VCC Live’s study of 2.5 million IVR minutes: 3 key takeaways

We need to talk about IVR.

While contact centers can, of course, get excited about newer AI and automation tools, IVR remains “the most important connection between companies and customers” (McKinsey 2023). It accounts for twice as many interactions as calls with live agents and its market value is estimated to almost double; from $4.9 billion to $9.2 billion by 2030.

That’s pretty incredible for a technology that’s been around since the 1970s. Needless to say, there’s plenty of life in the old tool yet.

Given its huge importance to call centers, there’s surprisingly little data published about IVR; in terms of real benchmarks and its impact on the total caller journey.

With this study, we aim to try and fill this void by analyzing IVR performance data across 3.8 million inbound calls and 40,902 hours of caller time spent in IVR. This data was collated from a defined 12-month period and from 11 different client companies; covering BPO, financial services and retail sectors.

Here are the three big takeaways from our study.

1. Average time in IVR

The measure of average time in IVR provides insights into important operational factors for call centers. These can include efficiency in resolving caller queries, reducing call volumes for agents, and keeping call times low.

Given the wide scale of this study, it makes sense to try and define a singular figure that sets the bar for the average time an inbound caller spends in IVR. 

Across all data collected, the average time spent in IVR was 48 seconds. The median number for this was not too dissimilar at 40 seconds, which suggests there is a slight skew, potentially due to some outliers (we’ll explore these later).

There are very few current benchmarks for what’s considered to be a “good” average time in IVR. One reason for this is that IVR has a variety of use cases and it’s not always about routing calls.

Another important use of IVR is self-service; in which a caller can resolve their query without the need for assistance from agents. Examples of this include tracking orders, appointment scheduling, and over-the-phone payments.

These diverse use cases mean that a good average time in IVR is difficult to quantify. However, we can bring some more definition to this through segmenting our collected data by industry – namely BPO, financial services, and retail.

In the BPO sector, the average time in IVR was 32 seconds. While the use of IVR for BPOs can of course vary, they are typically outsourced call centers that companies pay to handle calls on their behalf. Therefore, it makes sense why an inbound caller’s time in a BPO’s IVR would be lower than the overall average; because a primary goal is to connect callers to agents as quickly as possible.

For financial services, the average time in IVR was 39 seconds – a little higher than BPOs. This is likely because we see more elements of self-service involved in the IVR process. Checking account balances and making repayments are just two of the uses that our financial services client companies employ in their IVR processes. These actions naturally take longer to resolve in an IVR than merely connecting callers with agents.

The main outlier here is retail with an average time in IVR of 1 minute and 8 seconds. In this sector, self-service tends to be just as dominant as live call handling. Order statuses, returns and store opening hours are just a few of the options that we often see on a retail company’s IVR menu. 

While this study sheds light on average time spent in IVR, it’s important to remember that IVR setups can vary significantly from company to company and are dependent on their respective goals. Therefore, striking a balance between efficiency and self-service should be a prominent consideration when developing IVR processes.

2. Disconnected in the IVR

The disconnected in the IVR metric can tell us a lot about the effectiveness of an IVR setup. It’s a measure of the number of callers who terminate their call during an IVR process.

In our study, the average disconnection in the IVR rate was 30%. Again, as with average time in IVR, the median number wasn’t too far away at 26%.

Disconnected in the IVR is seen as a very important measure of customer experience. A high disconnection rate suggests that callers can become frustrated at an inadequate or overly complex IVR menu and, consequently, give up on the call. Conversely, a low disconnection rate is a telltale sign that the IVR is working well and is efficiently routing calls to agents.

What’s interesting, however, is that a high disconnected in the IVR doesn’t always equate to a poor IVR structure. This is because a call can often be resolved through self-service options in an IVR menu and, when the caller subsequently ends the call, this constitutes a disconnection in the IVR – despite it being a successful resolution.

This is most apparent for one particular company in this study – one of Europe’s largest electrical retail businesses – a sector that we know heavily utilizes IVR self-service options. Their IVR disconnection rate was a huge 81% – far greater than the overall company average of 30%. However, this is by no means an indication of a poor IVR structure, but rather a signal of effective self-service. This is further evidenced when taking into account their average time in IVR which is the highest of the entire dataset at 2 minutes and 8 seconds. It demonstrates that callers are willing to spend a longer amount of time in the IVR to resolve their query themselves because they have the options to do just that.

Relying on a single metric alone rarely tells the full story and can actually sometimes misinform on performance. When assessing IVR effectiveness, it’s important to take a broad view of all available data to determine whether your IVR is working well or not.

3. Impact on call queuing performance

Lastly, we’ll take a look at the study’s data concerning call queues and examine any impact that IVR processes may have on queuing performance.

Call queuing is a quintessential feature of contact centers. It helps companies orchestrate inbound calls to better manage wait times, prioritization, and resource allocation. It’s also benefited by processes such as automatic call distribution that routes incoming calls to queues based on predetermined criteria; such as call type, caller language etc.

Here, we’ll take a look at two metrics concerning call queuing: 1) average time in queue and 2) disconnected in the queue.

In this study, the average time in the queue was 47 seconds and the median number was 31 seconds. The reason for such a skew in data was chiefly due to one significant outlier with an average queue time of 2 minutes and 26 seconds across the company’s 563,000 inbound calls handled.

Additionally, across all companies, the average disconnection in the queue rate was 16%. 

Unlike the metrics of average time in IVR and disconnection in the IVR – for which there can be valid justifications of why a higher rate can sometimes be a positive indicator (i.e. self-service performance) – higher queue times and disconnection rates in the queue can only be seen as negative. Inbound callers have chosen to connect to an agent and are only waiting in the call for this reason.

Therefore, it should be no surprise that a higher queue wait time correlates with a higher disconnection rate in the queue. This is illustrated below when looking at some of the study’s high and low performers in call queuing.

Above average call queuing performance

Below average call queuing performance

While the fact a high queue time results in a high queue disconnection rate isn’t exactly a significant finding, it provides a basis to look at what impact IVR may have on all of this.

One slight anomaly observed in the sample above is Retail 2 company. Its average time in the queue of 1 minute and 22 seconds is considerably higher than the dataset’s average of 47 seconds, however, its disconnection in the queue rate is exactly average. 

This begs the question whether IVR could be playing a role here. Perhaps the precise question to ask is: Does a higher time spent in IVR correlate with a lower disconnection rate in the queue? 

Looking collectively at all 11 companies included in this study, this certainly seems to be the case.

Callers spending around 30 seconds or less in IVR have an average disconnection rate in the queue of 22%; compared with callers spending 30 seconds or more in IVR whose disconnection rate is almost half that at 12%.

So, why does a higher time spent in IVR equate to a lower disconnection rate in the queue? Here are three factors why this could be the case.

1. Caller perception of effective call screening: Actions such as providing preliminary information (call triage) and a positive IVR experience could give the caller confidence that, when their call connects to an agent, it will likely be resolved
2. Better caller expectation management: The caller is being well communicated with during the IVR process – e.g. informed of the department they’ll be connecting to and expected wait times – meaning they’re less likely to disconnect in the queue
3. Callers who are happy to queue: A higher time in IVR suggests self-service options were offered but the caller is choosing to resolve their query via an agent – which therefore suggests they are willing to wait a little longer to do this

All of the points above allude to several key considerations for an effective IVR setup and why keeping IVR processes as concise as possible isn’t always the best approach. What’s more, it demonstrates the extent to which IVR can affect the entire call handling process – from the very start of the call all the way through to its resolution.

How to easily build your own IVR setup

VCC Live’s contact center software comes equipped with a fully customizable IVR system that can be created, tested and rolled out for all of your inbound calls.

To see how our IVR feature works, check out the video below: