[MUSIC] So Rich, what type of measurement systems are there and then, of course, how do they differ? >> I think of there being seven categories of content measurement systems. They each measure different things in different ways. The first is what I would call panel-based measurement systems. These are similar to the tools used to measure television and radio usage, for instance. There's an agency of some kind that provides ratings. They look for a panel of users. Those users sign up and provide some information about themselves like their demographics. In the case of these web rating agencies, those users, those panel members, install monitoring software that enables the agency to understand where they visited, what sites they've spent time on. The agency can then extrapolate the behavior of the entire audience from that panel. These are widely used by sites that sell advertising online and businesses that advertise on those sites. These are how they decide what sites to buy ads on and how much money to spend. They have some limitations. They undercount workplace usage, because in the workplace the technology department may not allow that monitoring software to be installed. And they don't work well for relatively low traffic sites, because the panel is unlikely to have a representative group of users for any site that's not widely used. >> And depending on the country, I suppose there must be parts of a country or a population that don't have access to this that they can't reach. >> Yes. And in fact, I think there are many countries where there aren't actually ratings agencies that provide this kind of service, because there isn't a well-developed, for instance, online advertising economy that supports it. >> Mm-hm, the next one? >> The second category I would call server-based measurement systems >> These are by far the most widely used systems by sites of every size. Google Analytics is a good example of a server-based system. In China, the comparable system is Baidu Analytics. There also are more expensive and sophisticated tools like SiteCatalyst from Adobe. The way these work is that the site administrator puts a very small piece of code on every page of the site. That code runs when the page is opened and that generates data that the server compiles about both the page that's viewed and the computer that's being used to access it. Then the content strategist can log into that system and see the data compiled for their review. Typically, for most content sites, or content strategists, this is the best source of audience measurement data available. >> So I could conceivably set one up even in my organization, to work within the organization, and it would measure how many times people went to some page, some part of the site? >> Yes, and you can use these systems to measure public facing websites, to measure sites that are available only to people who have logins, to pages that are on an intranet that only the people inside your company are using. They all work the same way, and they can measure the same things. >> And beyond that, what's the third one? >> A third category I would refer to as realtime analytic systems. These are newer. They are designed to help people who manage very popular websites to understand exactly what's happening on the site at this moment. A major news publisher, for instance, would use a site service like this, Chartbeat is the service that's best known, to see what's happening on their home page right now. For instance, they may have literally tens of thousands of users visiting their site at this particular moment. They could look at what content on the site is being viewed the most, where that traffic is coming from. They could look at the homepage for the site, at a particular headline. They can say this particular site, this particular headline is not getting clicked on as often as the last headline we had in this spot. Maybe we need to rewrite the headline or maybe we should put a different story in that spot. And what this enables an editor of a site like this to do is to make data-driven editorial decisions in real time in a way that was never possible in traditional media. >> And the next one? >> The fourth category are what I would call publishing tools that accomplish things that a website wants to accomplish, and along the way, generate analytics. So a good example of that is a URL shortener such as bit.ly. You would use that system to create a shortened web address to include in a social media post. So it accomplishes a particular purpose where your social media post, you want it to be short because maybe there be a number of characters limitation. But along the way, once you've used that shortener, you can tell how many clicks that link got, when they were clicked, what channels generated those clicks, and that can be very useful to you as a publisher. Another example of a tool that generates analytics are the very popular share buttons that exist on many websites, typically alongside or at the bottom, that basically make it really easy for your visitors to click and share this content with their friends on a social network. It accomplishes this one click goal for the publisher and for the user. But along the way it generates data that allow the content strategists in the organization to understand which kinds of posts are getting the most engagement, which social media platforms are the most valuable, when is the clicking taking place, etc. >> And I suppose the fifth one, then, must be social media analytics. >> Yes, every social media platform comes with its own analytics system. For instance, if you were using Facebook as a tool to build your audience and build engagement with content, you would have access to the Facebook Insights Dashboard. And the Facebook Insights Dashboard allows you to see which of the posts you have published are getting the most engagement and even what form that engagement takes. Is it people clicking on the links and going to your site, is it people sharing it with their friends, etc. And if the social media platform requires the user to provide information about themselves, for instance, Facebook, most users put in their age or their gender or their occupation. It's possible then in Facebook insights to understand the demographics of the people who are following your page, to understand when they're most likely to be online so you could choose to use that platform at a particular time of day when people are most likely to be there. >> If that's true for social media, which is enormous, there must also be a way to measure video, which is exploding and is every bit as large. >> There is a whole separate set of tools available for measuring video consumption and video engagement. They measure different things. So for instance, if you were measuring engagement with video, typically something you'd want to know is, how far into the video do people go before they stop it, pause it, or leave the page? That would be a very important thing to know and a video analytics tool would give you that information. >> And if that's true, then there must be ways to divide video or think about it as it moves elsewhere. Let's talk about mobile. >> Generally, if you're looking at content on a mobile device, you are either looking at an app that you have downloaded and installed on your phone or your tablet, or you are looking at what would be called the mobile web, which is a website typically formatted so that it looks good on a small screen, a technique that's been called responsive design. Mobile content viewed on the mobile web can be measured through Google Analytics and Exact or any of these other analytic systems the same way as it would be on a desktop machine or a laptop. Whereas if you're viewing content through an app, you need a different system to measure that. >> [MUSIC]
