Patti Freeman Evans of JupiterResearch shares some key findings from the analyst firm’s recent survey that looked at best practices in the use of Web marketing metrics. Her insights are key to anyone responsible for driving revenue via the Web.

ATG: Patti, tell us about the research project and its major findings.

FREEMAN EVANS: We surveyed retailers – both multi-channel retailers as well as online-only – and wanted to find out what factors were most responsible for improving business results. What analytical tools and measurements are the best? What areas of the marketing process are likely to deliver the best returns? What returns on investment can be achieved? And so on.

And one of the key findings was that fully half of the retailers we spoke to – that is, 50% – are getting either no return, or a negative return on investment on their use of metrics. So, they may be spending money on a Web metrics package or even using Google Analytics for free, but have not been able to derive business benefits from these performance metrics. At the other end of the spectrum, we found that 22 percent of retailers are actually doing a very good job, driving some very significant gains using their metrics. In the middle, about 28-30% are getting some benefits, but have lots room to improve. And these percentages haven’t really changed from two years ago when we did a similar study.

What was really interesting was that it didn’t matter what tool they were using, it didn’t matter what size they were, and it didn’t matter how many people they had thrown at their metrics efforts – none of these made the difference between who was good and who was not good. The real differentiator seemed to be their approach to using metrics, their process.

ATG: How is it even possible to have access to analytics, and yet not improve results?

FREEMAN EVANS: When we were doing research for this report, we interviewed an awful lot of retailers, and many of them said, “You know, I’m just overwhelmed by the information.” Now, as you know, the level of detail that’s possible to track can be excruciating, and picking out which metrics are going to be useful is really not that easy, necessarily.

But I would argue that it’s not simply a matter of too much information. In fact, we found that across the board, the best performers actually use metrics at a higher rate than the others. So the real difference isn’t an overabundance of data, it’s that the poor performers haven’t figured out how to approach and apply the information they have at their disposal in a strategic way. Because this is all new to them. Retailers in particular have, over the last 20 or 30 years, operated against a very straightforward set of metrics: gross margins, turns, dollars per square foot, same-store sales.

The retailers who have figured out how to make metrics work understand that you have to get out ahead of the wave: they actually approach the Web in a proactive way, where you come up with a hypothesis, test specific alternatives and then see what the numbers tell you. The poor performers on the other hand were far more likely to use their metrics in an ad hoc way – look at today’s click reports, and react.

We heard a lot of, “Well, that change we made – we didn’t measure it because it was clearly the right thing to do.” But if you don’t measure the effect, you don’t know how right it was, and you don’t know how to make better decisions going forward. And you can be sure that sooner or later your competition is going to start making those measurements, they’re going to get smarter about how they’re making their decisions, and you’re going to be paying the price.

ATG: What kind of ROI improvements are we talking about for those who adopt a more scientific approach to performance metrics?

FREEMAN EVANS: Our survey found that the top performers have extraordinarily better results in terms of moving the needle on conversion rate, on their ROI from search and e-mail marketing, on average order value, and so on, across the board. Our middle 30% of retailers, for example, earned a 5 percent increase in their return on investment from e-mail marketing. That’s good, but a 19 percent bump is an awful lot better – and that’s what the top performers averaged. The top performers also averaged a 24% increase in conversion rates, a 19% improvement in SEO, 13% in average order value – each a multiple of what the middle performers did. So clearly, even the mid-level performance could be doing an awful lot better than they are.

I mentioned that half of retailers have a negative return on investment. What we found was that the best performers averaged a 68% return on their investment performance metrics, and that investment amounted to .9% of their sales. Their ROI came from both improved sales and gross margin numbers. The mid-level performers achieved a 12% return on investment, but the impact on the business was even less given that their investment level was a third of the best performers.

ATG: All right, so for these 80% of retailers that still have a lot of upside on the table – How do they turn the corner? Sounds like analysts who know how to do this in retail aren’t exactly growing on trees.

FREEMAN EVANS: Well, for the most part, retailers give the job to a single mid-level or junior person; we found an average of 1.3 workers dedicated to metrics. And even though it’s the right thing to do to have a person devoted to analytics, these people tend to be report generators. And while generating reports is necessary, what really makes a difference is getting a senior person involved, who’s there saying, “Here is this problem, here’s what it means… Let’s try to think up some ideas of how we can get better results.” A senior level person, a VP or EVP, not a $50,000-a-year person.

Another thing that the best performers do is to bring together cross-functional teams to tackle the problems, with weekly review team meetings focused on getting results. This is quite critical because no single person knows all the questions to ask. So having a cross-functional team meeting – or better yet, a cross-channel cross-functional team meeting – will get people to start asking questions that can really impact the business.

For example, something like raising the issue “I don’t think that our free shipping offers are producing enough incremental revenue to pay for the bill.” That’s what Timberland did. In that case, it turned out that they needed free shipping to drive a 40 percent lift on incremental sales to be a worthwhile offer, and they found they were not getting anything like that. So as a result, they not only improved their margins on this particular issue, but they have a better sense of how free shipping impacts shopper behavior, and can use it in a more strategic way to impact the business. And they wouldn’t have done this if they didn’t have regular team meetings with the purpose of identifying areas for improvement, setting out hypotheses in advance, and then using metrics to figure out.

This is not an easy thing to do. Your internal structure, the sensibilities and culture all have to work together to make this happen. If you can’t get that level of commitment internally, it’s very difficult to make inroads. But you can make things work better by bringing external talent to the table. Maybe your commerce platform vendor has a strong analytics group, or your analytics package vendor has a consulting arm. Or even an outside consultant can help the company face up to the tough questions, and help change the culture of the organization.

ATG: Okay, changing the corporate culture takes some time. Are there some simple steps or areas of focus that can make a difference in the short term, while we’re working on the cultural sea change?

FREEMAN EVANS: We identified a couple of them. If you’re looking for the one place to start in terms of using your metrics most effectively, it’s checkout. We’re loathe to touch checkout because it’s such a critical component of the shopping experience, but it actually offers the largest ROI impact. Yet among the weak performers, fewer than half even look at checkout metrics, much less test variables strategically.

In addition to that, look at improving your “add to cart” rate – because obviously, the more that people add things to their carts, the greater likelihood they will check out. Another metric that the top performers are starting to use is their zero search results. Right now only about 30 percent of retailers are actually tracking or using zero search results to make decisions, but these metrics can be such a wealth of knowledge in terms of understanding how the public actually sees your product offering – knowledge you can then use in your search, in your merchandising, in your marketing efforts, and in your Web site copy. It can be tremendously valuable.

So those are all critical things, and each step can help put you on the right track. But the real kicker is to approach these issues not in an ad hoc way, but in a proactive way. And that’s easy to say, but not so easy to do. And ultimately it means breaking down the silos that naturally exist in our organizations, and implementing a cross-functional process that drives continual improvement. Start asking those “What If?” questions, setting hypotheses, and then being much more focused in terms of what metrics you look at and how you’re going to use them. But that’s how you move out of the ad hoc approach.