Finding a Multimillion-Dollar Gap Thanks to Web Analytics

I get asked a lot about how we use web analytics to identify issues specific enough to justify the cost of the toolset. Many people are skeptical when told it will cost them $40,000 yearly to receive reports on how their website is used. What I’m about to explain involves a client who, thanks to the analytics tool we’ve provided, can potentially increase their annual revenue by $1,000,000. This article will explain how we first identified the issue with the client’s website by monitoring a key performance indicator and then how we fixed it.

KPI stands for Key Performance Indicator.

KPIs, huh? It’s the hottest term being bandied about right now. What critical measurements to employ is vital, but how they are used is crucial. I’ve discussed page views per visit as a key performance indicator before. I always consider page views per visit a key performance indicator, regardless of the site’s purpose. This measure serves as a “tripwire” for me, and here’s why. It serves as an early warning system, similar to a tripwire, for when something is amiss. Before using this key performance indicator, you need to determine how many pages are required to accomplish the goal. It took seven pages to get to the result (a buy). Think about what you’re looking for in a pleasant browsing experience. From a business perspective, we deemed it a successful visit if the customer viewed between five and seven pages and ultimately made a purchase (another seven pages). This implies that the customer learns there is more to choose from than just the one item they came for. Then, we tack on another seven pages to trigger the “too many pages” alert. So, the minimum number of pages is 7, the optimal number of pages is 14, and the maximum is 21.

This red flag seems a bit excessive.

Our analysis suggests that if visitors view more than 14 pages throughout their session, they are either delighted with their experience and are exploring the site or irritated because they cannot find what they are looking for. In this scenario, we determined that a positive user experience would be achieved if the typical user spent fewer than 21 seconds browsing the site to locate what they were looking for. Our findings demonstrate the significance of this statistic. The key performance indicator skyrocketed, revealing that visitors typically browsed 22 pages during their session.

The next step was to evaluate whether or not this development was ultimately beneficial. If a typical visit lasted 22 pages, the user was thoroughly exploring the site, in which case our client would be overjoyed, or there was a problem with the site’s navigation.

Is it a good idea or a bad idea? What’s your mood?

The key performance indicator had triggered an alarm, and we wanted to identify which visitors were affected. Visitors can be sorted into groups according to their behaviors in HBX (and many other technologies). If visitors were fast clicking between sites, indicating dissatisfaction, or visiting a large number of pages and staying on the site for a typical amount of time, indicating satisfaction, we needed to know.

Because of this, we divided site visitors into two groups: those who stayed for less than two minutes and those who went to the shopping cart. This would allow us to determine if visitors who spent only a few minutes on the site were viewing many pages during their stay or if it was the cart abandoners having problems navigating the site.

Within two minutes, the conduct was typical. People who visited the site for less than two minutes typically looked at two or three pages. Again, the people who hit the shopping cart went over the edge, but the situation was even more dire this time. A total of 58 pages were viewed during the average session. We located the troubled individuals.

Fifty-eight pages each visit, on average?

Now that we knew which visitors were experiencing issues, we wanted to determine what they were doing wrong. How in the world is the typical user viewing 58 pages? We even had the developer double-check that the tracking code was installed correctly because it seemed odd. However, the website issue was evident once we ran a route analysis.

Only one user had browsed all 97 pages. As we investigated more, we found that every single link in his journey led back to the same place: an error page for search results. When we looked at other, more isolated trips, we saw the same central pattern: the error page for search results.

This prompted us to investigate the website’s unsuccessful search results. When we combined all the unsuccessful keyword searches, we found that almost 2,000 were for product numbers. Most product codes are made up of numbers and letters, but the site’s internal search engine cannot decipher them. The issue had been isolated.

Therefore, the search engine needs to be fixed. This one improvement has enormous implications. More than a thousand people tried to buy but only got as far as entering those faulty keywords. Our client receives about $160,000 monthly from a little more than 1700 customers who made purchases online. That’s why it’s simple to calculate the potential using some basic math. It costs us more than a million dollars a year in missed income.

Summary

The price of a web analytics system is a common cause for concern. They are costly, and most companies already have so much on their plates that it might be challenging to make the most of them. To get the most out of the systems, you need either an in-house specialist to analyze them and identify the issues or to hire a consulting firm. However, failing to take advantage of web analytics is the same as wasting cash.

Steve Jackson (Editor, Author) – Conversion Stories Author and CEO of Aboavista, Steve Jackson, edited The Conversion Chronicles. If you sign up for the Chronicles newsletter, you’ll receive a free electronic copy of his book.

admin