Posted on: January 28, 2016
The Seven Deadly Sins of Automated Phone Systems
Automated phone systems, everyone hates them, everyone has to use them.
These systems force callers through an excruciating menu system only to leave them on hold forever waiting to hopefully talk to a human before their call is dropped.
These systems cause callers to repeatedly press ‘0’, be hostile when they finally reach an agent, and, if they don’t reach one soon enough, to eventually hang up in frustration.
Make no mistake; these phone system interactions are part of the users’ experience with your company’s system or service. This could be the experience that causes them to leave.
All is not lost though. It is not a mystery why people hate calling your company; there are very concrete, very fixable reasons for their hatred.
This is Good News
1.Forced to Repeat
It is frustrating for the caller to have to repeat themselves over and over, especially if the reason for the call is a company error. Information that is collected by the telephone system or by other agents is often not transferred, forcing the caller to provide information again and again.
It’s surprising how often this happens in the first question of the system.
“For service in English, say ‘English’.”
I say ‘English’ as casually as I can. Which is really not casual as I am currently in a dead silent room full of people.
“If you meant to select English, please press ‘1’.”
Yes, FedEx wants me to tell them twice in a row that I really do want English. Why would they force me to provide redundant information? If I had to guess, it would be because voice input is not reliable for some people. Although if this is the case, voice input should be avoided as the default interaction method.
“Welcome to Fido! For service in English, please press ‘1’.”
I press one and hear as the system transfers me to another number. There’s a pause as I wait to either talk to someone or to hear the next menu. And then…
“Bienvenue a Fido! Pour le service en Français, fait le un. For service in English, please press ‘2’.”
Now, while I could explain some reasoning behind the FedEx redundancy, this one seemed like it must be an error. Why would Fido want to ask me twice in a row which language I want? Were they testing my knowledge of English number words? Were they upset that I didn’t listen to the original message in French?
This just seems like an error that either no one at Fido knows about or no one considers important enough to fix.
These problems occur so early in the process and take up just a little bit of your time, so they are unlikely to serve as a major annoyance. However, as you get further on in the process, your tolerance begins to wane.
If time is wasted when the caller gets to an operator, the effects can be felt.
If your operators need to spend one minute of time collecting information that is already in a company database, then this quickly adds up to a lot of wasted time.
Look how much one minute of redundancy per caller costs your company
|Cost of 1 Day of Callers*||Cost of 1 Week of Callers*||Cost of 1 Month of Callers*||Cost of 1 Year of Callers*|
|1 min of Redundancy||2 days time||2 weeks time||2 months time||2 years time|
*Based on the average number of daily callers to 911 centre in Washington D.C.
If your operator spends one extra minute on their first caller, then that’s one minute the second caller is delayed. If the second caller is then delayed one minute, then the third caller is delayed for two minutes (because they have to wait for the second and first caller).
Your operator also has to spend three extra minutes serving three people. After only 60 callers, the phone operator has to spend an extra hour. This very quickly starts to add up for everyone.
For a call centre that averages only 3,800 calls a day, that’s a lot of time that could be wasted for a one-minute delay. If your call centre gets more than 3,800 calls, the cost is much greater.
2. Forced to Wait
When something goes wrong, your caller is already not very happy. Then add on a long hold time and you have a recipe for disaster.
Every minute on hold means one more minute to feed your caller’s anger towards your company. Callers end up with endless amounts of time to contemplate how much time is wasted to fix their problem.
Even if your company didn’t mess up and the caller is calling for information or to open an account, this is going to damage your relationship with the caller.
If this long wait and poor experience is true before they hand over their money, they might decide that it’s a better idea to go elsewhere.
If your caller is simply calling for information, they get frustrated that it’s so difficult to do something simple. As they wait, and wait on the line, it seems like your company is telling them that they aren’t important. By the time the caller finally does get though, they might have a very different question in mind: can you transfer me to cancelations?
The worst situation though is when the caller’s problem is very urgent. Calling a utility company when there is no heat or calling a credit card company after losing a credit card are both high stress situations that leave callers already predisposed to anger and stress. For these callers, any amount of waiting can seem like forever.
3. Annoying Apologies
“Sorry for the wait. Your call is important to us. Please stay on the line.”
Does that make you feel better? If you’re like most people, that probably actually makes you feel worse.
North et al. found in 1999 that apologies from the system negatively affected customer reactions to telephone waiting, yet almost 20 years later, companies are still using them.
People listening to these apology messages do not feel like they have been apologized to. It is unclear why a machine with no feelings of remorse constantly delivering apologies to untold numbers every five minutes is supposed to make you feel better. Nothing about that seems sincere.
The message only serves to draw attention to the fact that you are still waiting and you are still not being helped. It also doesn’t provide any useful information to the caller about if they are even progressing in the queue at all.
4. Not Being Able to Talk to a Real Person
If it sometimes seems that systems were designed to prevent the caller from speaking with someone, that’s because they probably were. Many companies have a goal to get your questions answered without getting a person involved. It’s cheaper to have a machine answer your questions than employ someone to answer the same questions over and over again.
However, sometimes the caller’s question isn’t answered by the automated messages and they still need to talk to someone. When it becomes hard to figure out how to do that, they get frustrated.
Being unable to talk to anyone further pushes this image that your company does not care about the callers.
5. Bad Menus
Bad menus can ruin the caller’s experience even before they end up on hold. The options provided on menus need to be well thought out and concise.
There have been a number of times that I have called local stores and had to listen to a four-minute greeting before I was able to get to the phone menu. If people commonly call about store location and hours, then make these your first two menu items instead of listing all this information in the greeting. Anyone who isn’t interested in this information starts to get annoyed as the message drags on and on.
When people finally reach the menu the next problem they can run into is a series of options that don’t fit their situation. This forces users of your menu system to select an option that doesn’t apply to their problem to hopefully eventually reach an individual who can direct them where they need to go. If someone wants to call about finding a shoelace in their bag of rice and the manufacturer’s options are Order Status, Place an Order, and Request a Sample, then they’re just going to have to choose one of those. Even though they know it’s not right.
Users also choose the wrong option when their terminology does not match the menu. When you’re calling Dell about a laptop problem and you don’t hear the term “laptop”, you might end up selecting the option for all other devices instead of realizing that Dell calls these devices “notebooks”.
Too Many Options
Choosing the right option gets even harder as you increase the number of options users have to select from. You should never be pushing your menu out to nine options and I’d argue that even seven is quite a lot. By the time the caller gets to the last option they’ve forgotten what the first few were. So either they have to repeat the menu, if they are able to, or just guess a number and hope it was right.
6. Technical Problems
Let’s for the sake of argument say that calls are dropped due to technical problems and not on purpose. This is something you should clearly fix. Even if your callers don’t assume it was on purpose, they are still going to be annoyed for waiting all that time only to have to go back to the end of the line without ever being served.
Another common technical problem is having a bad voice recognition system. These systems can go bad quickly. If the caller is calling in a noisy environment or if the system in not in the speaker’s native language, or if they speak a regional dialect.
Even if the system can understand what the caller is saying, if you provide an open menu, people might try to say words that the system doesn’t understand and get frustrated when they end up having to use the menu anyway.
If you feel that you absolutely need to use these systems, then you have to have an alternative option provided for when there are problems or if users are not comfortable with using them.
7. Bad Music
Yes, let’s talk about the music, because it does matter. Everyone has a different taste in music and often there isn’t a clear choice about what type you should play.
However, it is abundantly clear that you probably shouldn’t just play one song on repeat.
Or the same three songs.
Fortunately, there is some research to guide you here as well. Instead of forcing everyone to listen to the same music, why not let them choose?
According to research by Kortum et al., being able to choose your own music increased user satisfaction while on hold.
Good music can make the passage of time more bearable for callers and maybe even help them forget that they are waiting. Bad music, however, can make the wait excruciating.
Here’s two callers, same call centre, one hates the music, one loves the music.
Music is a very individual preference. One size does not fit all. If you don’t let callers choose, then someone is always going to be unhappy.
Until callers can choose their music though, perhaps go with jazz (Kortum et al., 2009).
So What Can You Do?
You don’t have to be doomed to a series of negative tweets, angry customers, and lost business. You can do better.
Recommendation One: Provide a call back service
No one likes waiting. Don’t make them. Provide users with options besides putting their lives on hold for your company. A quick look on the Apple website shows that users here are provided with a couple of options of how to proceed with their problem.
Apple doesn’t force the caller to sit on the line listening to hold music. They’ll call you when someone is available to help or the caller can schedule a time when it is convenient for them.
A word of caution about call back systems, however. If your company has this service, then when users get a call from your system, they should be immediately connected to a human, not just put back in a queue.
Recommendation Two: Look at the experience as a whole
Look from the beginning of the call experience, when the caller first decides that they need to contact the company, to the end and consider how various aspects of it could be made easier and faster.
Talk to a few of your callers about their journey through the process and highlight common pain points. Then focus on fixing these areas.
It’s important to talk to users rather than making your own assumptions, as without users you might miss common trends. For example, if you have a smartphone with a data plan, you might never consider how users report an internet outage if they have no internet to look up the provider’s number.
It’s also a good idea to make use of the information that you already have. Any information you collect from the user is assumed to be known across the organization, it’s your job to make this the case. If a user has typed their account number into the IVR they shouldn’t be asked what this number is again when they connect with a person.
Recommendation Three: Take a look at your menus
Do callers often end up connecting to the wrong department? Yes? Sounds like you have a phone menu problem.
Conduct an open card sort to get an idea of how users categorize reasons they might call and what they call these categories. Then, based on these results make a new phone menu. Finally, before releasing this new menu, test to see if users categorize their issues in the buckets you expect they will.
An open card sort ensures that your phone menu presents options in your users’ terms (and not company jargon). Testing it before release, also ensures that you interpreted results correctly and allows you to catch any problems before the system is live.
Finally, to account for user fatigue and memory, you should ensure that option lists are short and that the number of layers is shallow.
Recommendation Four: Hire more people
No matter how great your IVR system is, in the end some callers are going to need to talk to someone. If you don’t have enough people employed to accommodate your volume of callers, wait times will be long, and people will be unhappy.
Implementing these recommendations will not only improve the experience of your callers, but also the experience of your call centre employees.
Michelle believes that good design is like silence. You never seem to notice when it’s there, but its absence is always missed. With a thorough understanding of end users, Michelle Brown creates these silent designs that support users through every step of their journey. She delights in crafting pleasurable experiences through a variety of research and design methods and is always pleased to use her knowledge to take designs to the next level. Her experience spans wireframe creation, usability testing, persona development, design feedback, and card sorting. As an Experience Architect, she has proven that she can meet aggressive schedule objectives and deliver actionable results. Michelle has a MSc. in Computer Science with a specialization in Human Computer Interaction and is an Experience Architect at Akendi, a firm dedicated to creating intentional experiences through end-to-end experience design.
Sign up for our UX Blog
Don't miss an article! We'll notify you of each new post.