Customer Satisfaction Surveys: Learning From American Express's Mistakes

So American Express really want to know what I think of their service and how they can make it better.



Since I contacted them recently by phone I've received a customer satisfaction survey to complete... oh joy.


Rather than simply ignore it, I thought I would share the experience here so that you can learn from their mistakes.

Q. Overall, how would you rate the service you received when you contacted American Express by phone recently? (Excellent, Very Good, Good, Fair, Poor).


I was changing my address. How hard was that to do well? American Express identify the nature of my call in a later question (Q3), but you can bet your bottom dollar that someone will aggregate all the responses so that they can tell themselves their service is excellent.


The fact is the service I received was adequate. It was as good as it could be. But the transaction was so trivial it could never be evidence of American Express's customer service excellence.


Most respondents will answer this question based on the outcome of their experience; since they got their address changed without a hassle they'll select "excellent" - they won't really mean that, but the head of customer service at Amex can feel good.


Q. Based on this recent experience with American Express, how likely are you to recommend American Express to a friend or colleague?
Please use a scale from 1 to 10 where "10" means "extremely likely" and "1" means "extremely unlikely".



Is the fact a company can change my address a sound basis for recommending them to my friends? Of course not. But again, people will provide a 'halo' answer, and corporate backs can be patted!


Overall, how would you rate the service you received from the representative you spoke with? (Excellent... Poor)


Enough already! Haven't we established that all you did was type my new address into your system. And incidentally, now I think about it, I haven't received a new statement yet so I don't even know if you managed it or not!


Q. Please rate this representative in terms of...

• Having the knowledge and authority to resolve my enquiry
• Listening to and understanding what I had to say
• Communicating in an understandable way
• Spending appropriate time to address my enquiry Making me feel like a valued customer

I won't labour the point further - but, really, who can answer these questions honestly? Are the answers all "excellent" because nothing went wrong with this simple process and I'm happy? Or should I say "fair" - the second lowest possible rating and one that, were everyone to answer in this way, I know the company would be crestfallen with the research results - because this unexceptional event was handled "fairly"?

These pre-coded consumer research questions are really a language all of their own.

I don't know how much money American Express waste on this survey. Of course it won't just be the cost to design and administer it - fairly trivial given that it is conducted on-line. The bigger issue is that, presumably, someone considers the responses and does something on the basis of them.

There is a simple, inexpensive way of gauging the quality of your telephone service. Use statistical sampling (a process which is invariably compromised by non-response in consumer surveys) in a pure form to collect a sample of actual telephone calls (they are all recorded).

Listen to them - or better still get an expert in consumer behavior to listen to them - and evaluate how well they have gone from the customer's perspective.

Much simpler, not compromised by non-response, not open to the vagaries of customer survey language, and far, far more accurate than the post-rationalised 'halo' responses these questionnaires typically illicit.

Philip Graves
Consumer Behaviour Expert