Expert's Angle: The People in Your Neighborhood: The IVR

Welcome back to our series, “Who are the People in Your Neighborhood?”

In previous entries we looked at the Customer and the Agent. In this article, we're going to look at the voice of your company. Agent? Manager? Salesman? Marketing? No, it’s none of these. It’s the IVR!

I know this series is people in your neighborhood, but we’ll define people broadly enough to include robots. Ask any IVR developer, and they'll tell you that the IVR is the "voice" of your business. In many respects, this is true – most contact centers have an IVR that handles the first contact from callers.

Your Contact Center is not monolithic, but a system of people, technologies, and processes. These components must not only work individually, but work together in harmony and synergy to deliver the level of quality customers expect (at a cost you can live with). The IVR, as a key component, must work effectively as a standalone application, but it must also be coordinated with the rest of the enterprise.

In this age where corporations are people too, the corporate voice is embodied in the IVR. It’s often true that the IVR is the first voice customers hear, and in most contact centers, the IVR is the busiest agent.

That makes a lot of sense. Because IVR/Auto Attendant has a few advantages over human agents, we assign it simple repetitive tasks, like the initial greeting to the automated IVR system.

Other IVR benefits include:

• The IVR never needs to take a break.
• The IVR always sticks to the script.
• The IVR always handles irate customers professionally.
• The IVR can handle many tasks without requiring a live agent.

Owing to the fact that it’s a robot, the IVR is not only the busiest agent, by far the IVR is the most consistent agent – it acts as it is programmed to - and only as it is programmed to. Therefore, the design and subsequent support of the IVR is critical to its success.

The aforementioned IVR design needs to happen intelligently because the IVR needs to be transparent. By transparent I mean that the IVR provides no impediment to the customer’s ultimate goal. The customer’s ultimate goal is the solution to their problem. (If you want to go somewhere, you get in your car and go there. You don’t have to work through a problem of, “How will I use the car to get where I want to go?”)

IVR has emerged as a dirty word in some circles. When I talk about IVR, people say, “Oh, I hate IVR!” (I won’t lie - they do!) But inevitably what they really hate is bad IVR. People can have bad experiences with IVR and let that be representative, but in general attitudes about IVR are evolving as technology improves. If I may quote Patrick Barnard at TMCNET:

“As a result of their improved functionality and capabilities -- not to mention the increasing number of deployments -- consumers aren’t just growing tolerant of IVR systems they are actually coming to prefer them. With these speech-enabled, self-serve systems, consumers can carry out basic transactions -- such as getting one’s bank balance, carrying out an over-the-phone transfer, or requesting a copy of a brochure, etc. -- without having to wait on hold to speak with a live agent.”

When we’re trying to give customers access to their solutions, the first point of contact is almost always the IVR. Given that such a huge percentage of customer traffic hits the IVR, perhaps we should give IVR quality some thought.

Good vs. Evil

Accepting that the IVR is important, let’s talk about good and bad IVR. Remember, people don’t hate all IVRs - they only hate the BAD ones! There are two sides to every IVR: the customer side, and the business side.

Customer P.O.V.

GOOD IVR for Users

So what is a good IVR?

To be considered GOOD, the IVR must:

• Allow user to reach a live agent at any time they choose. The IVR must recognize that ZERO=TRANSFER to Live Agent, always.
• Provide self-service whenever possible. As technology improves, and the devices we hold become more and more powerful, we’re more and more capable of handling our own chores. In general we prefer it. According to the 2009 study, Driving Consumer Engagement With Automated Telephone Customer Service by Forrester Research, “Consumers rate automated telephone customer service higher than live agents for certain straightforward interactions.”
• Use professionally produced voice recordings. If you are serious about service, you’ll have prompts professionally recorded. If you don’t, you aren’t.
• Limit choices in a menu to 5. Keep menus simple. It is better to create individual IVRs for different tasks than it is to create on convoluted system.
• Limit IVR Menu Layers to two. (See previous point.)
• Pass caller information to agents upon transfer. The technology that allows agents to receive screen pops with caller information along with the call is not new, and it is no longer cutting edge. It is expected by the customer. If I provide my account ID to the auto attendant, and then have to give it again to a live agent, I know I’m in a bad system.
• Be tested! Thorough testing is critical. Don’t let your customers find the problems in your system.
• Allows experienced users to get through the system quickly; “barge-in” lets users who are familiar with the system skip prompts they’ve heard before.
• Be used by management! The people in decision making positions in your organization need to experience what customers are experiencing with regularity.

BAD IVR for Users

Quality has a transcendent quality - we know it when we see it. That goes for poor quality as well as high quality. Bad IVR systems share some common characteristics.

A bad IVR:

• Will prevent users from speaking to a live person. Some contact center operators have figured out that it can be expensive to staff a contact center. So they become stingy with this resource and trap customers in the IVR. This is why people hate IVRs.
• It’s not IVRs that annoy people, it’s criminals who use IVRs to prevent customers from getting what they want, need, and deserve - good service.
• This is also known as the IVR MOAT. No ability to get to a live person. Ironically, whatever a contact center saves on agent resources, will be lost due to customer churn.
• Will transfer calls blindly to the agents - If and when the call moves from IVR to the live agent, it should pass all caller information available to the agent as well.
• If the IVR asks for my account number and I give it - the AGENT had BETTER HAVE IT when I get transferred. No way should I be giving that information twice.
• Will have poorly designed menus - you’ll know these when you see these, and they are revealed through testing.
• Menus will have more than five options - three would be better and keep levels to two-three at the absolute maximum. A laundry list of options is a tell-tale sign of a cluttered and user-unfriendly system.
• Note: This is easy to fix: callers of different varieties can be given different direct phone numbers. The best menus have three options.
• Have irregular voice prompts - prompts recorded by a variety of non-professionals by a variety of means. (If you hear more than one voice during your transaction, that’s a bad sign.)
• Inconsistent volume and quality lead to a poor experience. Instead, use professionally recorded voice prompts; they’re inexpensive.
• Is untested. If your organization’s leadership is not using the IVR, this is a problem. When changes are made to the system, they must be tested and documented. Otherwise, your customers find the problems, and after a while small changes add up to a system that’s no longer in control.

Business P.O.V.

The IVR serves customers, and customers are top priority, but the IVR serves very real business needs as well, and they need to be addressed.

Good IVR for Business

A good IVR:

• Is reliable and available. The best IVR is the one that’s working and taking calls.
• Handles as much self-service as possible while maintaining positive customer experience.
• Is maintainable - how easy is it to make changes to your IVR system?
• Note: Changes happen under change control, are documented, and are tested. On-the-fly changes lead to problems.
• Is a strategic fit - who well does your IVR integrate with your other IT systems?
• Has data traceability - how well can you see in reports what actually happened on your system? When people say they had a problem with your IVR, can you recreate the call and verify?
• Is well documented. If your contact center processes enough transactions, there will be complaints, and the typical complainer will say, “The IVR told me X.” The first step towards solving the perceived problem will be to verify what happened - to compare the report of bad behavior with what your system records tell you. We’ve discovered two amazing (well, maybe not...) things over the years related to this.

People blame the IVR for their own mistakes. An astounding number of people with IVRs don’t have the call flow of their IVR documented. You can’t address complaints if you know neither what the system was supposed to say, nor what it actually said!

Bad IVR for Business

A bad IVR for business:

• Is unreliable - you spend a lot of time fixing it.
• Is proprietary - Proprietary is not always bad, but vendors who lock you into inflexible platforms limit your choices going forward.
• Is difficult to maintain/change
• Systems that are poorly documented are difficult to maintain. No one knows exactly what they do or how they work. Take away change control, and no-one can be sure exactly what they’re dealing with from day to day. This is a recipe for outages and long, tedious troubleshooting.
• Is a black box and is difficult to test
• Is undocumented
• Is untested
• Have I mentioned that testing IVR is absolutely critical? Changes can create unintended consequences. No changes to the IVR should ever be pushed without testing. It’s impossible to test effectively without accurate documentation. (You may be seeing the pattern here!)

Is My IVR a Good IVR?

How do you know if your IVR is well designed? Test it! Document it! The list above refers to GOOD and BAD IVR from the customer perspective.

A documented IVR is well-known and predictable. When the IVR is tested against the documentation, and change control ensures the documentation and the actual deployed system are in sync, then you have complete confidence that your system is going to provide excellent service to customers without killing management and IT with heartburn and stress.

We’ve probably gone on too long already, so we’ll just say in conclusion, the IVR is critical, and both GOOD IVR and BAD IVR happen as a result of a lot of decisions. Good IVR pays rewards and BAD IVR punishes organizations with dissatisfied users. Do yourself a favor and put the time into creating an excellent IVR.