Optimization Techniques for Your Publisher Teams

For all the talk of automation in programmatic advertising, the truth is that the business wouldn’t work without people. Pretty much every company in the space relies on a front line of practitioners to interpret data, guide clients on tactics, and tell the machines what to do. These are the services teams at exchanges, and the specifics of their jobs vary by where they work and who their clients are.

Using our interactive analytics, publisher services and sales teams at exchanges can optimize performance for their publishing partners by evaluating advertiser and content category blocks, implementing new ad formats and augmenting ad requests, and adjusting price floors.

 

Evaluating Advertiser and Content Category Blocks

Publishers typically block advertisers and content categories for two reasons:

• To ensure the quality of their advertisements
• To minimize channel conflict

However, unnecessary blocks can dramatically reduce competition for your clients’ inventory, thereby limiting fill rates, CPMs, and revenue. In the Facet Dashboard, you can quickly estimate the revenue a publisher could potentially receive by removing a block by filtering bid data down to a single publisher and changing bid status to “blocked advertiser”. Then, add splits to show the “adomain”, aka advertisers, that are being blocked.

 

 

Sorting by total bids, you can see that for the past few weeks this publisher blocked over 23 million bids from Brand 1. By also checking the average bid price and average bid floor, you can see that Brand 1 is bidding over 10x the floor price. So how much money is your publisher leaving on the table with their current block?

23M bids x 9.4 cpm / 1000 = $216,200

Now, multiply by a typical buyer win rate, say 10%, and your client is looking at a potential revenue gain of almost $22,000 for the last two weeks.

While not all publishers will be open to removing advertiser and content category blocks, it’s worth informing them of the opportunity. At the very least, your advice may lead them to seek out a direct deal with the advertiser via private marketplace or another sales channel.

 

Implementing New Formats & Augmenting Requests

The highest-performing publisher services teams spend significant time understanding the demand landscape for their exchange, and for individual clients. One way to do this is to study bidding frequency and pricing signals on various ad formats, device types, and user attributes. Your publishers can use these insights to inform their pricing strategy, fuel direct sales efforts with new leads, and create inventory packages for private deals.

Educating sellers on the pros and cons of different inventory types is critical, especially if your exchange supports a variety of media and formats. In the example below, eCPMs for video ads are about 7x higher than banner ads across the exchange. However, overall demand for video ads is much lower, with a bid depth of only 0.36 bids per auction compared to 0.71 overall. This, along with higher price floors, results in clear rates about 7x lower on average for publishers.

 

 

In this case, it may be best to advise the publisher to leverage video ads in addition to banners, which will yield them higher CPMs. Again, talking to sellers about marketplace trends doesn’t just help them think more creatively about the levers they can pull, it helps you forge a deeper understanding of their goals.

In another example, ad requests with Latitude/Longitude data (in the row marked “Yes” in the image below) clear 30% more often, result in 20% higher CPMs, and are generally in higher demand with 11% more bids per auction. Sharing this information with your publishers will help them determine the cost/benefit of collecting location data, and it will help them break their average CPM into more understandable parts.

 

 

Adjusting Price Floors

Price floors are another business rule that significantly influences performance. Typically, exchanges enable publishers to set floors based on geography, ad unit, and other inventory types. If a publisher simply wants a high fill rate they will usually go with a low price floor. If they want a high CPM they’ll go with a high price floor. The trick is to find the sweet spot so your publisher can maximize their revenue.

To figure out the optimal price floor for a specific slice of inventory you must first understand the bid landscape. By sorting bids in price buckets, you can analyze how buyers are valuing your clients’ inventory so you can determine where to set your floor.

In the scenario below, the publisher has set their main price floor at 2 dollars for 300 x 250 in the United States. This has resulted in a eCPM of 2.07, a clear rate of 12%, and revenue $42.3k. In this case, we have a good indicator that we can improve fill and keep eCPM high: the Avg. Bid Depth suggests a lot of demand for this inventory.

 

 

Looking at the bid price buckets below we can see there are some CPM ranges that have a fair amount of bids, particularly at 1.15-1.20 and 1.85-1.90.

 

Since the publisher already has their floor at 2 dollars, they might be willing to lower it to 1.85 to increase fill rates and still have a good chance at keeping their CPM close to $2.