Customer self-service ROI: What should you measure and how?

How to best monitor and measure your customer self-service programmes. 

As with any initiative, in order to manage your self-service project, you will need to be monitoring and measuring it. And in particular, focus will need to be placed on the usual suspects – performance, costs and return on investment.

In this feature, we’re going to take a look at some of the key figures you should be keeping an eye on, and dish out some advice on how to best monitor and measure your self-service initiative. 

Performance monitoring

When it comes to ongoing monitoring of the performance of a self-service initiative, there are a variety of different dials to keep track of.

The Inner Circle Guide to Self-Service, for instance, recommends that effectiveness can be successfully tracked by comparing the number of self-service requests that are logged with the number of times that customers request that the self-service session is escalated to agent assistance.

The report notes: “If analysis of this ratio can be done at a granular level - for example, by product type, customer profile or other easily identified group - it will be possible to see which parts of the self-service solution work most effectively, and which need fine tuning, or indeed a complete overhaul. Understanding how knowledge content is accessed is also vital information, as it may be that certain information is requested far more often from mobile devices then PCs, meaning that businesses may want to reconsider the presentation of such information to customers, for example.”

Aphrodite Brinsmead, senior analyst for customer engagement at Ovum, believes particular attention should also be paid to resolution and customer effort, in order to determine the effectiveness of self-service.

“Businesses need to think about how easy it is for customers to complete certain tasks – whether it is a simple interaction such as checking an account balance or resetting a password, or something more complicated like understanding a fee on their bill," she explains.

"To track task completion rates, businesses need review customers behaviour across different channels but this can be challenging because customers often interact in a personalised way with many channels. Businesses should encourage customers to provide a unique ID, such as an email addresses or account number, when they engage with web chat or a virtual agent. When using a website, customers should login, so that businesses have insight into their behaviour and challenges. That way, when the customer reaches an agent, the business already has insight about the customer’s historical interactions and actions. By linking data and measuring task completion rates, businesses can improve pain points for customers and ensure smooth interactions with high first contact resolution rates.” 

Best practices

For the purpose of best practice, Michael Maoz, research vice president and distinguished analyst at Gartner Research, summarises the key areas your business should be monitoring as a matter of course to keep track of self-service performance and effectiveness:  

• Track each person who has attempted self-service, and ensure that this is extended to mobile. “There are software companies out there that allow you to track customers across all of their mobile applications, so if they start on your mobile app, then went to your site, or out to a competitor’s site, they can monitor that, as well as the functions they used or failed to use on their laptop or mobile device,” says Maoz.
• Measure completion rates of self-service.
• Measure escalations.
• Detail escalation paths. For instance, a path may be from Google, to a community, to a company website, to a live chat, to an IVR and then to an agent. You should map the most prominently-used paths and pinpoint the paths/destinations you don’t want customers to take. By having this picture, you can change the roadmap and get them to use self-service by finding out why they aren’t using it at present.
• Breakdown factors driving Net Promoter scores (NPS). Self-service is not just to save money, but it should also be driving NPS up. If it’s not doing this, then you’re doing something wrong.
• Measure and publicise satisfaction with each channel, and overall. This is important internally and externally so that people can see that your organisation is going a great job – customers want self-service and you delivered it.

Of course, just because a self-service initiative is performing well for the customer, it does not mean that it is cost-effective. One of the main purposes and benefits of self-service is to cut costs for the business, and a self-service session should be far cheaper than any interaction type that requires a live agent input. But because there are related expenses, such as ongoing maintenance and administration of self-service communities, costs still need to be closely monitored. All of the related costs need to be detailed to get a clear picture of how much is being spent on self-service.

As the Inner Circle Guide to Self-Service notes: “When considering the cost per interaction of a web self-service request, such as searching an FAQ database, incremental cost becomes negligible. However, the time and effort - both from an initial and ongoing basis - to keep any knowledge base up-to-date and accurate, is anything other than negligible.”

However, it adds: “With the proliferation of channels and devices, this effort can be seen to be shared over all of these new channels, as much, if not all of the knowledge, is accessible by customers and agents regardless of channel or device.”

Return on investment

Nonetheless, done correctly, the self-service initiative should be realising more cost savings than additional costs. Indeed, when calculating return on investment, the main component is traditionally the cost reductions. This usually takes the form of cost savings due to call avoidance.

The simplest metric to monitor in this sense is the number of live contacts that are no longer made because the queries are being dealt with via self-service.

The Inner Circle Guide to Self-Service highlights: “If the knowledge base and self-service application is available to agents as well, there is also the opportunity to improve first contact resolution rates, and to decrease average handle time in the cases of those queries which were not handled by online self-service (for example, the customer may not have attempted to self-service before picking up the phone), but which the knowledge base has effectively answered for the agent.”

Anand Subramaniam, VP of worldwide marketing, eGain Corporation, adds: “Self-service ROI can be measured by operational metrics such as deflection rate of agent-assisted service (phone, email, chat and branch-office, for example), deflection rate of unwarranted field service and a percentage increase in queries handled without staff expansion.  These metrics can then be translated to cost savings.”

Finally, soft metrics such as improved customer satisfaction, website stickiness and improvement in Net Promoter Score are also important, and should be considered.

“Customer satisfaction with usability of web self-service is important," says Brinsmead. "Often the easiest way to understand that is via a survey following the interaction, though ideally this will be sent in a non-intrusive way.”

With an eye on all of these dials, businesses can ensure they have a clear picture of performance and ROI as they forge ahead with their self-service initiative. And as self-service technology continues to mature, many are optimistic that returns are set to grow in this field in the future.

“If you consider the wages of agents in call centres, they are clearly more expensive than a software programme that can be scaled to enormous amounts of interactions with people,” says Steven van Belleghem, professor at Vlerick Business School and author of ‘The Conversation Company’. “If you look at the list of jobs that are disappearing due to automation over the next five to ten years, number one on the list is call centre agents. The technology is evolving so fast that in a short time we won’t hear the difference between talking to a computer and talking to a real person. We’re getting very close to the point where for basic routine questions and complaints, you can just replace people with a robot. And at that point it is much cheaper. So a lot of companies are waiting to do that.”