The Power of Words: how customers use words to communicate their experience and why interviewing captures them so effectively
Semi-structured interviews are the core of qualitative research. They are a powerful tool to collect thick data on customer experience.
Business loves Big Data. And numbers. 87% of top industrial firms believe big data will shift the competitive landscape.Well I guess you can’t argue with that. Numbers are objective, clean and nice to work with – surely, with enough of them you can learn everything there is to know about your customers’ experience of your product.
Human nature is not objective, clean and sometimes not even nice to work with. And because your customers experience your product as part of their everyday, humanlives, researching their experience necessarily requires a method that captures the subjective, messy and counter-intuitive.
Words Do the Job Nicely
People already have an extremely valuable way of communicating their experiences: words. We have refined our use of words over thousands of years to the point that we have an extensive vocabulary that we can use to communicate extremely effectively how we feel, what we want and how we experience the world around us. Why, then, would you want to rely on another form of data – say, numbers – to interpret what we are already so effective at communicating?
Interviewing – more precisely, semi-structured interviewing – is the best way of capturing these words and turning them into data. This is a method borrowed from the social sciences and is the most widely used in qualitative research. The interview is guided by a series of key questions or discussion points that the interviewer wants to explore. But crucially, it is open enough to allow the interviewee to discuss these points in a way that makes sense to them and to explain how they experience the phenomenon at hand.
But How Many Interviews?
Because of the aforementioned love of numbers, it is pertinent to ask how many interviews are needed to make a reliable analysis. In a meta-analysis of interview-based research projects, Mark Mason established the average number of interviews conducted is 31, with some studies do as few as one single interview!
Are the alarm bells in your head ringing?
Surely, a few interviews cannot produce reliable data? How can you know the data wouldn’t change if you interviewed another 100 people?
Qualitative research requires a different outlook. Unlike with Big Data, the data set is not laid out in advance and stuck to. Rather, as applied research Dr Daniel Turner explains, it is an iterativeprocess in which “the investigator should be continuously going through cycles of collection and analysis until nothing new is being revealed”. There is no pre-determined, ideal number of interviews to conduct. Qualitative researchers talk of reaching “saturation point” when they are gaining no more new data from new interviews; they are just seeing the same patterns being repeated so they can stop interviewing, satisfied that they have gathered sufficient, reliable data.
“The investigator should be continuously going through cycles of collection and analysis until nothing new is being revealed”
A Large Quantity of Quality Data
Once transcribed, a 30 minute interview can produce up to 5,000 words. So, if you’ve interviewed 20 people, that’s 100,000 words; which all of a sudden looks like a lot more data points than 20 interviewees. Obviously, though, having a lot of words is not enough – otherwise you could just rely on quantitative data. What makes these words powerful is that they combine to make a comprehensive picture of your customers’ experience.
The value in interviewing comes from its ability to capture the thoughts, opinions, values and beliefs of the interviewee. Collectively, these produce “thick” rather than “big” data about your customer. As a key component of qualitative research, interviewing enables you to hear the stories of your customers and understand how and why your product fits into those stories.
Behavioral Economic Analyst