Mike Premo, ARC’s President, gave a presentation during the CTDA conference in Lake Las Vegas last week. Mike showed us how ARC has turned to the Satmetrix Net Promoter Score to measure their performance. It turns out ARC’s score is a 70 – which basically means a lot of their customers are raving fans. See more about other companies’ scores here.

The Satmetrix Net Promoter Score is gaining traction as a way for retailers to measure key performance indicators that lead to increased profitability. I like it because it’s simple and offers powerful insights about your company’s performance. Keeping in mind the profit equation is profit = revenue – costs, then NPS can help.

Let’s look at the online retail world where site bugs can have an outsized impact on NPS, but are difficult to quantify in other ways. Bugs lower revenue. When they affect a Website’s shopping or check-out paths their impact will show up initially as lower conversion rates and higher bounce rates. Fatal errors and other bugs that annoy customers to the point they shop elsewhere will also reduce visitor numbers to your site over time. Since it’s difficult to measure this effect it’s easy to ignore. Online Travel Agencies (OTA’s) face another problem – interactions across different lines of business. Visitors who start their search for a particular product, flights, may end up purchasing a hotel or package (usually flight plus hotel).

For this discussion the site architecture for flights, hotels, packages and cruises is designed as discrete shopping paths with a bespoke checkout process that ties into consolidated back office accounting and reporting systems. Great online marketers design intuitive paths that minimize customer frustration, but bugs can undermine every designers’ best efforts. Customers will tolerate errors in their own way and may forgive slow response times but bounce when an ad flashes at them. Once the customer chooses a product, an error during checkout, or a price jump (we’re talking about travel here) can drive customers away before the transaction is complete and these problems add up. How much do bugs cost?

Most Web sites track numerous statistics and site performance metrics to uncover friction points or areas to improve, but these reports often miss interactions across paths or temporal changes. I’ll set up an example and describe several assumptions to show you how. For simple math, let’s assume that Hotel and Flights each drive 40% of an Online Travel Agency’s site traffic, while Packages and Cruises account for another 10% apiece. There are many use cases, but most leisure travelers who are flexible about their destination and timing will search for flights fist, and then hotels. This experience suggests that the ratio of flight searches to completed checkouts is higher than it will be for Hotels. Moreover, once a customer settles on a specific flight, they’re able to begin shopping for a hotel. But what if the customer experienced a major bug during checkout? They’ll leave the site and never become a hotel shopper. In this case the poor experience with the flight path caused the customer to search another OTA and they never showed up as a visitor in the hotel path.

Bugs are typically rated by their frequency, severity and location in the purchase path. Frequent, severe bugs that occur during payment have the greatest effect on financial performance since the customer was extremely likely to complete their transaction. Digging deeper you’ll find a second layer that’s often missed too, since the customer has already invested a lot of effort in the process, they are more likely to bolt later in the path to avoid finding the same bug the next time around. Sharp marketers must estimate this behavior to gain better visibility over a particular bug’s impact. Managers who identify a bug that occurs during 1% of visits when typical conversion is 100% would incorrectly assume that the bug will lower conversion by the same amount. That’s wrong from the start if, as most travel sites know, buyers visit to shop many times before they make their purchase, so errors in early parts of the path that frustrate visitors and interfere with shopping will drive shoppers away and site visitors will decline over time, thus conversion rates may decline by 1%, but shoppers may decline 10% too, which would compound the losses and be indistinguishable from other problems.You might spot this by tracking changes in the ratio of new visitors to returning visitors but this metric is affected by new browser and device releases and doesn’t provide the detail you need.

In the previous example as one line of business drives a customer away permanently, than visitor numbers in the other lines of business will experience a steady decline. Once again the bug’s effects are invisible in the conversion rates throughout the checkout path, and it will be unclear why direct visitors and returning visitors have declined. The OTA will need to spend more on paid search to drive ever more new customers to their leaky bucket.

This thought exercise demonstrates that it’s important to measure bugs across the enterprise and the Site’s 1-N list should be discussed widely and in the context of corporate strategies about staffing, and marketing spend. The Net Promoter Score can rescue Travel Agencies from internal bug lists and give you actionable intelligence about the ways you might be preventing sales. NPS is independent from the hidden correlations and mountains of data that overwhelm online marketers and site-health professionals each day. Fix the biggest problems and watch your revenues and profit climb.

For more information about these topics check out Avinash Kaushik on twitter @avinash Author, Web Analytics 2.0 & Web Analytics: An Hour A Day | Digital Marketing Evangelist, Google | Co-Founder, Market Motive; Bryan Eisenberg @TheGrok, Marketing Optimization & use the Data expert (small or big data), keynote speaker & New York Times best selling author. Austin, TX  bryaneisenberg.com; and great dashboard ideas: Juice, Inc. @JuiceAnalytics “We craft applications that help people understand and act on data.” Reston, VA · juiceanalytics.com