SoftwareAdvice.com, a technology evaluation firm for customer support software, conducted a survey of 170 managers and found that many of the most used metrics for self-service are process metrics that have very little to do with customer satisfaction.

Web Self-Service (WSS) should be your first line of defense

Almost everyone would rather self-service than call an 800 number. This is especially true for millennial customers. However, if they cannot find what they want quickly, they will call. The keys to success are easily found tools that take you to the right answer within two clicks. Two great examples are:

• a highly visible (top of home page and every other page) link to a website map or FAQ listing that get you to a cluster of hyperlinks, which answer your exact question in no more than one more click, and
• a search or virtual agent function that allows you to put in a phrase with enough detail that takes you to exactly the right answer

What turns customers off are huge lists of FAQs and search functions that either return 20 documents that must be read through or, worse, return “no records found”. One bad experience results in the customer never using WSS again.

Three quarters of respondents said that WSS improved cost effectiveness metrics such as speed of answer, cost per contact, resolution rate and abandonment rates. About one quarter of service managers said these metrics were worse, but one would expect that decline because, as simple questions are siphoned off by self-service, the cost and difficulty of the remaining contacts will go up on a per contact basis.

What are the most prevalently used metrics by service managers?

For evaluating web self-service, which included, search, virtual assistants, support pages, FAQs, virtual agents and knowledge bases, the most prevalent metrics were page views and user ratings. The first three of these metrics are a waste of time at best and can be very mis-leading at worst.

• Page views basically tells you where the customer was sent, not whether the customer found what they wanted.
• Time on page may actually be an indicator of worse service not better. It may mean that the customer spent a lot of time studying the page to determine that it was not helpful.
• Return visitors could easily mean you tried the solution and then had to return to try to find out why the solution did not work.
• My only surprise with User Ratings is that 15% of service managers did not use them.

Outcome Ratings Are Most Useful and, In Website Settings, Immediately Available

Three types of outcome metrics are useful.

• User ratings based on a yes-no question – did this answer your question? A problem with this metric is that the answer could partially respond or might be confusing, so the customer response is forced due to lack of a middle category.
• User ratings based on a more granular question – how satisfied are you with the information provided is a much more useful metric but even then you often should ask why the rating was given.
• For search and virtual agent functions – two telling metrics are the percentage of time no record or answer is found (I get this about 30 percent of the time from search functions) and the percentage of time the function is simply abandoned.

Avoid Common Actions That Enrage Customers

CCMC’s National Rage Study has identified several common actions taken to promote WSS that pour gasoline on the customer fire – DO NOT DO THEM!

• Put a message on hold that, for faster service, go to the website. Customers most likely have already been there and did not find the answer.
• The website does not provide an 800 number or “call me” button on every page forcing you to go look for it.
• Provide an offer of chat and then talk longer than 30 seconds to initiate the conversation