How to Prove the Monetary Value of Your Self-Service Channels

Self-service channels such as communities, publically-facing knowledge bases and FAQs, were seen in the past as more of a “nice to have” for companies, rather than a “must have.” This had a lot to do with customer preference. Most perceived that they would get their answer fastest just by calling or emailing. This is changing, however, as more and more customers have acclimated to just “Google-ing it” when they have question.

As HP Worldwide Consumer Support Forum Manager, Jason Duncan, told me once, “While it used to be that customers would start on your website and navigate to support, now Google is their homepage.”

Don’t believe me? Check out this recent Forrester report showing a significant rise in self-service and community usage, with customer preference increasing 12 percent and 25 percent, respectively, in just the last three years. Just the same,as a reviewer of this kind technology, I still hear companies all the time questioning the true value of investing in these channels. These systems can get expensive, they require labor to manage, and content must constantly be updated.

So earlier this year, I interviewed nine experts in the space to try and come up with a way to respond to these questions about proving the value of self-service. Finally after two months of swapping emails and going through dozens of iterations, we came up with three data sets that can be used together in an equation.

The following is a description of this formula, which proves the monetary value of the service you provide through your self-service channels, relative to what that same level of support would have cost had the customer called, emailed or chatted with an employee.

Data Set 1: Quantity of Issues Resolved in Communities

The first data set in the equation looks like this:

([A x B]) + C)

Variable “A” should represent the total number of community questions that receive a response per month. Ideally, you’d want 100 percent of questions in your community to get a response, otherwise you risk customers deciding not to use the channel because they are unsure if they’ll get a response.

Variable “B” in this data set stands for the average percent of issues that are responded to by a customer, rather than an employee. In the most successful communities, a majority of replies come from non-employees. This is typically enabled through gamification tactics. A x B will find the total number of issues per month that are resolved by another community user.

Variable “C” should represent the total number of “this article helped me” votes, thumbs up, or three or more star ratings (depending on what system you use to rate content) in the community per month. This finds the number of issues that were resolved by a customer finding their answer in an existing discussion. These votes can also help you identify the most popular topics and discussions so you can make those answers easier to find (through search engine optimization and other community design tactics). If you don’t have much engagement activity, or it isn’t growing, this could be a sign that you need to proactively create content. Start with writing about solutions to the most common issues that come up in the call center.

Data Set 2: Quantity of Issues Resolved in FAQs and Knowledge Base

The second data set in the equation looks like this:

([D + E]) x .10)

Variable “D” in this equation stands for the number of FAQ page views per month, while variable “E” stands for the monthly page views of knowledge base articles. It isn’t reasonable to assume that every page view accounts for an issue that was resolved. Many of these could be people doing research, browsing through several pages before finding an answer, or still eventually calling for an answer.

Our experts each provided a range of page views that are likely from people who did resolve an issue; 10 percent was on the far conservative end, which is why the sum of these two variables is multiplied by “.10.”

Data Set 3: Average Cost to Resolve Issues Through Direct Contact Channels

The second data set in the equation looks like this:

([F + G + H]) /3)

F, G and H represent the average cost per trouble ticket to resolve an issue through phone, email and live chat, respectively. This total is divided by three to find the average.

All Together Now

Altogether, this formula looks like this:

([A x B]) + C) + ([D + E]) x .10) x ([F + G + H]) /3) = i

The “i” variable is the monetary value of the support you provide through your self service channels. If you know the average percent of tickets on average that go through each direct contact channel (and the cost per ticket varies significantly), you can get a more precise value by multiplying the total of Data Sets 1 and 2 by those percentages. Then you would multiply those totals by the cost per ticket for each channel. So basically it would look like this:

([A x B]) + C) + ([D + E]) x .10) x (average % of tickets solved through phone support) x (cost per ticket of phone support) +
([A x B]) + C) + ([D + E]) x .10) x (average % of tickets solved through email support) x (cost per ticket of email support) +
([A x B]) + C) + ([D + E]) x .10) x (average % of tickets solved through chat support) x (cost per ticket of chat support) = i

This formula doesn’t assume in every case that the customer would have contacted you through another channel if they didn’t solve their problem through self-service. It just puts a dollar value on the service you do provide, relative to what it would have cost to solve the same number of issues through more manual channels.

In fact, all of our experts agreed that self-service allows you to connect with a lot of customers that might never have otherwise reached out. When customers don’t resolve their issue, this can impact their satisfaction and loyalty. So this is the final, less tangible value you should consider relative to self-service -- the ability to mitigate the risk of customers not solving their issue at allow.