Serving more than 10,000 sites, with 400-500 publisher applications on a daily basis, Ezoic is amongst Google’s largest certified publishing partners. Their platform provides publishers with access to machine learning technology making it easier for them to increase efficiency within their organizations while increasing ad revenue.
Zach Measures had the opportunity to sit down with John Cole, Chief Customer Officer at Ezoic, to get a glimpse into how they do what they do. It’s like a CMS platform on steroids: virtually anything that can be construed as cumbersome for a publisher can be managed and optimized with Ezoic.
Here are 3 key takeaways that fully support Ezoic’s success and appeal:
Unifying the moving parts. A big part of the challenge for publishers is the myriad of agendas they are managing between the 50+ plugins their sites are often hosting. Depending on what your goals are, some of these plugins might be producing conflicts of interest between one another. Ezoic operates on the server side, so it’s able to pick out what’s working and what’s not. It is constantly trying to improve metrics such as time on site, page views per visit, bounce rates and scroll rates.
Working out the kinks caused by various plugins also helps to improve the factors considered by Google’s Core Web Vitals. This specific set of metrics related to speed, responsiveness and visual stability of a site influences its ranking potential for search results.
Furthermore, many publishers enlist tools like Ad Exchange, Ad Manager or AdSense in order to integrate ads with their site. Once Ezoic can optimize the previously mentioned aspects of the site so that everything is working together, the potential for ad revenue increases by 50-100% almost immediately.
Multivariate testing with actual user behavior. Rather than referring to samples and extrapolation to make changes to a site, Ezoic relies on actual reality. They consider that the better performer in an A/B test doesn’t necessarily translate from desktop to mobile or from the US to Italy. In other words, A/B testing simply doesn’t cut it.
When different versions of your website will be shown to different people based on multiple factors, proper testing is crucial for finding the most ideal solution for each use case. Ezoic relies on multivariate testing to guide decision-making for what to show which users, ensuring that every user is seeing the best possible variation of the site at any given time.
There’s no catch. There’s not even a contract. John says they’ve structured the customer model based on what he’d want if he were in the publisher’s shoes. He doesn’t like handing over his credit card for a free trial, noting that it could take months before realizing you’ve been unintentionally paying for something that didn’t resonate.
You can simply plug your site into Ezoic, allow its machine learning technology optimize your user experience and watch your ad revenue increase. If you decide you want to go back to a less optimized experience and lower monetization, simply turn it off and stop using it.
“It might sound too good to be true. That's probably our biggest hurdle - this piece of machine learning runs on my site and it makes me more money - there's got to be a catch.” – John Cole
At the end of the day, John suggests we ditch our assumptions. Or, at least doubt them. Pay attention to the data as it relates to how people are responding rather than clinging to your fixed idea about how things work. AI and machine learning are great tools to employ that ensure our emotions and biases aren’t getting in the way of our own success. Weeding out what is no longer serving your site and enhancing what is will accelerate the growth you’re looking for.