WebTrends seminar series features case studies on how to use the numbers to move the needle.
As any good seminar should, WebTrends' "Moving from Metrics to Results" series sells the need for the company's solution as much as the need for good web analytics itself.
I attended the event in Orange County, California on Tuesday. The company has been on the road with the seminar since early April, taking it from New York to Dallas, as well as internationally.
Much of the first half of the session was spent defining terms such as hits, page views, visits, visitors and sessionization, and describing various methods of tracking these things, ranging from IP address (most popular) to persistent cookies (best approach).
Barry Parshall, director of product management at WebTrends, did a nice job of describing the difference between data collection methods, and providing the advantages and disadvantages of each. In a nutshell, log files (IP addresses), he said, are inaccurate for capturing traffic from real people, but imperative for tracking spider traffic, i.e. visitors coming through search. On the other hand, client-side data collection -- a.k.a. tagging, java script tags, pixel technology, and page beacons -- is far more accurate for tracking visitor behavior, but filters out spider traffic automatically. The best choice, therefore, is to use both.
"Cookies are still the best technique for tracking," said Parshall, but he admitted that unique user counts will be inflated with consumers deleting cookies. He went on to define the difference between first-party and third-party cookies, and explained how spyware programs are designed to delete third-party cookies -- including those placed by WebTrends. The company is addressing this, he said, with a new version coming out next month, which essentially will be convert cookies to first-party.
After the break, Jeff Seacrist, director of product marketing, gave a product demo, and then presented "Mastering the 7 Key Areas of Your Web Initiatives." The key takeaway: analytics in and of themselves are nearly useless (my words, not his); it's how you use them to optimize results that matters.
Okay, so that's kind of a no-brainer, but worth exploring nonetheless.
Here's why. According to a CMO Council report, 90 percent of marketing executives believe measuring marketing performance is a top priority, yet less than 20 percent have developed meaningful, comprehensive metrics for their marketing organizations. What's more, Forrester reports that 53 percent of marketers cite acting on the findings of analytics as the most difficult aspect of the measurement process. So despite the fact that everyone knows the reason for collecting data is to improve performance, more than half of the people gathering information struggle with how to use it.
Part of the problem is with the information they gather. Jupiter Research reports that 66 percent of marketers measure clickthrough rate to determine campaign ROI. But "on the web, conversion is king," said Seacrist, using conversion expert Bryan Eisenberg's definition: "Conversion is your site's ability to persuade your visitors to take a desired action on your site."
So the reason for analytics is to understand how consumers are getting from point A to the point of conversion -- whether that be a purchase, request for more information (lead), to resolve a customer service issue or to interact with a brand -- and then to shorter that process.
For example, if a large number of customers are leaving your site during the purchase process, figure out where they're leaving and why. Are there too many steps for them to go through? Consider combining a couple of them. Are they becoming distracted? Eliminate some content. Are they getting lost in your navigation? Figure out how to redirect them.
In the case of Motorcycle Superstore, WebTrends found that two percent of people leaving the product page went to on-site search. In analyzing search terms, they found that the number one search word was helmets, which generated more than $5,000 in helmet purchases during that particular tracking period. The number three search word was helmet, which, unfortunately, sent searchers to helmet accessories rather than actual helmets, resulting in just over $1,000 in sales -- an opportunity for improvement.
Speaking of search, WebTrends suggests marketers invest in both paid and organic search. Here's an example of why. In an analysis of organic versus paid search response, Motorcycle Superstore found it generated 80 percent to 90 percent of response from organic search, resulting in purchases of more than $115,000. The company's sponsored links produced just under $18,000 in revenue. Had the company not tracked it, it might have assumed its second-spot sponsored link was all it needed.
Another important metric when analyzing search: change of positions and revenues. The number one position doesn't always produce the best results, but in most cases will cost more.
WebTrends explained that cost-to-reach versus revenue-earned applies in other cases as well. For example, according to a Bain & Co. study published in the Harvard Business Review, it's six to seven times more expensive to gain a new customer than it is to retain an existing customer. Thus, a five percent increase in customer retention can increase profits by 25 to 95 percent.
Motorcycle Superstore tracked results of its new buyers versus repeat buyers. The company found that new buyers were purchasing more products, but spending less per order. The solution: feature lower-priced merchandise on the home page -- where new buyers were most likely to find the company's products -- and higher end products in retention newsletters.
Another case study WebTrends shared was results that T-Mobile gained from changes made to its customer service elements online. A Gartner Group study shows that the web can be 20-plus times less expensive at solving customer problems than email, phone or text chat. T-Mobile redesigned its site, making online self-service features more visible and saw a 67 percent growth in traffic to self-service features. The increase in traffic resulted in significant cost savings.
Through analysis, T-Mobile also learned that 17 to 24 percent of visitors were abandoning at the step of "selecting location" when researching phone plans. Via A/B testing, the company optimized the scenario and increased online orders by 27 percent.
In sum, Seacrist reminded the audience that it's all about turning your analysis into action, and that optimization is an ongoing process with clear steps:
- Clearly define your site objectives
- Establish metrics that align to those objectives
- Leverage the analysis to take action, which includes testing different factors and optimizing on best performers
- Repeat again and again.