At last month's eMetrics Marketing Optimization Summit in San Jose, Calif., BtoB
Editor Ellis Booker assembled a roundtable to discuss the current role and use of Web metrics. One theme that emerged: Marketers are asking for analytics to defend their efforts and, not trivially, their budgets. Roundtable participants included: Anne Commisso, Cisco.com marketing Web metrics team lead, Cisco Systems; Moe Rubenzahl, executive director-Internet marketing, Maxim Integrated Products; Michelle Rutan, Web program manager, business intelligence, National Instruments; and Jim Sterne, founding chairman of the Web Analytics Association and producer of the eMetrics Marketing Optimization Summit.
What's been the impact of the current economy on investments, use and strategy when it comes to metrics?
We're seeing heightened interest and lowered investment. It's obvious. It's valuable; but it's not a fire extinguisher. And, right now, there are fires. And unless you can show me specifically how buying this tool and training all these folks will cut costs, I can't do that right now. But some of those tools, products and systems can show how to cut costs.
It creates a real change in the whole funding of the marketing priority. A year ago, there was a real emphasis on brand awareness. Now there's much more of a focus on lead generation and conversion.
We've been diverting resources from some of the items that involve expenses to other ways to do those things. So our advertising budgets are heavily cut, and that's an opportunity for the Web because the Web is seen as “free” marketing. We are moving resources that used to be focused on creating and running ads [to] some of the subjective aspects of Web measurement in areas where we haven't had the resources to do so before.
We've been seeing increased interest. We have a direct sales force. It's very large, and they're engineers. They're not cheap, [and] our products are not cheap. We do e-commerce on the Web, but we're trying to establish where the boundaries are. Who should be buying on the Web? When should they buy on the Web? How can we utilize [the Web] so we don't have to hire 200 more salespeople? How can we do that but at the same time keep up with the customer service level? We've been doing a lot of analysis for that.
We've been spending a lot of time lately trying to use Web traffic data to give us more understanding of our customers. We've looked at engagement—people who have bought products, people who are prospects. We're looking at a second formula for engagement—things like recency and frequency. The one I'm the most excited about, and that has the most traction right now, is trying to take the visitor traffic, the visitor history, and make that something tangible that we can use to categorize pros- pects, to categorize customers and what they're looking for now.
BtoB: What shifts are you making?
I've seen a big increase in social networking. If you ask marketing teams now, you actually have people who say, “My role, my responsibility is social media.” They're blog-posting; they're Twittering. I see a lot of marketing assets being placed on Facebook. They're placing Cisco videos on YouTube.
The main thing is the decrease in advertising. The other is that we're doing more voice-of-the-customer examinations so that we can find out what is and isn't working for the customer.
We're using that information to decide where to put our investment in terms of Web content development. The message we're getting is that content works. [We're asking] where we can put our internal resources to develop higher quantity and higher-quality content that will track the right kind of customer who engages and who goes to the appropriate destination.
I don't see enterprises making interesting shifts because of the ability to analyze campaign response, but I do see individual campaigns tactically making interesting shifts in the middle of a campaign. [For instance] people who are shifting money around to different properties based on not just the amount of traffic they're getting, but the quality of the traffic they're getting from a given keyword, phrase or a given banner ad.
We have noticed people defending their budgets. They're a lot more interested in being able to prove not just “I drove this many people to the site” but carrying it over to “Did it make a conversion?” Before, I never heard from our print newsletter. Now they're like, “I saw someone do this. I want to learn to do this.”
We're in a paradigm shift in my group. A couple of years ago, we played a consultant role. They wanted to know if there was any interaction, any level of click. So we had to run around and do all sorts of “tech-wizardry stuff” to make sure everything was tagged and everything was measurable. Now we're moving to more of an adviser role and recommending what should be tagged. We don't need to tag everything. We need to think in advance, “What is success? What are the five things you really want to measure?' ”
There are some constant challenges. One is when people arrive at our site, we don't have a conversion of that. How do we create the equivalent of a conversion for that? How do I evaluate that this customer was more valuable than another customer. I know what criteria would go into that measurement, but I don't know how to measure it. That's a constant challenge: How can I evaluate the traffic from one page compared to the traffic from another page?
Cisco's tried to do some behavioral targeting. We do weighted score ranking of our visitors so we assign a “topic interest” score. We're just looking at three basic measures: Yes/no—Do they visit our product? Did they do it recently? And did they do it frequently? Based upon those three common measures, we can give you a topic interest score. So if you are on a different area of the site, or if you're on one of the primary pages, we can try to target content to you.
BtoB: What are some mistakes people make with metrics?
Not doing it at all. The next problem is not understanding the numbers themselves—looking at numbers and assuming they mean X when in fact they mean Y—and making bad business decisions because you're dealing with false assumptions. The important one is not measuring far enough. The mistake that's made is people stop short of measuring to the business outcome.
We got ahold of [“Web Analytics Demystified” author] Eric Peterson's calculation and everyone was like, “We need an engagement metric.” So we went down this path. We worked with Unica [Corp.] and created all these metrics. And after a while, it's not going to mean anything. I would tell the director, “OK, 47%.” What are we going to do with that? Then it turned into, “What is the function of our site? Do we really want to measure it on this high level?” So now we have what we're calling “presales engagement.” [We] figure out what they're doing that's going to rank them higher, and then do that. That's been a lot more successful.
BtoB: The analytics discipline is getting better and better about locked-up things: Search engine optimization, keyword purchases and auctions, e-mail campaigns, click-through analysis. And then we have this tsunami of social media. What do you measure? How do you rank it? How do you score it? How do you magnify its presumably positive impact?
Within our Web marketing team, we have a small team that's devoted to the social networking space. They're measuring responses back to Twitter, how many followers, how many mentions in the blog. They're using third party tools like Second Opinion and others that measure the tonality of the blog posts. So we're starting to see a lot more numbers with social media.
We're looking very hard at the social aspect, but over the years we've had small success in engaging our customers because engineers tend not to be social in the same way. So Twitter, discussion groups, Facebook ... that's a real push for us. We have a Twitter presence, [but] it's not much. One of our competitors has a much stronger presence. ... It also is not much. Where we do
think there's value is in China. And that's a much harder nut to crack because of language issues, because we can't just set up some social media in the U.S. and try to track Chinese engineers. We have to have some local presence.
Two parts: Part one is exactly the same as you measure everything else. Does it bring traffic to the Web site and is that traffic qualified, doing what we want when they show up? The second part of it, though, is the sentiment analysis. And there's a whole new realm of tools that are out there measuring [whether] what's being said about you is positive or negative.
In our top tier of metrics, social media hasn't infiltrated yet. It's more on the third tier. Last year we had a conference, [and] three or four people from the team were Twittering while there to get people to register. They wanted to be able to measure how many people had clicked through, and they did the tiny URL things to see people coming into their site. It was very exciting to their campaign. It was free.
One more thing. What we see with the rich media trend is ... a lot more third-party hosting of this content. And that poses a whole other challenge on the measurement perspective. The whole experience may happen off of our domain, but [our marketers] still want to know about it. It's a new challenge [for] data collection.