IVR, Abandoned Rate, and TSF

[Carrigman]

We have recently introduced an IVR system and our customers are still getting used to it. There is a high abandon rate of callers who, when presented with the IVR, simply hang up, or, when they reach the options, fail to continue. Apart altogether from the possible flaws in the IVR as constructed - and clearly there are - my question is: is it normal to discount such abandons when reporting to senior management on weekly/monthly performance? In other words, should I be counting only those who abandon while waiting to speak to an agent? Likewise, I am counting those transactions that are handled to finality by the IVR as delivering 100% TSF. I presume that this is standard practice.

Regards, and A Cool Yule to y'all from Ireland.

John

[JTaylor801]

There are two schools of thought on this question. I don't count abandons while in the IVR in my weekly, monthly, qtrly metrics. However, this doesn't mean that I disregard that data. I review reports of calls that abandon in the IVR and plot them on a control chart. I pay special attention to this report when changes are made...but I have rarely seen any "special cause" exceptions on the control chart that show cause for alarm.

If you're just implementing the IVR concept and before didn't have any kind of complex menu's then I'm sure that is the cause for your high abandons. As I believe you mentioned, you're doing your do-diligence work to make sure the IVR is functioning correctly and if that is the case all you can do is wait until your Customers get used to the system.

One idea (if you're not currently doing it) is to add a special greeting to your IVR that alerts your Customers that you've made changes to the system to better serve them and to listen carefully to the menu etc. This might help...especially if before when they dialed your telephone number they heard a live body answer.

Hope that helps.
Jerry

[JoeB]

The answer to the question whether to report “short abandons” and “IVR abandons” is not a matter of opinion or “standard”. The answer should be based on the business impact of abandoned calls.

You did not describe your business environment, i.e.. who are the callers, do they have alternatives, what is their business perspective concerning abandoned calls (or, for that matter, long wait times) and so forth. So IÂ’ll use simple examples to demonstrate my point:

Abandoned calls are business critical if dissatisfied callers may defect and take their business to the competition. On the other hand, if the callers are internal staff, do not experience significant downtime, and will call later, then abandoned calls are much less critical.

My recommendation is to record abandoned calls and understand their causes. The relative importance you give them in your management report depends on their business significance. Same factors should determine how much effort you invest in improving abandonment.

See more at http://www.diagnosticstrategies.com/abandonment.htm

Joe

Hi John,

As Joe said, is not about how you have to present the data, is about the importance of the matter. If they just hang up the phone, you have to understand why?. Like Jerry said probably they don’t understand the message of your IVR or they don’t like be attendant by a machine, happens all the time. Sometimes the companies take decisions and not always know the real impact in our customers, I don’t know you company profile and services, but those customers can go to the competence if you can’t understand them.
We call here in Mexico this abandon calls, the Abandon Ghost and we count always, because is people don’t understand your message, or you take so long to answer them and here is link with your ASA (average speed of answer).

Take a look of this matter and make an investigation of why they hang up.

Regards

Miguel

[pavlov]

A significant variable is whether or not your IVR has any true self serve functions avaiblable, or if it is primarily for skill-routing to agent groups. It sounds like it has true interactive self serve functions in it.

What I prefer to do, is track calls that terminate in the IVR, as a "termination rate", which is different from an abandon rate. My "termination rate" is the % of calls that enter the IVR and never get to an agent queue. Abandons remain to be calls that disconnect after leaving the IVR, and hang up while in a queue.

Now, is a high termination rate a good thing? It depends on how many of the calls are functional terminations (service provided and caller dosconnected), versus disfunctional terminations (caller hung up without getting correct service provided and likely called back). This can often be gleaned from IVR reports that indicate where in the IVR structure your "terminations" are taking place, and wht type of interactive data was sent back. This assumes you have such reporting capabilities.

Also, a good (more rudimentary) thing to measure is "calls accepted into the IVR per 100 subscribers/customers". If this number (your total call volume) climbs with the implementation of your IVR, then you are probably getting alot of disfunctional terminations.

Good luck