second step after getting the actual CPMs, you need to be able to look at your actual volume.
This is a two step process:
- First, in a complex stack, you can get your volume data from several sources (your adserver and your SSPs), which can be very confusing. You need to know which source has the most accurate data and feed your database / data lake with this data.
- Second, when you identify discrepancies between several sources, if these discrepancies affect your key metrics, you need to identify the root cause and fix it. In a header bidding setup, the impression reports from the programmatic partner and the publisher ad server impression reports very often differ because of integration / implementation issues - we’ve discussed this topic in a post about the right use of header bidding analytics. In the process, the right analytics tool can be a game changer!
An example of volume discrepancies in a header bidding setup:
Source: Adomik CrossView - Real Data from International Publishers - February/March 2018
3. Implement the right naming conventions
When you sell inventory using multiple monetization partners, it can get very difficult to “map” websites, pages, formats, devices, positions… across all partners. That said, to increase yield, you need to have a clear vision of how website X or format Y is performing across all channels.
If you’re not using a tool that absorbs your custom mapping to show your data in a clean UI, the best way to avoid very long and tedious hours on excel is to implement a simple and unified naming convention in all your partners.
Two good examples that we've seen:
- Device_Site_Type of Format_Position_Format_Context_Page - International publisher, March 2018
- Product ID_Media type_PageCategory_Position_ChannelName - NL Publisher - March 2018
This simple naming technique will make it easy to aggregate data and get visibility on your key (actual :) ) metrics for a given product, media type, category...
At Adomik, we can help making this work. Please contact us to know more!