Will utilities be able to fit their Contact Centers in "phone booths"?
- Nov 7, 2016 8:00 pm GMT
- 2410 views
Co-Author: Andy McKenna
One goal of most contact centers is to minimize live voice communication. Utility customer service is presently delivered through a combination of self-service and voice channels, with the former gradually increasing their share. The speed of that change depends on technology adoption rates and conscious utility decisions to transition customers to use self-service channels. The use of self-service channels reduces expenses and is consistent with general business trends of such service delivery by other providers.
Some optimistic projections have pointed to near-elimination of live call volume. However, due to several factors, we do not foresee this happening in the next few years. Even after aggressive automation and process changes, a utility will still have to deal with about 0.4 live calls per customer per year or 400K calls per 1 million customers in the near future. We base our estimates on customer demographics and point to (overlapping) populations that will still need personalized help to obtain assistance to satisfy their needs.
Most utilities handle between 1.1 and 1.4 live calls per customer per year—or 1.1M to 1.4M live calls per 1 million customers. Typical live annual call distribution is split as follows:
- Payment arrangements: 30% to 40%
- Payments: 5% to 15% (depends if real time payment processing is available)
- Billing: 25% to 35% (depends on the number of exceptions, billing issues)
- Start, stop, move service: 15% to 20%
- Energy efficiency, outages, other: 5% to 10%
This distribution of call types is the reason for the barrier to get call volume to “zero” with current customer demographics, their needs and habits. The above distribution depends on a number of factors—stringency of collections rules, website service capability excellence, IVR capabilities and upstream process quality. At a typical utility, 40% of customers never call during a year—representing “passive” customers. The remaining 60% of customers call on average about 2.5 times a year— defined as “active” customers; practically, some call at least once; there is a subset of customers who call progressively multiple times. On a typical day in a North American utility Contact Center, between 10% and 12% of customers called at least twice—pointing to First Call Resolution challenges.
With the current volume of between 1.1 and 1.4 live calls per customer per year, we suggest that 0.4 live calls per customer per year will continue—meaning about 30% of live calls; the majority being multiple calls from the same customers. Repeat callers will be most likely calling about the following issues:
- Payment arrangements – repeat calls (need to negotiate delayed payments)
- Billing – possibly repeat calls (issues with bills, understanding)
- Start, stop, move – depends on move calls (cannot use automation)
We estimate that these repeat calls will come from about 20% of customers who need “manual” assistance. We consider this to be a reasonable estimate based on these overlapping customer demographic groupings (percentages in parentheses refer to US population):
- Poverty (28%) – using 200% poverty level, these are customers who are most likely to call requesting payment arrangements; inability to make timely payments is the most frequent driver for repeat calls (we estimate that nationwide that many households fall under this category)
- Retirement savings (14%) – according to various surveys, about 14% of individuals 65 and older, do not have any retirement savings (with unexpected expenses, they have difficulty making payments including utilities and call to get extensions)
- Adult illiteracy (14%) – according to the government statistics, 14% of customers fall under this category (and some overlap with those who are below poverty rate); they will need manual assistance
- General population (5%-10%) – do not fall under above categories, but will still be calling to complete transactions for the reason of convenience or unfamiliarity with self-service options (cryptic error message on the website, inability to be services by IVR)
We take all these overlapping customer groupings and make a generalized assumption that they represent as much as 20% of the total customer base (not just “active”). We also assume that each of these customers on average will call twice in a year—which represents 30% of the current call volume.
These customers need more support and maintenance to assist with payment arrangements and protocols that are outside of utility rules and cannot be completed by self-service. They also need help understanding and paying their bills.
Additionally, on a 10 point scale, the level of difficulty of these remaining calls will be perhaps “7-8” on average, while it is about “3-4” for current call distribution (remember, there is a “long tail” of easy calls in a typical Contact Center, which is why “5” is not a median).
The table below lists three markers associated with transitioning towards predicted minimum call volume.
To reduce call volume to the lowest limit today would require key processes re-design and to draw a “red line” on repeat collection calls (risky—possible Commission intervention), implement real time payment posting, create excellent upstream processes ( meter reading and field services) and create an amazingly intuitive and feature-loaded website including mobile interface. These changes might get the call volume down to 0.4 live calls per customer per year—with today’s customers’ cultural habits—but note that this is achievable through radical, drastic and expensive actions.
Self-service automation will reduce call volume with customers who want and can play by the rules, but will do little with customers who need to work something outside of the rules, and need to speak with a human being because they cannot handle all that automation. The word of advice is to plan for an extended and slow live call volume glide path and for upgraded training for remaining agents as they will be dealing mostly with complex calls. As such, nearly all of these agents will be required to have advanced skills.