The Joy Of Callback Queuing

I was in a taxi on my way to LaGuardia on a snowy Friday afternoon. A proactive text message informed me that my flight was cancelled (drat!), launching my desperate effort to find a route home. Of course everyone was calling the airline, so the queues were backing up. They offered a callback, so I took the bait. I “sat” in queue while I simultaneously scoured online booking options and nervously watched the clock and the New York City traffic. Impatient, I tried the call center again, and received a brilliant message saying they had me in a better position in queue based on my previous call. Shortly thereafter, the callback came, they rebooked me out of JFK, and the taxi and I changed our path. That was the moment I fell in love with callback queuing, like so many other customers and contact centers.

Why Does Everybody Love It?
First, a confession: I used to really dislike this concept. My mantra: “Nobody calls a center because they want to talk to you later.” I saw centers using—and abusing—the opportunity to understaff and kick their calls down the road. In the era of “customer experience  management,” this seemed like quite a disconnect. But after hearing center leadership, customers, vendors and even friends and family members tout the advantages, I started to focus more on the positives. I also started linking it to some social science research I heard about on NPR as it helped me to understand what was going on here.

Nobody wants to wait in queue, and there are times when a center just can’t staff to meet the volume demands. My little snow storm is an excellent example. A callback offer with a successful reconnect can deliver a “Wow!” customer experience. Now this is where that research comes in. As Shankar Vedantum, NPR’s science correspondent, recently reported, “It turns out that when we think about experiences, we don’t think about the experience as a whole—we are significantly biased by how the experience ends.” So, “You weren’t there to answer my call when I needed you!” is readily overshadowed by, “That was great that you called me back and I didn’t have to sit in queue!”

Now we know why customers love it. But why do contact centers love it? It can help them meet service level or other metrics. It can improve customer satisfaction or Net Promoter Scores. Perhaps they can even avoid staffing up for peaks (and, therefore, overstaffing in general), overtime, on-call, outsourcer coverage or other complex or costly staffing maneuvers. I’m quite happy to keep my newfound positive outlook on callback queuing (CBQ) and for centers to gain these benefits, if they do so with the customer front of mind.

Getting to Know CBQ
Let’s step back and get to know this technology a bit better. Many think of it as “Virtual Hold®,” which is a registered trademarked of the Virtual Hold Technology® company, a pioneer and continuing leader (more on that later). Others may refer to it as virtual queuing or automated callback.

Regardless of what it’s called, the “killer application” is for peaks that you can’t anticipate or staff for, or situations that are short- lived. But keep in mind, if your volume is up the whole darn day, you need to consider staffing solutions. If you have long seasonal peaks that are sustained all hours, across weeks or months, don’t think callback is a cure-all for that entire period. If you have routine callers and they get a callback every time, they might be justified in getting a bit cynical about how important their call is to you.

Here are some quick examples that highlight good callback opportunities:

• Electrical utilities live with the reality of outages. While much of that volume gets handled by IVRs, the callback option helps those who really want to talk to someone.
• Technical support (internal or external customer facing) can use CBQ to respond to outages, issues or shaky system launches.
• My previous example shows the value for airlines hit by weather and the havoc it can wreak on flights.
• Insurance companies with catastrophic events can improve their emergency response.
• Government centers that just can’t staff up can appear a bit more constituent-friendly.
• Retail, travel, entertainment or consumer goods with promotions that exceed expectations (or worse, cause unexpected confusion), may also benefit, but could risk revenue loss for every callback that goes uncompleted.

Logistics of CBQ
There is more than one way to implement callback. A simple first in, first out will do, but you may want to prioritize these calls, just as you would for inbound contacts (e.g., based on customer profile, relationship or issue). Some callbacks offer the option to schedule a time. This may fit if you have the hours of operation and staffing to meet commitments on scheduled calls without impacting the other inbound calls. These typically use a window (e.g., 30 minutes) rather than a specific time.

Two connection options can come into play, as well. One vendor, Hold-Free Networks, refers to them as “obtain agent first” and “obtain customer first.” When obtaining the agent first, the system dials outbound to the customer after securing an agent. This approach increases agent time but is more customer-focused. Of course, the risk is the customer isn’t there, resulting in lost agent time with no connection. When obtaining the customer first, the system will confirm it has connected with the right person and then connect to an agent on an inbound call that is bridged to the customer. This approach adds efficiency for the center, but may not be as good for customer, especially if he/she is placed into a queue for an agent at this point. So the algorithm needs to be spot on or “reserve” an agent (making it much like the agent-first approach).

Getting It Right (and Not Getting It Wrong)
Now let’s look at some keys to success in callbacks. First is to set conditions on when it kicks in; for example, when the wait time is over 3-5 minutes (not just when it’s over the service level). Offer, don’t force; it should be a customer’s choice to use it. Set proper expectations with your messages, letting customers know what the experience will entail. Many systems will check expected wait time (EWT) in queue. Use this functionality with caution, as it is often difficult to accurately estimate EWT, especially if the center is small and/or has long handle times. A best practice is to check if a caller is already in queue (per my example) so you don’t have multiple entries for the same caller or compromise a better place in line.

The flip side of getting it right is to avoid the applications of CBQ we’ve seen where the issue is really staffing. Imagine this: A center offers callback. The customer is not available when the call is placed. The interaction is closed as “complete,” but the caller’s issues have not been addressed. Other scenarios such as forcing callback, endlessly looping with callback offers while in queue, or shutting down the queue early to prevent more calls coming in are all inappropriate use of the technology. And whatever you do, don’t just use CBQ to manipulate and meet target service levels. The bottom line is, CBQ is not a solution to an understaffed center. If that’s your problem, increase staffing or change your metrics.

Next is to understand how callback works and define how it will be applied in your environment. Address the “what-if” scenarios, and define the business rules you will use to reach customers. Callback is much like setting up an outbound campaign. You must define the number of callback attempts you will make, what you will do based on various outcomes (busy? ring no answer? voice mail?), whether you will leave messages and, if so, what they will convey, and how soon you will try again. If your callback attempt will be a “one- shot” deal, make that clear on the front end so your customer knows what to expect.

Finally, it is important to understand the impact on reporting, which can depend on your architecture, level of integration with your ACD, CTI and IVR, and callback approach (e.g., inbound or outbound calls)—more on that below. You’ll need to look at the impact on call volumes, as well as potential abandons and, of course, service level.

Considerations for Sourcing and Vendor Options
Most contact center technology vendors offer some form of callback, whether their own solution or via a partner. Often, it is a packaged application using their IVR. If you are using the vendor’s IVR for prompting and CTI for screen pops, define the approach to integrate the callback application with the rest of your infrastructure and applications. Some solutions will bypass full integration to simplify, but could compromise things like screen pops and reporting. The integration is also important for reporting, capturing information about the original call as well as the experience on connecting the callbacks. You may see “abandons” based on selecting callback, and you may see inbound and/or outbound calls to accomplish the callback connection.

Virtual Hold Technology (VHT) has technology patents, enabling them to take a somewhat exclusive approach to CBQ. They use SIP messages and telephony events (through CTI) to monitor calls holding in queue as well as the “virtual” calls (callbacks) in queue. Calculating EWT and predicting agent availability, the system calls the customer, confirms connection, and brings the customer and an available agent together. By using this approach, the system creates a placeholder which ensures that the agent receives an inbound call, and the ACD does not peg additional hang ups or abandons. Some vendors resell VHT and others integrate with VHT.

As with any contact center technology today, a key decision is whether to go with a hosted solution or use a premise solution. Some vendors offer both. If you are cloud-based with all your other solution elements, you are likely to use the cloud for callbacks, as well. However, if you use a premise-based ACD, you could still turn to the cloud for callbacks. Cloud solutions are often easier and quicker to deploy, but may lack tight integration to ACD systems for reporting purposes. You’ll need to explore if the vendor is providing the network services as well, and what the cost structures are (e.g., driven by volume, ports or minutes). A cloud solution example is Hold-Free (now owned by Five9).

One final consideration as you pursue solution options: Decide if your callback interests are calls only, or whether they extend into web, mobile or other channels. As the sidebar shows, callbacks are not just for phone calls, so consider solutions with broader channel options, if your business requirements dictate.

You, Too, Can Fall in Love
Applied properly, CBQ can be a true customer satisfier that also benefits the center. It offers payback through network cost savings (not paying for calls sitting in queue) and revenue protection or generation if you can assign value to abandons or increased customer satisfaction. It’s not a replacement for proper staffing for “normal” volumes, but it offers efficiency for those hard to staff for peaks. Add this tool for connecting customers with your highly trained agents executing well designed processes, and you could leave a positive, lasting impression.

Lori Bocklund is Founder and President of Strategic Contact.
– Reprinted with permission from Contact Center Pipeline, www.contactcenterpipeline.com