Customer retention is paramount to successful businesses, and the importance of customer loyalty is further heightened in our current economic situation. In this article, Cisco IronPort Customer Support will share how they gage their customer’s experience through customer satisfaction (CSAT) surveys and show how the results become actionable items that enabled them to evolve their CSAT Program, implement continual improvement, and retain customers by exceeding their expectations. 

We Rule! We think… Our customers are ecstatic with customer support. We brag… A lot. How do we know? We have a 90-plus percent renewal rate, we receive kudos from the field, and we handle a small percentage of escalations. All great input, though a business thriving in an entrepreneurial environment of continual improvement, this was not sufficient proof to show us where to focus. Of course we’re not perfect…but then it suddenly dawned on us…we were sitting behind the 8 ball!

Implementing a transactional CSAT survey for customer support tickets became a top priority very quickly, which was prompted by nothing more than the need to confirm our beliefs by none other than—you guessed it—our customers. Collectively the small Cisco IronPort team, a recent acquisition and independent support business unit of Cisco Systems, pulled together a rudimentary survey, which was soon followed by a sophisticated survey solution that continues to evolve to this day.

Our first step was to begin accumulating data, which consists of ratings and feedback directly from our customers who open a support ticket with us. The first month provided an acceptable participation rate of 9.1 percent and an average CSAT score of 4.38. Given the feedback we had received, this number was lower than our expectations, but at least at this point we now know what our customers think and can take action.

Data Data Everywhere
We were very excited to obtain this type of customer feedback. And so the analysis began. After the first month of data collection, the primary areas of focus around support resolution were obvious. Both scores and comment analysis indicated a primary trend that tickets were being closed too soon according to the customer. Armed with this data, the global support manager launched the “two-strike rule”—a new best practice, which raised CSAT 1.4 percentage points in just one month.

The following month, actionable items were less obvious, and we started to drown in data. Requests to slice and dice data every which way were coming from every angle. The enthusiasm was fantastic, but overwhelming, and what we realized later in hindsight was that we lost focus. There were tons of slides with tons of numbers, but no actionable items. We resurfaced in month three realizing that we missed a period to take a step forward. Now it was time to pull our head out of the weeds and take a new look.

Of course, some analysis leads to dead ends but is worth assessing to ensure that opportunities are not missed. One such exercise consisted of calculating the median. Although our median was 5 on a 1-5 scale, which was great news to share, no actionable items could be derived from this. Likewise, while overanalyzing the data and finding nuances enabled us to address a few targeted issues, such as varying RMA distribution globally, it did not improve our overall score. At this point, we needed a more systematic approach to data analysis and a keen focus on taking actionable items, communicating them to the staff, and measuring results.

Looking Beyond the Numbers
Taking a macro view proved invaluable once we set a baseline analysis outside of the numbers. Over the course of the following two months, we took a new approach to looking at the survey data and saw a whole new perspective, including the rating distribution and the identification of an ambiguous question in our survey.

The first tactic branched from the median derivation. If our median is 5, then we’re doing really well, so why is our CSAT low? We broke the ratings into high (4-5), medium (3), and low (1-2) scoring; then we looked at the percentage distribution, and didn’t stop there. There were no consistent trends in low scores. The distribution was at an acceptable +/-3 percent. The reality is that there will always be circumstances leading to low scores; therefore, as long as the percentage is acceptable and there is an established process to manage them, it is not worth digging further into this area. So we moved to the next tier. Questions rated a 3 out of 5 were over 6 percent. There is nothing egregiously wrong with this distribution, but by moving just half of these ratings from a 3 to a 4 would have increased our average CSAT from 4.49 to 4.52! This is an easy message to communicate to the support team: “Focus on how you can provide a support experience that is good as opposed to just average.” A month later, 3s dropped by 1.3 percent; not half, but a respectable decrease, especially when low-score distribution also decreased. This small but powerful move increased our CSAT from 4.49 to 4.54 in one month.

Although the survey had been running for only four months, we already had two questions that were consistently lower than the rest. Again, instead of dissecting the data to death, we viewed these from a macro perspective and looked for trends. By selecting to analyze our “Effectiveness of Resolution” question, we realized that customers were using the question to rate either their experience with the customer support engineer or their experience with the actual response provided. This quickly brought to light that some scores for our engineers were being artificially lowered when bugs were identified and workarounds were provided. We created a new question to specify a response specific to the engineer separately from a response regarding the actual resolution. In order to keep monthly and quarterly data relativity, we will be implementing this change next quarter and expect the results to provide more accurate engineer ratings. It also should enable us to more easily identify problematic bugs and products and communicate back to the product support engineering team real data to complete the 360 internal feedback loop, yet another CSAT-related initiative that keeps communication flowing throughout the company.

Raising the Bar
The incremental improvements in our CSAT strategy have increased our ratings from 4.38 to 4.52 in just four months. This accredited to several key factors. First and foremost, we hire the best support staff and set them up for success. If an engineer thrives and excels at handling phone calls, then their primary role is to provide phone support. Another key factor is the flexibility and support to roll out incremental improvements and the ability to continually evolve the program. We never would have come up with a perfect survey from day one, even if we have given it 10 times more thought. It was through launch-and-learn that we have improved, and continue to evolve the survey, the process, and the analysis. Tied closely to this approach is also focusing on the whole picture. Don’t get tied to the CSAT number; look at where improvements can be made, and make it a collaborative effort. 

Conclusion
Cisco IronPort Customer Support is now on a learning curve with CSAT and is enjoying the learning experience, but by no means did this come to us easily. We learned from Cisco Customer Support, information on the Internet and as explained, and analysis from different perspectives. The customer voice is very important, but also even with a strong participation rate of 20 percent, there is a silent 80 percent of our customer base who we want to hear from.

Jennifer Klaflin will present a key session on this important topic at Technology Services World 2009, being held May 4-6 in Silicon Valley.  Please join Jennifer at her “Cisco IronPort Customer Support CSAT” session scheduled for Tuesday morning. Learn more about what Cisco IronPort Customer Support has done with CSAT, and what is on our roadmap for the future. See you there!

About the Author…

Jennifer Klafin is a program manager at IronPort, now a part of Cisco. She leads the Customer Satisfaction (CSAT) Program for Cisco IronPort Customer Support. She joined IronPort in 2004 as a project manager and has been with the company through exponential growth and their acquisition by Cisco Systems in 2007. Prior to IronPort Jennifer held various project management roles at fast-paced entrepreneurial companies and most recently launched a SLA alert program at EMC. Jennifer also leads the company-wide survey at IronPort and thrives on the collective rewards and collaborative solutions of making CSAT a success story. Jennifer will be presenting on this article topic at the Technology Services World (TSW) conference, being held May 4-6 in Silicon Valley. She may be reached at jklafin@cisco.com. Visit www.cisco.com.