This month, we're featuring the contributions of the 2022 HDI Thought Leaders and Featured Contributors. These leaders in the IT service and support industry regularly take time out to share knowledge to help their peers succeed. Doug Rabold, who makes both lists this year, has long been a pillar in the HDI community, and we always appreciate his contributions.
If – like me – you are or ever have been in a leadership role within a Service Desk or a Customer Contact Center, you know intimately what Service Level Agreements (SLAs) are. They are those commitments – typically contractual in nature – that your service or support center promises to deliver. They encompass some or all of the usual suspects of metrics: First Call Resolution (FCR), First Level Resolution (FLR), Customer Satisfaction (CSAT), Mean Time to Resolve (MTTR), Average Handle Time (AHT), Call Quality (QA), and Ticket Quality (QC).
In fact, most of the things we measure in service and support centers either are or could very well be established as SLAs. Typically, the metrics that have been selected as SLAs represent what is important to ensure that business can be conducted effectively and efficiently. If we cannot meet what is agreed upon, the customer suffers. One could really consider these SLAs to be the minimally viable product of the support team. Failing to deliver results in negative impacts for the entire organization.
The reality is that some of these metrics within SLAs work in tandem with one another while others are in opposition. An example of where they work in tandem is FCR and FLR. The more issues resolved by the initial agent (FCR), the higher the resolution rate will be within the service or support center without escalation to other teams (FLR). In this case, improving FCR also helps FLR.
Other times, two metrics within the SLA may work in opposition. An example is the relationship between FCR and AHT. If the organizational emphasis is on resolving as many issues as possible by the first agent, this means that those calls may take longer. In an organization where FCR is critical, there should be a lot of shift-left issues going from higher level teams down to the desk. This means that higher complexity issues are being handled by the desk. Higher complexity means longer duration. Ideally the MTTR improves because fewer issues need to be escalated outside the desk, but the AHT will increase because frontline agents are now handling issues formerly handled by higher-skilled teams.
This brings us to one of the main challenges with SLAs. Being that they are contractual means they are very seldom (if ever) renegotiated until the contract expires (if then). SLAs are usually static and represent the minimally viable product of the service or support center.
I would assert that there are certain events that should (but rarely do) trigger SLA renegotiation even during the life of a contractual agreement. These may include changes within the desk itself and within the larger organization.
Leaders of most service or support desks are exceptional at delivering against SLAs. Far too often though, it is because of a delicate balancing act that is performed hour-by-hour, day-by-day, week-by-week, and month-by-month. They have armies of team leads, coaches, and managers constantly monitoring performance. They have balanced scorecards developed that monitor performance of each agent, each team, and the entire desk.
In all of this analysis is where gaps manifest themselves. As soon as one of these all-important metrics begins to slip, a full-court press emphasizing quick recovery is initiated. The managers, leads, and coaches blitz the agents with the SLA flavor of the day.
And it works! There is recovery – usually fairly quickly. But no one ever circles back to tell the agents when it is time to return to business as usual, so they continue to focus on that particular metric. As a result of that focus, they lose focus on some other SLA and performance slips there. Now the treadmill begins all over again with the new metric gap.
Which brings us to the customer sentiment. Ask a roomful of desk leaders to raise their hands at various levels of CSAT performance. At 95% you typically still see about one-fourth of the hands in the air. We should commend those people for their team’s excellent performance, but then follow up with, “Please keep your hand in the air if you really believe that is how your customers feel about your team.” At that point most of those hands drop.
This is what is known as the Watermelon Effect. From an SLA perspective, we look like we are doing great! But the word on the street says otherwise. Like a watermelon, everything is green on the outside but it’s bright red when you dig into it.
Which brings us to something that has recently gotten significant traction in our world – Experience Level Agreements (XLAs). These differ from SLAs in several key aspects:
- XLAs supplement or complement, but do not replace, SLAs
- XLAs are dynamic and should be reevaluated on a regular cadence with changes that impact the experience
- Beyond their existence and what will be measured, XLAs should not be contractual
- XLAs can be indices that look at performance holistically to reflect the sentiment of the customer more accurately
The last two years of the pandemic present us the ultimate case study for XLAs. Reflect back to January 2020 when all was right with the world and COVID was still just an unrealized risk that may or may not impact us. Now think about how many experiences changed and how significantly almost overnight.
- The first shoe to drop was a massive, global shortage of hardware (PCs, laptops, monitors, cabling, VPN appliances, etc.).
- Next came the need for new processes to be developed to onboard new employees at home – including setting up their own equipment and remote training.
- Next came a whole new way of collaborating in virtual meetings and the explosion of telepresence applications to replace all those meetings we used to attend.
- The next was the challenge of performance monitoring agents who we no longer had in line of sight since they were sitting in their spare bedroom.
And the hits just kept coming – for months! Process after process had to be revised or developed to adjust to the new mode of operation. We got past each of these hurdles but there was no real agreement on what it would look like. We accomplished all these great things by the seat of our pants!
Everyone in service and support got kudos from our organizational leaders for how we handled the transition to working from home, but do we really know that this was their sentiment and not just another Watermelon Effect?
Had we thought it through in March 2020, we could have established XLAs that would have measured success against each change. We could have implemented means of capturing customer sentiment more effectively and on a more timely basis to ensure that we were executing at the speed of the business. The reality is that what was important changed dramatically for that period of time – the minimally viable product was put on hold – as other priorities came to the forefront. This is the type of scenario where XLAs can truly reveal customer perceptions of the desk. And once we confronted and overcame one hurdle, we could have eliminated associated XLAs and developed one or more to address the next obstacle.
This flexibility to address changing priorities while accurately capturing the customer sentiment is the difference-making aspect of the XLA. Let’s use them properly and remain vigilant that they don’t get weaponized against us.
Doug Rabold is Chief Experience Officer, Founder and Principal Consultant of Bold Ray Consulting in San Antonio, Texas. As an IT Operations Leader he has had direct oversight of over a dozen different ITIL processes. He is currently serving as President of the HDI National Board of Directors and is member of the ITAM Forum Strategic Advisory Board. He previously served two terms on a Customer Advisory Board helping guide development of an industry leading ITSM tool from a customer perspective.