I remember when I started making money on banners, in 1996 with a company called Flycast, where I ran impressions at something like $100 cpm on a page that loaded advertising for an online game. I made about $15,000 in a few days, and used that to really launch my online marketing business. Until 2005 or 2006, if someone saw an advertisement in their browser, they would respond to it – but even then people were talking about banner blindness, and how to really measure impressions.
However, with Agencies making a killing selling impression based media, and billions being invested into technology that just counted clicks and impressions, no one wanted to really try to change the model.
Now we’ve come to the realization that gauging impressions is completely useless, and anyone selling them is usually looking for a quick buck. Audiences dole out their attention in tiny, goldfish-sized parcels, and online publishers are starved not only for these attention particles, but for information on where it’s coming from and how to get more of it.
Jacob L. Nelson and James G. Webster of the International Journal on Media Management at Northwestern University recognized the need to further evolve media measurement.
They designed a research study to analyze the relationship between the amount of time visitors spend on sites, news sites in particular, and the number of unique visitors the sites receive.
They used comScore data to analyze 887 news websites in their study “Audience Currencies in the Age of Big Data.” Through then analysis of this data which measured an individual’s time spent on a news site and the average number of minutes spent on a specific page, Nelson and Webster found that these two variables most accurately represented an individual’s level of engagement.
In further probing, they also found no significant correlation between the number users a site reaches and the amount of time visitors spend on that site.
In other words, the number of visitors to a site does not necessarily correlate to the level of engagement. In fact, unique visitors and attention measures are very different measurements, one identifies the quantify of visitors and the other the quality of visits (engagement).
But how do you measure human attention?
MOAT has some ideas. They’re able to measure how long a user is on a page, how much and how quickly the user scrolls, and other attention-linked interactions. Measurements like in-view time stand out because they’re clear and simple — they indicate how long the ad was in front of the user on the page. Measuring scroll velocity helps to understand whether or not a user was reading content, or just skimming.
Kik recently rolled out: a switch to measuring “attention metrics” – looking at how engaged a user is based on how much time passes between messages being sent. This new attention metric allows Kik to understand the habits of the person behind the screen. Kik can now account for multi-tasking, conversations that continue outside the app and even the individual user’s chat style!
Media Intelligence Co. – a joint venture between Nelson-Field and AI tech firm Jemsoft – is working on projects with some of the largest companies in the world to utilise cutting-edge AI technology to measure attention, facial recognition and brand effectiveness. One of the projects, known as Attention Plus, measures consumer attention above and beyond the standard gaze technology currently offered in the market.
Whatever you think about Attention Metrics, they are definitely the key to online advertising growing in leaps and bounds. The day of Impression Based advertising needs to completely go, and never come back.