Yield management can be described as analyzing data and performing certain optimization techniques to maximize performance and revenue. One of the simplest forms of inventory optimization for digital publishers is accurately pricing inventory. While the concept may sound pretty straight forward, a yield manager’s job is highly technical – and far from easy.
To manage yield, the goal is to allocate impressions to wherever demand is highest – across various advertisers, devices, platforms and channels – for both direct and programmatic sales. Maintaining a holistic view of all revenue streams and how they talk to each other is key, making data analysis a mandatory skill for yield teams. Adding new tech partners, pricing rules or ad formats can have a major impact on revenue, and yield managers determine whether that impact is positive or negative.
Perfecting a yield optimization strategy takes time, skill and endless A/B testing, but there are best practices every digital publisher can implement before diving into the details. Follow along for part one (of two) articles sharing yield tips for publishers of all sizes and levels of expertise.
1. Understand the metrics that matter most by analyzing historical data
Perform a general site audit and start by analyzing historical visitor data to determine which metrics matter most to your organization. Know which pages or sections of your website have the highest and lowest traffic by day, month, quarter and other seasonality factors.
Do the pages with the highest traffic also have the highest sell-through rate or the lowest? How does your sell-through vary by vertical, device and ad format? How many articles do your visitors read and how far down the page do they typically scroll?
These types of questions often spark curiosity and lead to new questions, insights, and discoveries.
2. Leverage your math skills
Successful yield optimization does require some everyday math and logic, but it’s easy to understand once broken down. In order to drive overall revenue, you’ll want to find a healthy balance between the fill rate and eCPM for all digital ads not sold direct. Once you’re able to do that, you can further optimize between direct and programmatic. Let’s start with some important terms and definitions.
Ad Request: ad requests are simply the request to fill an ad slot with an advertisers ad once a page loads. They are triggered each time a page is loaded, but that doesn’t mean an advertiser will fill that request and serve an impression. Ad requests are counted whether an ad is shown or not.
Fill Rate: fill rate is simply the number of ad requests answered divided by the number of ad requests sent, expressed as a percentage. If an ad appears every time a page loads and triggers an ad call, the fill rate would equal 100%, which isn’t common.
eCPM: eCPM is the effective cost per 1,000 impressions. To calculate an eCPM, you’d divide the total earnings by total impressions served, and multiply by 1,000.
Floor Price: the lowest price at which inventory can be sold. Learning where to set your floor prices and other pricing rules for all types of inventory is another important piece of yield management.
After your floor prices are set, the next step is finding the optimal balance between fill rates and eCPMs. For example, let’s assume ‘Yield = eCPM x Fill Rate’ and compare two different reporting scenarios.
Report A: $1.80 eCPM & 85% fill rate. 1.80 x .85 = 1.53
Report B: $5.50 eCPM & 20% Fill Rate. 5.50 x .20 = 1.10
Despite report B’s higher eCPM, the final yield was lower than report A’s due to the lower fill rate. It’s easy to see why analyzing historical performance data to find the optimal balance between the eCPM and fill rate is so important.
Next, the yield manager in this scenario would probably lower the floor price of report B so additional advertisers would bid on the lower price and increase fill rate. That optimization strategy would produce a greater lift in overall revenue.
3. Understand the impact of unsold inventory
Analyzing and optimizing yield based on eCPM and fill rate is a given, but don’t forget to measure lost opportunity as well. Unsold impressions affect the overall value of a publisher’s inventory as much as the sold impressions do.
Consider calculating the eCPM based on ad requests instead of sold impressions, or measure RPM (revenue per thousand page views) to forecast total ad revenue. Pageviews and ad requests measure the inventory on a website as a whole, not just the impressions a publisher manages to sell. Measuring lost opportunity allows for greater insight into trends and more intelligent sales planning for the future.
Recapping the highlights
Remember to focus on quality inventory, analyzing historical data, understanding your audience and optimizing yield by considering these key metrics: eCPM, fill rates, rCPM, RPM, ad requests and floor prices.
Keep an eye out for part two, where we plan to deep-dive into best practices specific to the latest industry challenges. We’ll cover measuring advertiser buying behavior, seasonality, unified pricing rules, restructuring your ad server setup, and more.