Advertising Guest Columns
3 mins read

Publishers: Always be optimizing…

The use of algorithms to optimize ad performance has provided tremendous value for digital advertisers over the last decade. Publishers need to demand similar technology to optimize their performance, too.

The major benefit of optimization algorithms in content recommendation ads is their ability to analyze thousands of ad-to-content combinations. This enables publishers to run ads that will generate the greatest revenue while providing advertisers the highest conversion rates and users the best experience.

In the past, media planners used their intuition to match ads to content. Today, technology can analyze thousands of ads against a piece of content and recommend the best performing ad faster than one can say ‘John Wanamaker’.

But in content recommendation advertising, the optimization seems to stop.

For example, if a publisher earns $0.17  per click after the initial optimization, within a month or two, the Cost Per Click (CPC) will likely decrease to about $0.12. 

During the initial publisher – ad optimization period, the content recommendation vendor’s algorithm runs ads from a broad range of product and service categories before landing on the category/categories which perform best. So for a financial services website, the best performing ads might be crypto-related, and for a recipe site, the highest performing categories could be cooking ingredients and kitchen appliances.

Instead of continuous optimization, content recommendation ad serving remains focused on the winning ad categories from the initial optimization. And advertisers run their own optimizations, resulting in lower bids which reduce the publisher’s revenue by 25-30%. This is particularly painful for publishers locked into long-term revenue-share contracts. Once again, content recommendation appears to tip the scales towards advertisers to the detriment of publishers.

So what can publishers do about this?

First, if publishers experience a double-digit decrease in CPC, it’s their right to request the content recommendation vendor to analyze the situation together to understand the problem. One potential culprit is ad fatigue, so publishers should focus on ad frequency as part of this analysis. But more likely, the competitiveness (or lack of) among bidding advertisers is the reason. Based on this analysis, publishers can work with the content recommendation vendors to re-optimize the serving of content recommendation ads and do so on a regular basis. With advertisers having access to conversion data, this process with the content recommendation vendor will level the playing field for publishers.

Furthermore, publishers should make sure that their contracts have contingencies enabling them to renegotiate when performance decreases unexpectedly and significantly. This means not agreeing to contracts based exclusively on revenue share but instead a hybrid combination of revenue share and fixed-price. Therefore, if revenue drops 20-30%, publishers should have the ability to renegotiate or even exit a contract, particularly if they are blocked from working with competing vendors due to an exclusive contract.

The use of algorithms to optimize ad performance has provided tremendous value for digital advertisers over the last decade. Publishers need to demand similar technology to optimize their performance, too.

According to data from the US Bureau of Labor Statistics demonstrating the challenges publishers are experiencing, 60% of newspaper jobs were lost between 1990 – 2016. And the trend hasn’t slowed down since 2016. Though newspapers are only one segment of publishing, it’s pretty clear that technology-driven ad optimization has followed the money trail. In digital marketing, the advertisers are the ones with the money. But without a robust publisher ecosystem, where will marketers advertise their products and services? That’s why it’s important for advertisers and publishers alike that publishers always be optimizing.

James Mortazavi
VP Revenue, WhizzCo

WhizzCo is transforming the native content recommendation space into a transparent, fair, and competitive ecosystem. By opening publisher inventory to multiple, competing vendors, WhizzCo empowers publishers to harness the best each has to offer with just one integration, thus raising content recommendation revenue by approx 37%. The company’s proprietary algorithm, a machine learning neural network, predicts the CPM from the 40 content recommendation vendors worldwide, considering geolocation, device, site, widget location, and more, and then serves the ad with the highest predicted CPM. Performance is measured on WhizzCo’s unified and intuitive dashboard.