Tom Seymour’s thoughtful piece on CSI in AM (Car dealers and CSI surveys – how much would you give them out of 10?) contained some quotes of mine. I would like to clarify those views and to provide some of the detail behind them.

It is a fact that a very large, long term, survey of customer behaviour in a number of industries showed that the correlation between customer satisfaction and loyalty was 0.11.

To be entirely clear, a correlation of zero is none. No correlation at all. A correlation of one represents a direct and complete relationship between the two. So 0.11 is as near as makes no difference, none.

There are very good reasons why this survey is reliable. Firstly it was carried out scientifically, subject to independent validation, its numbers were large and therefore statistically relevant and it was long term. Secondly it just makes sense.

As Jim Saker said, satisfaction is a very weak emotion. It is also difficult to define. One person’s satisfaction is another’s dissatisfaction and yet another’s delight. So even if there was a link, you would not know what to do about it.

And think about it. If I do what you ask me to do and am well paid for it, will that make you loyal? I’ve just done what it says on the tin, nothing exceptional. Why would that create loyalty? I have to do more things – make it simple to do business with me and develop a relationship that engages the emotions beyond the process of delivery.

This means understanding the customer fully. What they need the product for, how they plan to use it, what they value and what they don’t, how they like to be approached and other things that it would take too long to rehearse here.

In short, you need a plan for every customer.

The point of the latter in car retailing is that you can have a plan for every customer. You have a small enough number of customers and comprehensive knowledge of the small number of products and therefore you can.

The big supermarkets would love to have a plan for every customer. But they can’t, it’s too big a job. So they try to approach such a thing by using Big Data and sophisticated algorithms and analytics.

Achieving loyalty has to start with avoiding dissatisfaction. Getting it right first time, on time, every time. This and only this allows you to move on to developing the sustainable relationship that leads to loyalty. It’s a long term thing. And that’s what you want. Still servicing the customer’s car after ten years. Or getting the next purchase in three years. And not even considering going anywhere else for anything to do with mobility.

This requires investment in process and people improvement over the long term. Investing in trying to generate “delight” or satisfaction is a complete waste of money. Any effect, if there is any, is transient, does not lead to loyalty and does not change for the better the way staff handle customers.

Now to the CSI programs and the scoring systems

Any score resulting from questions asked of customers is subject to a margin for error. So a mark of 75% (say) is in fact 75% plus or minus something. The “something” depends on a number of factors

If the sample is small, the margin for error is large
If the questions are asked after a variable passage of time, the margin grows larger
If the question method is inconsistent (different people, different approaches, different times etc) the margin grows
If the outcomes are not corrected for extraneous factors (the customer is actually complaining about the product, not the service) the margin grows
If the scoring questions are predetermined (the score is not influenced by what the customer actually wants to talk about) the margin grows

So that 75% score could be plus or minus as much as 25%. Which makes using the scores for ranking purposes, completely unreasonable. And any reward system based on rankings, or on being above a target figure (e.g. 76% = reward, 74% = none) completely unfair.

And the score is fundamentally affected by how each dealer manages the process. Two identical businesses, one which ploughs significant resource into managing the score and one that doesn’t, would get wildly differing scores. But both handle the actual customer experience the same.

In any event, what is the point? Surely the point is to monitor processes and improve them in order to waste less in delivering the product or service (reduce cost) and to deliver value (improve retention and therefore profits and the business).

If that is the case, what is the point of a CSI program? Consider a simple example. You are entertaining some friends and acquaintances. You notice the absence of one of them and think you hear the sounds of someone being sick in the downstairs ‘facilities’.

On the return of that person, what do you do? Surely you act in a solicitous manner, see what you can do to put things right and try to get to the bottom of it as quickly as possible to ensure no-one else suffers the same fate.

Surely what you DON’T do is ignore the situation and two weeks later commission a complete stranger to get another complete stranger to ring your friend and ask them if they were completely satisfied with the evening’s entertainment.

In the process of providing product or service – selling and/or servicing a car – someone in your business knows if all has gone well, or not. The time to do something about it is then. Real time. While the experience is fresh and the facts available. Not weeks later through some third party.

And if you want to make a real difference by improving processes and the customer experience, you and your team should be measuring that for all the touch points, not just a sample. And certainly not a self-selecting and inconsistent sample. That way you can get a real picture of what needs fixing, with facts, in the right order of priority.

The simple truth is that the CSI system as currently operated is useless, except perhaps for those whose network consists almost entirely of dealers who actually do nothing themselves to measure the effectiveness of their processes.

If you need a CSI program to understand where things need to improve, then you don’t understand enough about your business. And you won’t learn any more about it reading CSI reports on a computer. You should be going to see what happens, where it happens, when it happens, with the people who make it happen.

Author: Piers Trenear-Thomas, consultant