WAA's Burby says new words will lead to better Web action
Jason Burby, the co-chair of the Web Analytics Association Standards Committee, spoke to DM News managing editor Cara Wood about the association's new report that provides standardized definitions for 26 widely used analytics terms. A writer for The ClickZ Network, Burby recently co-authored "Actionable Web Analytics: Using Data to Make Smart Business Decisions" and is chief analytics and optimization officer at Wunderman interactive agency Zaaz, Seattle. What follows is an edited transcript of the conversation.
Tell me about the Web Analytics Association and its mission.
Jason Burby: Web analytics has become a new industry and, even more important than that, it's taken on a new role within corporate America.
A few years ago a lot of people believed that there was a lot of power in Web analytics and there were a lot of people running in a lot of different directions. But training someone to do Web analytics wasn't like just taking a designer or a developer and just saying okay, now we're going to focus on something a little bit different. This truly is a different role and a unique role.
How the Web Analytics Association truly began is, æHow do we take the amazing power of Web analytics and how do we start to build a framework around it?' æAnd how do we start to get some common language and beliefs?'
Typically, in most organizations, some one might spend half of the their day on Web analytics. Or it might be one person sitting alone in a cubicle and their job is Web analytics. So, how do you help all of those people and get all of those people the same language, so that they can learn from each other and evolve the industry in a way that makes sense?
Really it came out of being a completely new industry without core set leadership other than the vendors and wanting to take it to the next level and start to understand the value of the data.
What is the impact of releasing 26 new standard terms for the industry?
Jason Burby: I've been on this committee since it started and it should not be as hard as it is to come up with these terms. The problem is that its like trying to have 15 people define something that all speak different languages. So while I would like to say that the publication of these terms is going to instantly revolutionize the industry; it's not.
It is a step in the right direction and it's a great, great first step. But we're not there. We are all very proud of the work that's been done. The key is where it's leading.
Our goal as a committee was to come up with some standard definitions and to come up with something, not for that uber-technical person who eats and sleeps Web analytics, but, more importantly, for the marketer, so that they can be the one that leverages the data. There are the campaign people, who are in touching the Web site every day. And now there's a language that they can understand and they can then put to work.
What are common misconceptions marketers have about Web analytics?
Jason Burby: A lot of them are based on just definitions. Someone might say, "We had a million people come to our site."
"People" is not a good measurement. Is it visits? Visits being, I may have come to the site 10 times and still, I'm only one person. It's very different if it's a million people coming once or 100,000 people coming 10 times. Decisions are going to be made based on that statement and depending on what your goals are those are two very different results.
A lot of times what you'll see marketers get and use this information and they'll start to compare things that shouldn't be compared. Or they'll start throwing around terms that can lead to making false or poor decisions.
What are some other terms that need clarifying?
Jason Burby: You used to hear it a lot in years past, people would say, "We got 40,000 hits to our site." Hits don't matter at all. In fact it's a term that we didn't define. We referenced it and I felt like you wouldn't really want to focus on it.
A "page view" was another area of confusion. How do you define a page? What is a page? How does a page work in a Flash program? How does a page work in the world of AJAX? How do all of those things come to together to a common definition of what a page is? A couple of years ago it was a lot easier to define what a page was. Now it's gotten very difficult.
What we've attempted to do is, we see the document and the standards to be a living document and something that we'll change and really focus more on technology but what our goal has been is let's put a framework in and offer some flexibility and offer how do you define a page?
For Converse [a client of Zaaz], the entire site is in Flash, so from the time you enter the site until the time that you buy a pair of shoes, you are in Flash. That's one page. But we've broken it out. And the definition allows us to say this is what we're considering to be a page for us.
People talk about bounce rates. And by bounce rates they really mean people leaving. People come into your site, only view one page and then leave. So people call it exit pages, exit ratios, bounce and there are about 50 different things and 50 different ways that you can tell the exact same thing.
So, how do we hone that and how do we understand that there are some differences? Someone immediately coming in and immediately leaving the site might be an exit, they might view multiple pages, or spend a significant amount of time on that page before leaving so again, we just want to know what a bounce is.
What it comes down to is almost not the individual terms themselves. The average VP of marketing is not going to know what a bounce is and they are not really going to care what a bounce is. They care about how the site performing. So getting down to some of these terms is about, how is the site performing and what does it mean to me?
Do all marketers care about the same metrics?
Jason Burby: First let me say that just because one tool provider measures a visit in one way, another tool provider might not count the same visit that way.
Going beyond that measurement problem, I really lump it into two categories. There are generic metrics, like visits, like number of page views, like how many visitors came to a site. Those are high-level metrics. The way we look at those in Zaaz is those are the high-level generic things. They are good to know but they don't drive any change to your Web site.
Web analytics is simply a means to an end. I think so many people get caught up in Web analytics, Web analytics, but none of it matters unless you can act on the data. And really with those macro metrics you can't really act on the data. Where you can now act on the data is when you dig in deeper. That is the second category.
What is the conversion on this particular offer or how do we get more people to click through? On the highest level a lot of companies are looking at the same things but the next step down they can be very, very different.
What role does direct marketing play in Web analytics?
Jason Burby: The DM world is interesting; the Web analytics practices, and then leading to A/B and multivariate testing, is just taking all of the direct marketing best practices and putting it towards the Web. All of these people are thinking that they are doing amazing great things on the Web. All they are doing is finding ways to use these really established direct marketing practices and apply them to the Web.
I think that some of the edge that the Web has is when you start to look at testing. When we are talking about on-site testing and we re doing things such as multivariate tests.
We have to take all of these different combinations, really an infinite number and then we can say after two days have good data and start to turn that over and impact everybody in a better way.
We are able to move a lot quicker. We don't have to worry about the printing. We don't have to worry about the mailing and some of those things. So we can respond and act quicker and typically test more things.
Then when you look at behavioral targeting technology we can really look at what's happening on the site, what are people looking at and what is going to work well for them. So, if someone comes from Google and has searched this term they are going to most responsive to this particular of offer. We can have five or six of those things throughout the site. Just because it's on the Web and it can be on the fly.
Where do you see the industry going, ideally, in the next 5 years?
Jason Burby: I think that the WAA has a big role to play in this. Right now I think people are still caught up in Web analytics for Web analytics sake.
We're starting to see the smarter companies saying, ok we've got to put this data to work. More of the top 500 companies are getting serious about the value that Web analytics can bring.
We're starting to see the combination of Web analytics behavioral data with attitudinal data, with customer satisfaction as well as competitive data. And then think the biggest trend that will occur is these off-site technologies.
In the past we've looked very myopically on our own sites. Now what we're starting to see with all the social media aspects is we need to understand the performance of our sites in relation to what the customers are doing in our retail stores and in relation to everything else that they are doing online. So when they go online and they are out reading a blog and then they come to our site, how can we better serve them?
What role does our site need to play in relation to all of those other blogs? Understanding your overall Web ecosystem and then acting on that data.
What challenges lie ahead?
Jason Burby: The analytics tools now provide so much data. So then the issue is how to you get people to focus on the right things? By helping people focus in not just by looking at every metric that's in Omniture or looking at what are the most important metrics. Or looking at, "Hey this is great insight, now let's change the site." If you start to look at the attitudinal or competitive media it starts to blow up things even further.
The problem is how can we as a collective industry focus in on understanding the things that matter, that drive the business the most and then acting on those.
To learn more about the WAA Standards Committee and to download the WAA Standards Analytics Definitions Volume I, visit: http://www.webanalyticsassociation.org/.