Causality or causation?

May 15, 2019

Outside of purely digital services, such as Netflix, it is very difficult to understand exactly what people are doing and why.

In science this is referred to as the difference between causality and causation. Just because b followed a doesn’t mean that a caused b. Without having detailed and in depth understanding of your customers, you have a very small chance of being right about their purchasing motivation and being able to suggest complimentary products and services, that don’t just seem like buying more similar stuff.

Analysing customer purchasing data as well as a myriad other information sources can paint a picture of a customer. But it is likely, that unless a person shares everything about everything, then that picture has significant and important gaps in it. And presuming a pattern based on incomplete data is unlikely to have satisfying results. Research based on couples indicated, that whereas the presumption is that you’ve lived with someone for a long time, that you can answer questions about that persons preferences on issues that you haven’t discussed, the truth was you can’t. You weren’t closer to the truth than a random stranger would have been, inferring something about someone merely based on conjecture.

For organisations, this simply means that without talking to customers, you don’t know them. It is presumptuous based on limited interaction data, to make proposals or referrals that are simply based on a) what is available to offer and that b) “similar” people have also purchased. Browsing on Amazon looking at baseball bats, the algorithm suggested products other customers had also purchased: a black ski mask and black leather gloves. How does that make sense? And how does that increase customer satisfaction (unless you like the humour)?

Now, yes, we know that you can’t go talk to all of your customers. But when was the last time you actually talked to any of your customers? Throughout all our work with client companies, the scariest job anyone has to do seems to be to talk to actual customers. The act of talking to customers to learn about their lives, preferences and what they actually think about you, without the filter of a questionnaire, is something put off. As long as possible. For two reasons:
1. It is qualitative research and, for whatever historical reasons, it isn’t trusted by companies. If a company has 100 thousand customers, then to get a statistical cross-section you need to get at least 1000 responses. Which means that talking to half a dozen people just isn’t worth the effort.
2. You have to get out of the office to do it.
(1 and 2 are related.)

The thing about qualitative research

Qualitative research on its own can indicate problems, where none exist. Talking to a few people may lead you to over dramatise something they said, because by chance these few people all experienced a segment of the service as detrimental to a good experience. Therefore, the best route to take is to combine qualitative research with quantitative data.

In-depth conversations (not interviews) with real customers, in their home or office, gives you a much broader picture of the context of your service within their lives. It is not only information, but can serve as inspiration because the dialogue with the customer paints a much richer portrait of why they like you / don’t like you, than 5 questions on a webform ever could. Moreover, the tangents of conversation, where they discuss not only your service, but that of competitors or similar functions in other services, can illuminate what you need to focus on next. Problem areas raised in conversations can be investigated further (hello quantitative research) and discussed within the company. After all, the customer is never wrong therefore his or her input can really help you improve your business.

But these conversations are, sometimes, painful. And they are time consuming. And it is true, many customers don’t want to give you a couple of hours of their time. And you have to do it outside the office, preferably 1:1 (or close to that). And after one conversation you just have some information. But you don’t know what is important and what is not. So you need to do several of these time-consuming, difficult conversations. And you only know if you’ve found something important once the answers start repeating themselves.

The good thing about qualitative research

The best aspect of qualitative research is that we have yet to encounter a situation, where it hasn’t been illuminating. More often than not, it has a direct impact on product and service development and is referenced inside the company for a long time. And those individuals that were involved in the conversations with customers, end up swearing by it.

The whole point of the exercise is to understand the difference between causality and causation in your customers’ journeys. Why they do what they do and how, what you do, affects what they do. Because looking a questionnaire answers, while letting you infer something, is about as accurate as a wild guess. Or, as accurate as something a couple that has been married for 50 years would say about each other, if they haven’t ever talked about it.

It isn’t as hard as it sounds

If the above makes you tremble in your sneakers, that talking to people is terrifying, then take a deep breath. And relax. While the first conversation may be scary, it will also be highly rewarding. Both in terms of information but also for social reasons. It is interesting to meet people you’d normally not meet and to get a broader view on life. So like Nike said – just do it!