As everyone knows, working in ad tech and digital media is much like shooting a moving target. This is a rapidly moving industry in which all pieces are constantly evolving: technology, clients’ needs, expectations, and capabilities that promise to make media buying more efficient, targeted, and simpler. However, in many ways all of these new layers of innovation have made buying and evaluating media effectiveness more challenging.
The good news: one metric that we all use as a currency to manage spend is the CPM (cost per 1,000 impressions). The CPM is the measurement of choice for determining the cost, or value, of media being bought and sold across all platforms – video, mobile, and display. However, when buying media, people look at the CPM and often calculate it a bit differently. There is the actual CPM, the cost of media for a single line item on a campaign IO (insertion order), which is fixed. Then, we have the eCPM, which is the cost of media across multiple line items in an IO where each line item may have varied pricing, or there is added value included to reduce the actual CPM of the order. And lastly, the dCPM (dynamic CPM), which is the new metric associated with buying media on RTB (real-time bidding), where there is a bid price and an actual selling price associated with the media.
Publishers evaluating the value of their media will look at CPM and RPM constantly. RPM (revenue per 1,000 impressions) is the revenue generated on a per-page basis, meaning the aggregate of all ad placements on the page. For the purpose of this column, I’m going to focus on the CPM and eCPM calculations and their impact on yield optimization and vendor evaluation at the placement level.
Publishers usually look at their sales in two modes: direct and indirect. Direct being when inventory is sold directly to an agency or advertiser by the sales team. Indirect applies to inventory sold through third-party partners. Typically, a publisher takes inventory that has not been fulfilled by their direct channels and sends the remaining inventory to a third party. This could be a network, exchange, or an SSP (sell-side platform – technology for publishers). When selling indirectly, the publisher must look at the eCPM generated on that placement from every demand source they’ve enabled. Therefore, the eCPM calculation would be total inventory divided by 1,000, divided by the total revenue. However, many publishers attempt to track eCPM by vendor, even when that unsold inventory is passed on to yet another vendor to be monetized, which is not a fair representation of value (e.g., total impressions, divided by 1,000, divided by one vendor’s demand). We often call this daisy chaining. What should be evaluated is the paid inventory that a demand partner is bringing to the relationship, thereby changing the calculation to paid impressions divided by 1,000, divided by vendor revenue. This will naturally result in a higher CPM as the number of impressions is decreased.
There are some cases where the vendor a publisher is working with is able to fulfill 100 percent of the inventory, in which case the calculation is dramatically simplified for the publisher. Or, if the partner is an SSP bringing multiple demand sources into the mix, then the eCPM across all demand should be automatically considered in the calculation.
In order to get a true sense of the value created, it is critical for publishers to evaluate vendor demand value based on the impressions that are actually monetized. This will enable a publisher to review all vendors on equal ground and to allocate inventory wisely to generate maximum yield for the business.
Image on home page via Shutterstock.
There’s a significant increase of video content this year, and as it still hasn’t reached its peak, we’re analysing the most popular ... read more
Verizon has agreed to acquire Yahoo's operating business in a $4.8 billion cash deal, sealing the fate of one of the internet's pioneering giants.
Facebook will take the lion's share – more than two thirds – of global ad revenues for social sites this year, according to a report from eMarketer.