When people think of contact centers offering a call-back instead of putting callers on hold, the first thing that comes to mind is improved customer satisfaction. Consumers hate long hold times, so removing that frustration is obviously going to lead to happier callers. For many Fonolo customers, this is the primary, or even sole, motivation for adding call-backs.
But there’s another angle to this story. When deployed correctly, call-backs can deliver concrete ROI through several paths: lower abandon rate, shorter handle time, reduced telco cost and more consistent call volume. In this series, we’ll devote one post to exploring each benefit.
These posts will be more detailed and analytic than usual, so if you want a more high-level introduction, try 5 Reasons You Need Call-Backs in Your Contact Center.
Intro to Abandon Rate
Today’s post looks at call abandonment. How is it measured? What are its implications? And, most importantly, how do call-backs change abandonment behavior in a way that benefits the call center?
An abandoned call is one where a caller was placed on hold and disconnected before reaching an agent. The Abandon Rate (or Abandonment Rate) is the percentage of abandoned calls out of total calls. Abandonment leads to higher repeat calling, which lowers both First Call Resolution (FCR) and, of course, customer satisfaction. As you would expect, abandon rate grows with ASA (average speed to answer). The longer callers are on hold, the more likely they are to give up.
Many call centers have a tough challenge with abandon rates. One way to reduce call center abandonment rate is to add agents until the rate shrinks down to the target level. But of course, extra budget for staffing is not always available.
(Call centers with “spikey” call volume face a tougher problem. If you staff to the peak volume, you will have a lot of excess agent capacity at other times. This is another area where call-backs can help, which we’ll cover in part 3 of this series.)
Figure 1
How Call-Backs Help
Call-backs reduce call abandonment by allowing callers to keep their place in line, without staying on hold. Once callers switch to this mode of waiting, they are much less likely to abandon. That’s because the mental effort required from the caller is much reduced. (For companies focusing on “customer effort score”, this is a good point to keep in mind.)
A study by Contact Babel showed that 32% of contact centers experienced fewer abandoned calls after call-backs were added. At Fonolo, we’ve witnessed this effect many times. Two recent examples are the deployments we did with Technology Credit Union and Bright Horizons, where we saw a 37% and 33% decrease, respectively, in abandon rate.
To learn more about these projects, you can read a case study about Bright Horizons, or watch this short video about Technology Credit Union:
(Those two companies have something else in common: They are both running Avaya Aura call centers. If your call center is based on Avaya, you’ll be very interested to read about our Avaya partnership.)
What’s the Cost of an Abandoned Call?
To get a quantitative ROI based on reduction in abandon rate, it’s necessary to assign a dollar value to an abandoned call in your call center. In a sales scenario this is fairly straightforward: it’s the “opportunity cost” of each missed conversation between an agent and a caller.
Here’s an example of that math for fictional company “ExampleCo”.
Figure 2
Some notes on cost per abandoned call:
- Depending on what you sell, “Lifetime Customer Value” might be better to use for the top line.
- In a scenario other than sales, it’s harder to get this number. However, companies do come to us with a number based on their own internal processes.
- This number has a very wide range. We’ve seen values as low as a few dollars and as high as $120.
Calculating Your Savings
Once you have the cost of an abandoned call for your call center, the next question is, how many abandons can be avoided. Judging from the case studies above, let’s choose 35% for the “Abandon Rate Reduction Factor” and continue working through our example:
Figure 3
Understanding the Numbers
To get a better handle on the abandon rate reduction, let’s turn the math around and look at the number of call-backs used, as a fraction of total call volume. Here, that works out to 10.5% (52,500 divided by 500,000). We call this is the “Overall Call-Back Rate” and it is determined by two factors:
1. The first, “Offer Rate”, is straightforward: What fraction of calls will hear an offer for a call-back? In order to make use of a call-back option, callers have to hear an instruction like, “Press 1 to get a call from the next available operator…” at some point. Is the offer message played for all calls and at the start of the queue? If so, then the Offer Rate is 100%. If call-backs are enabled only under certain conditions, or if the offer message is played only after X seconds in the queue, the Offer Rate is lower.
2. The second factor is the “Take-up Rate” rate, and it’s a bit trickier. What are the odds that an offer is accepted by the caller? It depends on many factors such as the demographics of your callers (How impatient are they?) and the nature of the calls (How urgently do callers need to reach agents?), whether callers have heard an estimated wait time, and even the wording of your offer message.
Figure 4
The Impact of Offer Timing
The “Time-Till-Offer” or “TTO” is the amount of time a caller spends in the queue before hearing a call-back offer. It is quite important because it impacts both the Offer Rate and the Take-Up Rate
It impacts the Offer Rate because as TTO gets longer, a larger fraction of callers will abandon or connect with an agent before hearing the offer. (In the extreme, if TTO is greater than your longest queue time, then the Offer Rate is zero.)
TTO impacts the Take-Up Rate because of the psychology of callers. Some fraction of callers will take an offer that is made immediately while others will opt to “wait it out”. But after a few minutes on hold, callers are more likely to take that option, so the Take-Up Rate grows. After a while, this effect flattens out. The chart below is an example of this relationship.
Figure 5
This chart is provided simply to facilitate the example calculations here. You shouldn’t take this as a guide for choosing TTO in your call center. The Take-Up rate is affected by so many factors that the only way to get an accurate picture of its relationship to TTO is to experiment with different values. Here is a short video of a Fonolo customer talking about his experience adjusting TTO:
[fonolo_overlay_video src=”105148079″ title=” When to Offer the Call-Back” time=”0:40″]
Take-Up also increases by making multiple offers for the call-back. In fact, this is the best way to maximize the overall call-back rate. (If you’re assessing vendors for a call-back solution, make sure the ability to make multiple offers once a caller is in the queue is on the feature list.) To keep the calculation simple, we will assume here that just a single offer is made.
Although this sounds complicated, it’s actually good news. Changing TTO is an easy “lever” to adjust and has a big impact on the performance of your call-back deployment.
Let’s see this in action. Figure 6-A shows a histogram of calls on a typical day at ExampleCo’s call center. The calls are grouped by wait time (or ASA). With the TTO set at 35 seconds, the 3 rightmost bars represent calls that will hear an offer. This adds up to 30% of all calls, which is the Offer Rate.
We see in Figure 5 that when TTO is 35 seconds, the Take-Up Rate is 40%. Figure 6-B shows us how the Offer Rate and the Take-Up Rate work together to yield an Overall Call-Back Rate of 12%. This exceeds our goal of 10.5% to achieve the scenario in Figure 3.
Figure 6 A,B
Figure 6-C shows what happens if we change the TTO to 15 seconds. More of the bars are now “in” which raises the Offer Rate to 80%. But the Take-Up rate drops to 30% (as per Figure 5). The net result, though, is the Take Rate doubles to 24%.
Figure 6 C,D
Referring back to the math in Figure 3, we see this also doubles the total savings from $630K to $1.26m.
Calculating ROI
So now we have a handle on the savings but, of course, an ROI calculation requires us to understand the cost as well. To keep things simple, let’s make some assumptions about the call-back solution you purchase:
- No initial fee
- Price based purely on the number of call-backs used
- No concurrency limit
With this pricing model, the ROI is quite straightforward. Going with the scenario above, the company saves $630K and uses 52,500 call-backs. To be in the black, the cost per call-back needs to be $12 or less. Not surprisingly, this is the cost per abandoned call. To get a 5x ROI (a common benchmark), the cost per call-back needs to be $2.40 or less.
[You’ll be happy to know that Fonolo meets the criteria above and would comfortably fit this ROI case. But we encourage you to explore other options on the market, as well.]
Some Extra Notes on the Calculations
- Call centers are sometimes directly tasked with a target abandonment rate, in which case it’s not necessary to calculate a “Cost per Abandon”, etc. to justify call-backs. (And the ROI is “because my boss told me so”!)
- Not every call-back used represents the elimination of an abandoned call. Some callers opt for a call-back who would not have abandoned anyway. For this reason, the only way to really know the decrease in abandon rate is to run a test and observe. The calculations here are to help estimate the amount call-backs that will be used, and illuminate the variables that impact success, but provide only a rough prediction of abandon rate reduction.
- Even when a caller chooses a call-back instead of waiting on hold, there is a chance they could abandon. In our work, we’ve found the “call-back abandon rate” to be very low – around a few percent. To be strictly fair though, you should include this factor in calculations like Figure 3, by doing a weighted average of the two abandon rates.
Blog Series: ROI of Call-Backs
Part #2: Reducing Handle Time
Part #3: Reduced Telco Cost
Part #4: Smoothing Out Call Spikes
