AJAX is commonplace now -- Gmail, Facebook, and many other popular media and social networking sites use it for navigation or other functionality. But when AJAX first started making a splash, it got a lot of people in the online media business thinking about different ways to present their own content as well as advertisements. What if readers didn't have to navigate away from the page to read a related article? Was it finally a chance to break away from the pageview paradigm for measuring content popularity, advertising impressions, and site growth?
A lot of people hoped so. Yahoo presented its case for moving away from the pageview in 2006. Journalists and media pundits looked forward to banishing slideshows, celebrity news and other pageview-friendly content to the sidelines. Some people in the advertising industry began to get excited about measuring "engagement" on media sites instead of raw ad impressions or ad clicks.
Fast-forward to the end of 2009, and have we realized these visions? Not quite. Time spent on site is a useful metric, but when it comes to online news and advertising the pageview still rules the roost. You don't have to look far to see the evidence -- slideshows, top-50 lists, and celebrity news are everywhere. Standard display advertising is usually sold and measured according to CPM (cost per 1,000 impressions) or CPC (cost per click). When new media kingpin Nick Denton talks about how Gawker and his other blogging creations compare to the New York Times, he cites pageviews.
So why hasn't time on site and AJAXy display models knocked the PV off its perch? I recently had a chance to answer that question in an email to a media scholar, who remembered the talk of change from the middle of the decade and wondered why publishers still haven't made the switch. I wrote:
... For years, I’ve been hearing about “time spent on site” being used as a selling point for online advertising, which would be wonderful for some very unique AJAX-powered displays of news content. Madison Ave. would probably love it too, as it would potentially allow for creative campaigns that leverage highly engaged users who are not jumping off to the next page right away.I didn't say this in my email to the media scholar, but Yahoo actually got it right in 2006 -- they were just five or 10 years too early. The pageview may still be king as the decade winds down, but it won't sit atop the online publishing throne forever. I think that the next decade will see a huge change in the way online advertising and news content is measured and sold. The pageview will fade in importance, and various measures of engagement will replace it to determine value and reach. It may take some time to work out the technical and industry roadblocks, or Google may speed things up with a new product announcement, but it will change in the teens.
But it hasn’t happened yet, at least not on the dozens of news sites I visit every day. I think there are several reasons why online publishers haven’t embraced AJAX-based news display, and instead stick with page views:
Now, in the mobile world things are a bit different. On my iPod Touch, I see that the NYT app does allow for long stories and menus to be slid underneath an advertising and navigation overlay. It’s not AJAX, but it supports a model for viewing content and ads that is not dependent upon the almighty pageview. I suspect that the NYT take this approach out of necessity – the iPhone API supports it, it reduces the numbers of calls to the NYT and ad servers to pick up and process new content, and relatively small pieces of content on a small screen can serve an outsized purpose simply because they are spending more time in front of readers’ eyeballs.
- Most online display advertising is still booked according to impressions and clicks (CPM and CPC) which favors page views and/or unique visits over time on site. If it came down to one user being exposed to the same ad for 1 minute vs. one user seeing four different ads for 15 seconds vs. four users being exposed to the same ad for 15 seconds vs. four users being exposed to four different ads for 15 seconds, I believe most publishers and advertisers would prefer the last three options.
- The technical foundations of the online advertising ecosystem -- display ad standards, technologies used by ad networks and ad serving companies, metrics packages (such as Omniture) and news content management systems -- are oriented toward CPM and CPC campaigns. Any AJAX-oriented system would have to be compatible with these technologies to make a big impact, which reduces the flexibility that an AJAX-driven display allows (for instance, you can’t have old content slide horizontally off the frame using AJAX if a standard “skyscraper” unit appears in the right rail – such ad units aren’t designed to be counted or turned off in such a manner)
- Costs for developing AJAX-driven news sites are high. There are some experimental sites out there, but I haven’t heard of any production-ready modules that can be easily slotted into popular CMSes, such as Drupal. That means publishers would have to do a lot of customization and integration work on their own, which can be very expensive.
But for such a system (AJAX or otherwise) to work on the Web, there would have to be a major industry shift among publishers and advertising platforms to support it. My guess is it won’t happen soon, unless a major player develops a product or service that makes it attractive to do so. Google has the clout and the resources to do something like this, but an news/ad industry task force could also get things moving.
Naturally, this will have a corresponding impact on the type of news and information that sites provide to their readers. Will it mean a return to long-form journalism? I don't think so. People don't have the time, attention spans, and freedom from distractions that once allowed them to plow through a 3,000-word feature from start to finish. Instead, I think that an increased emphasis on reader comments, online discussions and multimedia will be used to keep people on sites longer. And, although it will make journalism purists cringe, the 50-chapter "best of" lists will continue to thrive, albeit in AJAXy wrappers that don't necessarily generate new page views. As long as the definition of "engagement" or "time on site" is the most important metric to advertisers, any content that attracts eyeballs for more than a few minutes is fair game.
More posts by Ian Lamont on the future of media:
- The crisis in journalism: Short-term hope, long-term uncertainty
- The Tiger Woods accident simulated in 3D shows the future of newscasts
- The business of media: The newsosaur speaketh
- The business of media, part II: Blame journalists' ethical lapses?
- Boston Globe advertising: What's wrong with this picture
- Boston Globe relaunches forums, but they're doing it all wrong
- Google News: Biased or broken?
I see evidence of a lack of a standard metric of content's value or draw every time I explain how PostRank works, what it analyzes, etc. Especially when people argue with me or try to tell me how our technology should work. :)
ReplyDeleteTo some of us it's been obvious for years why a single metric -- pageviews, links, or what have you -- leaves out much of not only the content's lifecycle and "story", but, more importantly, also that of the audience.
To many, though, they're still back at figuring out how to draw, maintain, and grow an audience, let alone determining how to consistently measure and monetize them.
Even though PostRank focuses on online social engagement (more on what that all means here and here), we recognize that, as important and integrated as the social web has become, it's not the whole story. Hence why we integrate Google Analytics into our Analytics service and why lists like the AdAge Power 150 and Atlantic 50 integrate both PostRank and their own audience and publication-relevant metrics.
I'm not sure there'll ever be a gold standard in measurement, simply because there are so many different ecosystems -- news and social and mobile and whatever comes next. But I do agree that the focus on audience activity is the better foundation.
It's no longer possible to tell even a fraction of the story from a static number like a pageview. I went to that page... ok. Did I actually read it? All of it? Did I understand it? Like it? Did I save it? Share it? Did I share it with colleagues or friends or my mom?
Knowing what the audience is doing on (and off) the site with the content will tell publishers and advertisers alike far more about the content's success or value.
I do think the page view has found a longer life on the web because interactive agencies and ad sales teams know how to sell them, media buyers know how to price them and both have standardized their clients' understanding of them. Old technology works, and new technology risks revenue I know how to generate. Tough to change that behavior. How long has television held on to the 10, 15 and 30 second broadcast ad?
ReplyDeleteLeslie: Your point about broadcast advertising is interesting. I think the 30-second spot's longevity can be explained by several factors, including broadcasting's one-way nature, the dominant ratings systems' measurement criteria, a relative lack of innovative possibilities for TV and radio ads, and (until recently) a lack of competitive platforms.
ReplyDeleteMelanie: The PostRank concept is valid, but I wonder how dissimilar content types and a lack of standard design and interactive elements affect comparisons of different sites. For instance, the audience for an Apple-focused publication is probably more likely to comment and critique than the audience for a CRM site. I may also want to engage with the content on a ZDNet blog, but can't easily read the comments and don't want to deal with the irritating registration process.
This also brings up the topic of customization. In the age of paper publications, an advertiser dealt with standard formats and publishing technologies, with very little ability to do something truly innovative beyond fold-outs. Ditto for TV. Not so in the online world. Not only are sites very different in their designs and technological abilities, but it's very possible to do a completely custom campaign based on unique content elements or software abilities. Want a 12-second rollover widget that plays an animation of a well-known blogger from the same site manipulating product X? It can be done, for a price. This enables some interesting creative possibilities, but is very labor-intensive and ultimately not that scalable, compared to standard display advertising online or traditional print and broadcast campaigns.
Readers: I stumbled upon a very interesting essay by Shelby Bonnie, CNet's founder and the former chair of the IAB, entitled "Let's Kill the CPM." The conclusions are similar to what I've written above, but written from the perspective of an ad industry insider. He also offers a general way forward, based on (as the title suggests) abandoning the CPM altogether.
ReplyDeleteOn the other hand, TechCrunch points to a JP Morgan note that says online display advertising is headed for a rebound in 2010.
How quickly things change. Nick Denton, the Gawker Media head honcho who I cited above, is moving from pageviews to unique visitors. I am not sure this will end mindless slideshows as his memo suggests, but it's worthwhile noting that raw impressions are being downplayed.
ReplyDelete