I rang Telstra to cancel my data pack (see the post here for an explanation). The customer service rep, trying to be helpful and service oriented, asked me if she could ask me a few questions about my account usage to determine if I’m on the right package. I said ‘sure’ and this is what followed…
She asked me about other services I had (e.g. landline) and when and what I used my phone for (mainly business during business hours). I then asked about whether I had the correct bonus option on my account (i.e. 25% of all calls to mobiles or landlines vs. 50% off all calls to mobiles). And she said ‘Yes. looking at your latest bill, it looks like you’re on the right plan and option).
I said ‘So you can see my call volume to mobiles and landlines and determine that I make an even split?’ And her answer? I bet you can guess…and here’s the rest of the conversation between the customer service rep and me…
CSR: ‘No. But looking at your account it looks right’.
ME: ‘But if you can’t see the statistics of my calls, how can you make the judgement?’
CSR:’Look, I’m a specialist in mobile billings and plans and I do this all the time. You’re on the right plan’.
ME: ‘Don’t you have any statistical analysis tools to help you understand my usage patterns’.
ME: ‘I’m a bit uncomfortable that you’re making a judgement about which plan is best without a statistical analysis.’
CSR: ‘I do this all the time, it looks right. You just have to trust me’.
ME: ‘I’m a bit of a numbers guy. I like having the numbers to back up a position.’
And so the conversation continued with me feeling very doubtful about the quality of her judgement. Why? As a psychologist…
- I know that people are generally not good at numbers
- I know that people are even worse at understanding statistics and probabilistic data
- I know that people constantly look for patterns in what they see to reduce complexity and if they see erroneous or unusual data, they mistakenly generalise from that
- In her context, co-workers provide inaccurate advice and coaching on statistical data to people around them meaning that how one person mistakenly explains a pattern in the numbers is passed from one person to another until it becomes ‘law’
All Telstra needed to do was provide a small summary panel on the screen that shows my account summary to the CSR that shows the essential statistics used to ‘sell’ the various options to make sure I’m on the right plan. What do i mean? Something as simple as:
- Calls to mobiles: $50 (37%)
- Calls to landlines: $36 (22%)
- Calls to 13 and 1800 numbers: $20 (44%)
The CSR could quickly scan the statistics and make a strong decision. Even better, an algorithm could be used to identify the most appropriate option.
This is important, because when I first signed up to the account, I had to choose which discount option to go for. I think I’m like most people and don’t count up my calls each month to see what my pattern is. How can I possible judge which is the best plan and maximise the discount?
Get the computer to do what it does best – maths – and give me the data I need to make a decision. What would take me days of analysis, it can do in microseconds.
The Telstra bill I get each month (don’t get me started on the waste of paper and lack of electronic billing) should give me these statistics and offer alternatives. This kind of push marketing that is clearly in my favour does wonders for customer service, experience and loyalty where there is no cost to me to switch providers.
Churn is one of the biggest issues for telcos and other commodity service providers. You’d think that people would do anything to maximse customer retention. I’d bet it’s a heap cheaper to do this that spend hundreds of $millions on advertising aimed at new customers.