Slamming the Door in a Customer’s Face - The Good, The Bad and The Ugly of an IVR

If you are not familiar with the acronym IVR I can almost guarantee that you have used one.  Interactive Voice Response is the menu system that call centres use to get you to the right advisor - matching your problem to an advisor who has been trained to handle it.  You may be growling by now, you may hate IVRs, you would probably contest that they get you to the right agent.  Well, that is because they are generally really badly designed.

I like to think of the IVR as the front door to a shop.  Normally, you would go through the door, buy whatever you need, leave through the door and never notice it – but what if the door is really heavy, or it is a glacially slow-moving automatic door?  Then you notice, you comment on it and you certainly remember it without fondness.  Well, an IVR is the same – it should be intuitive, get you to the right person who can help you, and you shouldn’t really notice it.  I have sat on many focus groups where customers have been grilled about IVRs and the response is nearly always the same – “I am cool with pressing the buttons if it is quick, easy and gets me to the right person first time.”


So far, so obvious!  Why then do companies get IVRs so badly wrong?  Inside-out thinking is why!  The IVR is designed starting from some internal concern rather than from the customer's viewpoint.  Common pitfalls are:

  • Micro-Skilling.  If you have a whole population of single skilled advisors then you will have a complex IVR because you will have to ask increasingly detailed questions to get to the right agent.  An IVR will only ever be as good as the skilling that sits below it - the greater the degree of multiskilling the fewer menu options and reduced transfers.
  • Project fever.  The new flavour of the month gets all of the heat and light of the organisation and an over-zealous project manager wants the IVR restructured for the needs of his project.  In my experience, every project manager, tries this on every project and gets batted back by the guardian of the IVR.  But the real danger comes from the CEO backed pet project where the IVR is tweaked to satisfy his needs but no longer flows or makes sense to a customer.  The sales machine is equally dangerous!
  • Data Hunger.  Bizarrely I have seen the desire to understand customer behaviour dictate the structure of the IVR as it is a rich source of call reasons.  But if you design the IVR, in the company’s language, from a data viewpoint you are influencing the customer's actions - your insights reflects your own IVR structure not what the customer really wanted.  Doh!
  • White, middle-aged, middle-class male in the Boardroom syndrome.  Where a bunch of executives sit with lattes around a speaker phone and pontificate their strongly held views of the correct wording of the IVR.  There are two reasons this will only ever be a disaster – they are not representative of the demographics of their customer base and they are “insiders” – they accept as normal a language that is alien to the customer.

There should only be one rule in IVR design.  Design for the customer.  It's that simple - every compromise from that rule will take you a step further in the wrong direction.


So, time for an example of bad IVR design?  This weekend my trusty Samsung S3 died so I prepared to call my provider – T-Mobile.  It is a whole sorry case study in poor customer service that will be the subject of my next post, but for now – ladies and gentlemen, I present the T-Mobile IVR.


I went to the website to find a number and I was pleasantly reassured that the number wasn’t hidden – but, then again, T-Mobile is a telco so should really be comfortable with customers calling them!  I was even more pleasantly surprised to get a quick guide to the IVR on the website – such a great idea to help customers quickly navigate the menu for the most common options. 

I had a handset fault - so I need option 4, then 1, right?  I dialled, I typed ahead, a pleasant lady said “Sorry, that’s not on my list”.  Oh-oh, I redialled – there were only 3 options, there wasn't an option 4 to press.  I listened to the menu again, carefully.  After hanging up and redialling a number of times I cracked the code.  I found the route hidden in plain sight.  It was like negotiating my way to the summit of Everest.  I pressed 1, input my telephone number followed by the # key, then 4, then 1, then 4, then 2.  Then I was told that I would need to give my telephone number.  Yup, the one I had already input.  I cursed my blistered finger as I was told that "We are receiving a lot of calls right now......".

The T-Mobile IVR then began to intrigue me - how could an inanimate object make me feel so bad?  As customer experience professionals we often objectively study a business but it feels different as a customer.  So, what stuck out through the experience?


  • I am afraid, I can’t say much here.  It simply wasn’t a pleasant customer experience.  The only positive I got as someone with “insider knowledge” is that there was definitely an attempt once I had been identified to present menus relevant to me.  But although they were relevant they weren’t well structured or phrased.


  • There were no bail-out options.  As a minimum I would expect an option to repeat the current menu or return to the previous menu.  As a consequence I had to keep redialling.
  • Inside-out language.  The IVR narrative was written by someone who had been kept in a darkened room in T-Mobile Towers for several years without home visits.  PUK code anyone? No?  How about a content lock or a PAC code.  These evocative terms are there in the IVR, presented to customers without explanation, in all their glory.
  • Loading up options.  Somebody must have tried to thin out the IVR menu at some point.  It may look like there are only three options but cunningly hidden within each option are loads more “To get technical support, to get a PUK code, or to report your phone as lost or stolen press 4”.  When you have four options loaded up with three sub options you really, really need to be paying attention.  Especially when they don't appear, from a customer (outside-in) viewpoint, to be logically connected.
  • They didn’t manage my expectations.  It would have been nice to be told how long the queue was and then I could make an informed decision whether to hang on.


  • The hang-up.  In my opinion, this is singularly the worst sin that can be committed in an IVR.  If I didn’t pick an option quickly enough they hung up on me.  There is simply no need for it.  Shame on you, T-Mobile!
  • After negotiating the whole painful IVR process, a recorded message told me that the advisor would need to access my mobile phone menu and ideally I should hang up and redial from a different phone to their 0845 number.  Is it just me or is it really bad form when I am already needing help with my service to redirect me to a number that the company earns revenue on?  Take my call, and call me back on another number!
  • This is just aesthetics but there is a really loud and ugly welcome tone.  I heard it every time I had to redial.  I jumped out of my skin.  I still hear it in my sleep.

So, how was the T-Mobile IVR as a shop door?   It was a door that had to be pulled open but had no handle and a “Push” sign on the outside.  I face-planted into it and then it made me work really hard to get in.  By the time I was inside I was weary and exasperated.

It’s great, but so unusual, to hear a well designed IVR with customers at the centre of it's design.  I would love to hear when and where you last heard a really brilliant, customer-focussed IVR?

The IVR was not the end of this story.  You can follow the link to read the rest of my horrible T-Mobile customer service experience.