Analytics customers in the entertainment and media industry are increasingly using Adobe Consulting’s statistical clustering techniques to profile their visitors; almost all of them are also awakening to an important fact: most of the page views on their sites are driven by a very small percentage of their visitors.
I’ve personally constructed statistical clustering models for more than a few large media companies with sites ranging from over 10 million unique visitors per month to as high as 100 million unique visitors per month, and each of them have come to a similar and startling conclusion. Each found that 6% to 15% of their visitors are generating 50% to 75% of the total page views on their sites.
This skew is further underlined when you consider that bouncers (single page view visitors) make up 40% to 60% of the total visitors for most media clients, but will typically only contribute 5% to 10% of the total page views. Additionally, I’ve noticed that 75% of all page views are usually generated by less than 25% of total visitors.
Most online advertising is sold based on the total number of impressions without considering that most of those impressions will likely be consumed by a very small proportion of those visitors. Therefore, the ad reach may not always be what it seems and highly engaged users can become bombarded with the same ads over and over again, or they quickly blow through a few premium ads and are left with nothing but remnant ads.
Also consider that this could be compounded by the fact that media ad inventory sales teams are usually given incentives to sell huge ad contracts for whopping amounts of impressions rather than having greater high-CPM contracts with fewer impressions per contract.
This presents a two-fold problem that many in the media industry have not yet realized:
- You probably have an ad frequency problem which can hurt your user experience; if you don’t,
- Your advertisers probably cap impressions per user which presents the second problem: Your most valuable traffic will hit that cap very quickly leaving them with only pitiable low-CPM ads – diminishing your revenue.
Luckily, Adobe’s Digital Marketing products combined with Adobe Consulting have the tools you need to help you understand and combat this problem.
First, you’ll need to understand how drastic this situation is for your site. In my opinion, the best way to do this is using visitor clustering techniques provided by Adobe Predictive Consulting. We can take your SiteCatalyst data, summarize it at the user level, and run predictive profiling techniques to have the computer identify your high and low value audiences. These algorithms determine what makes these audiences more valuable than other visitors.
Second, you can export the list of these high-value users to Adobe Audience Manager. AAM is the premier data management platform for media industries since it allows you to combine analytics with ad targeting and experience testing. It also allows you to combine predictive cluster information with demographic information (for further consumer insights) and sell ad inventory to these specific visitor profiles onsite and offsite. With AAM, you can also make sure that your high value customers aren’t completely saturated by the same ads over and over again. AAM also gives your ad sales team the ability to sell the extremely valuable traffic since you can now personalize the ad messaging to each of your high value users.
Finally, AAM integrates with Adobe Test & Target. This integration allows you to present extremely personalized experiences to each of your high value customer segments. Not only is this shown to increase user engagement, but it will give insight into which parts of your site are working for each consumer segment and which parts are not.
If you own a media site, it’s going to be increasingly important in the future that you understand your visitors and act accordingly. Adobe has the tools and resources you need to make this happen. If you’d like to start this type of analysis for your site, reach out to your Adobe rep today.