Groundhog Day. Occurring every February 2nd, it’s a tradition-turned-holiday celebrated in the United States and Canada. It draws its history from the Pennsylvania Dutch superstition (derived from a German tradition): if a groundhog sees its shadow on a sunlit day when emerging from its burrow, it will return to its den and winter will continue for six more weeks; if cloudiness means the groundhog does not see its shadow, spring will arrive early. Such information might provide some hope (if the winter has been harsh) as well as set expectations for the weeks that follow. Today, much pageantry and reporting follows the inconsistent prognostications of not one but six groundhogs across North America.
As Wikipedia is quick to point out, while the custom remains popular in modern times, studies have found no consistent correlation between a groundhog seeing its shadow or not and the subsequent weather afterward. For this year’s Groundhog Day, the National Oceanic and Atmosphere Administration (NOAA) got in on the fun, tweeting “Groundhogs aren’t the best source of climate information—we are!” They went on to point out that the popular groundhog Punxsutawney Phil has only been right forty percent of the time in the past ten years. With a fifty-fifty chance of predicting correctly, Phil has not delivered great forecasts.
Now imagine if you tried to run customer service on data that was only forty percent accurate. Just as NOAA is a better predictor of weather, there are better and more diverse sources of data, greater insights to be derived, and the means of acting on them in customer service–and improving service accordingly. With the right customer service platform, this is easy; you just need to leave the burrow and take action.
Service quality
There are many ways to measure service conditions. Many are objective and based on operations in the contact center. All can be measured with analytics embedded in the customer service platform, delivering real-time information. These include:
- Channel hold, wait, and/or response times (e.g. telephone, chat, or email)
- First contact resolution
- Mean time to resolution
Other measures are more subjective, yet are equally if not more important. Customer surveys such as CSAT (short for customer satisfaction) and NPS (Net Promoter Score) are two popular methods. Sending surveys to customers’ channel-of-choice (e.g. text message or email) permits fast, easy responses and the data to measure their viewpoint.
Important decisions and actions are driven by analyzing both objective and subjective inputs. This includes adding resources, improving agent knowledge, and further developing customer service and other soft skills.
Channel volume
Another operational condition to monitor is channel volume. This isn’t limited to merely identifying how many calls, emails, and chats are coming in to live engagement channels. How that volume changes by day of week and time of day (as well as other seasonal factors like holidays and new product or service releases) must also be considered. From this information, the necessary staffing levels can be predicted and agents scheduled (or hired) appropriately.
It’s not always a simple matter of moving agents among shifting channel volumes, though. Different skills are necessary to assist customers by voice vs. email vs. chat. Identifying changing trends in channel usage can provide an early indicator of how current agents’ skills may need to be developed or new agents need to be on-boarded.
CSAT and NPS comments may also affect channel decisions. Customers may be seeking engagement options that don’t currently exist that should be considered.
Issue trends
Customers contact customer service because they have a problem. It could be something minor such as an unclear instruction in the manual, or something major like a product missing parts. Big or small, these issues must be tracked because they signal a failure in customer experience.
With the nature of the issue and the corresponding volume, customer service can work with the department responsible to address the issue. This is important for two reasons. First, it improves the customer experience for future customers. Second, it eliminates the ongoing cost associated with responding to customers with the issue.
Self-service opportunities
Besides giving insights into customers’ issues, issue trends help identify where short- and long-term self-service options could assist.
In the example of an unclear instruction manual, the long-term fix would be to update and reprint the guide. While that will help future customers, current customers need a solution now. The easy fix would be a knowledge base article providing the corrected steps; over time the article would be retired as the correct manual became the norm. (And the decision to remove the article would be driven by self-service analytics indicating when the article was no longer viewed.)
Self-service can also speed common customer inquiries found in issue trends that aren’t necessarily a product issue. Customers might frequently contact customer service to register their product. A simple form on the website could capture the necessary details and use workflow to send the information to the appropriate department–a faster process that is not only easier for the customer but also completely bypasses customer service.
Agent performance
Agents are without a doubt the most important component of customer service. Their knowledge and skill not only address frontline customer issues, but they help identify the underlying issues.
Many metrics and key performance indicators exist to measure agent performance. Raw productivity can be measured: calls, emails, or chats completed or cases closed. Knowledge can be determined from first contact resolution. Soft skills can be assessed using CSAT results from their customers. And of course all these measurements can be compared to similarly-tasked or tenured agents as well as the department as a whole.
More accurate than NOAA or a groundhog
At over one hundred thirty-three years old, Groundhog Day and its festivities have stood the test of time. Despite a poor accuracy rate, it’s a fun if imprecise means of foretelling what’s to come.
It’s not that easy in customer service. It will fail quickly if relying on erroneous predictions. Using the right customer service platform, data and analytics with the ability to take actions exist to improve both customer service and experience.
