The 1966 hit, “Hold On, I’m Comin’,” was recorded by Sam & Dave and written by Isaac Hayes and David Porter. I know for a fact that they weren’t referring to actually “being on hold.” However, with lyrics like “reach out to me for satisfaction; call my name for quick reaction,” one can start to see the connection.
Contact centers have been in the Hold On, I’m Comin’ business since their inception. Treatment for “hold” goes almost as far back as the phone itself (1876). By the end of the 19th century, there were records indicating operator frustration when attempting to connect a caller and asked to “Hold on while I find her.” So this is not new by any stretch of the imagination.
My mentor, ICMI Founder Gordon McPherson, had this great line: “We’re making a science out of what to do with people on hold.” Though Gordon has been retired for many years, today’s contact centers continue to struggle with this reality. I believe that contact center leaders must move away from making a science out of holding and, instead, use the science of proper staffing to avoid the issue! Alas, that is for another article. Here, we discuss hold options being used and their respective effectiveness.
Vendors are coming up with all manner of “features” to manage hold. One of the early ones was to provide the caller with an estimated hold time. Some early adopters had to grapple with the uptick in abandon rates after the delay message played. Well, HELLO… a percentage of people chose not to wait because the announcement worked. It flushed out the queue!!! Regardless of this very reasonable response, some contact center leaders had the feature removed because they struggled to explain the increase in abandons as being a “good” thing.
If you choose to provide an estimated time to answer, expect that some callers will hang up when the announcement is made; this will register as an abandoned call. If your organization is sensitive to this feature, please do not deploy it.
This feature, also known as “virtual hold,” offers callers the option to exit the queue and receive a callback from the system “without losing your place in line.” Depending on the system, the callback uses the ANI (Automatic Number Identification), the number the call is coming from. Other systems allow a callback number to be entered, or even a specific callback time to be scheduled. The system “promises” that callers won’t lose their place in line, but typically does not offer a response time. If the system is capable of offering a response time, the best configuration is to provide a range of time rather than a specific one.
A time must be selected to offer the callback. Reviewing average time to abandon will provide some insight. The callback option is best offered close to or just before the average time to abandon. Offering the option too early is undesirable.
There are a couple of additional factors to consider. Make sure that the system is configured so that when a callback arrives there is an agent available, or that any delay doesn’t exceed 15 seconds to answer. Otherwise, the caller will be dissatisfied. Certainly most of us would be irritated by a lengthy queue delay after a callback has been established. In addition, it is wise to quiz the vendor on how the callbacks are counted in the Automatic Call Distribution (ACD) data. Does the return call “count” in calls offered? Is it an abandoned call if the caller hangs up after accepting the callback? And is there is anything that resembles a “reach report” showing that of all callbacks made, how many actually were connected to the caller?
A word of caution… if extreme delays are the current state, the callback feature will likely do little to improve satisfaction. When callbacks take hours rather than minutes, larger issues must be evaluated in order to manage both the customer experience and the hold times. And don’t kid yourself into thinking that the callback is a long-term solution to a staffing problem. This feature does nothing to relieve the utilization or “load” on agents. Agents merely deal with a “displaced” load. While the caller may be “holding offline” the agent enjoys no relief in these models. If understaffing is the driver behind offering the callback feature, be certain to keep your eye on burnout indicators. If you don’t, high turnover might be inevitable and cause additional damage.
Bring Back the Busy Signal?
All ACD systems have the ability to send out a busy signal when delays have reached a threshold. For example, if any call has waited 10 minutes, a busy signal can be sent out to new callers until that call has been answered! Years ago, the busy signal was quite a popular option, though today I’m sure there are callers that wouldn’t recognize the “busy” signal.
I may be half-kidding, but I think that the busy signal still has potential for organizations that are so underwater as to be drowning. Busy signals inform the caller that, “We are not available” and essentially say, “Try your call again later.” I see far too many contact centers struggle with really long delays; sending out a busy signal not only informs callers, it saves agents from overutilization and burnout. It also transforms statistics so that there is a kind of “fake” but hopeful feel-good factor. If this feature is ever chosen, be sure to find out how to track busy conditions via your ACD. This, of course, ought to be a very temporary solution until the contact center gets back on track.
Music on Hold and/or Marketing Messages
Music on hold (MOH) is a common method of “entertaining” callers while on hold. Obviously, the objective is to keep them from hanging up! There is an abundance of evidence that MOH keeps callers on hold longer than silence. Wikipedia states that “a CNN survey found that 70% of callers in the U.S. that hold the line in silence hang up within 60 seconds. Meanwhile, research by North American Telecom found that callers hearing music on hold stay on the line 30 seconds longer than callers experiencing silence.”
There are many things to consider when determining music options. First and foremost, music must be licensed to be used for MOH. Professional MOH sources exist and cover all the legal aspects. The selection of music is also critical. The most common MOH selections are instrumentals in the soft jazz genre. It is risky to select anything too intense, loud, offensive or stress-inducing. We want our callers to remain calm and not get worked up because of the music selected. Some companies get creative. For example, Zappos offers a “Joke of the Day,” if you press 1.
Marketing messages are another popular option. These require serious oversight in order to remain current. Offers, services, seasons and holiday greetings have specific timeframes and must be properly administered to provide the right messaging at the right time. Marketing messages may contain a call to action: “Ask your representative about our extended warranty.” If agents haven’t been informed as to the nature of the marketing message, they can’t help sounding and feeling like an idiot. This further contributes to both customer and agent dissatisfaction.
Hold messaging can also be informational. For example, a healthcare provider may have a back-to-school announcement, such as: “Be sure to book an appointment with your provider if your child requires any forms prior to school opening,” or “This office requires a 48-hour notification for prescription refills.” The key to any announcement is to keep it current, keep the frontline informed of any action calls, and limit the looping. After five minutes of any message, the system must be programmed to “go to music.” No one wants to hear any more self-promotion when delays exceed five minutes!
Music, marketing and informational announcements are all external sources to the ACD. The ACD has its own internal messaging capability. When both are “on” they compete and are often really irritating to the caller. When using an external MOH source, there is NO REASON to program the ACD to interrupt with messages like: “Our agents are continuing to assist other callers; please stay on the line for the next available agent.” In fact, some of the marketing/informational recordings include the “Please stay on the line” script. So WHY is this condition so common? Chalk it up to negligence mostly, since there is no good reason to use both messaging sources.
Often during system installation, a technician or IT/Telecom person records a system message (and not always in the most professional manner). So carefully crafted and paid-for messaging is interrupted by basement recordings that repeat the obvious. Callers actually KNOW they are on hold and don’t need repeated instructions to continue to hold. Some messages are set to repeat every 20 to 30 seconds, and it is unbelievably annoying.
If you do nothing else this year, take a long, hard look at how your ACD queue announcements are programmed. Turn them off if you have invested in professional messaging! If only MOH is used, make any announcement interval a minimum of 60 seconds. Once the caller has been on hold for five minutes, program the system to “go to music.” While MOH has proven to impact abandon positively, repeating a hold message every 20 seconds will almost certainly increase abandons. People don’t need to be reminded constantly that they are on hold.
Solve at the Process Level First
Keep in mind that the best solution is to understand what causes excessive hold conditions and solve at the process level first. This will relieve both caller delay and agent load. Improved processes may also improve digital access and self-service as well as reduce demand on agents. And… maybe consider the hit, “Hold On, I’m Comin’” in your music on hold rotation!