Data Collection

Digital Marketing - Study Notes:

Qualitative versus quantitative research

When thinking about data collection, it’s important to think about your points at which you are going to collect robust elements of data. And if you do that, you need to think about what ways and methodologies you’re going to use to be able to do that collection.

We talked about the difference between qualitative versus quantitative data, and this is very important when you’re collecting and analyzing, because they both give you different sides of the coin.

Triangulation

One of the core concepts when thinking about what type of data you collect is “triangulation.” And triangulation effectively means using multiple data sources as a means through which to collect both qualitative and quantitative data, and when put together, they add to the richness of it all their own.

So thinking about the data collection process is very important. You need to think about what customers you’re targeting, what kind of questions you’re asking those customers, and when are you asking them. It may be the same customer, but at a point where they’re disgruntled they will give you a very different response to when they’re actually very happy and satisfied with your organization.

So you need to get a really good cross-section across a number of different dimensions in order to make that data analysis and collection meaningful to what you are trying to really achieve and the insights you’re trying to get on that.

Current trends

Let’s think about some of the recent or current trends when thinking about data collection. There’s a number of different ways to basically do data collection, and a lot of the time, some of the more conventional ways seem a little bit stuffy and non-responsive from a customer perspective. It feels like the rule of sending a survey through the post and expecting someone to fill that out both accurately is probably in the past. Some of the most common things we see is people visiting websites, for example, and you’ll be asked a short survey on the back-end of it to see how your experience is.

One of the key concepts around data collection is the point at which you do so. So for example, if you have an experience and then a company asks you for feedback five or six days before, it could be very, very different from when they may have asked you to do that at the outset of that process. So think about what time is the most relevant time to collect that data.

• Mobile market research: Mobile marketing research is a very good one, and one of the reasons why it is great is because not only can you get an understanding straight from the customer in real-time, because we’re always on our phones, but you can also get location-specific data as well, understanding the context. So remember, context is key. You get that wider context as to when someone is responding to you, which can actually act as something quite meaningful over the long-term to determine why someone responded in a certain way.
• Prediction markets and social media research: Both are becoming very, very active. Think about the number of Facebook surveys that you have out there. And again, these are just ways and means to try and, you know, get access to a customer base, based on a relevant channel that they’re constantly using in a timely way, perhaps quite soon after they’ve had that experience.
• Biometric market research: This is another really interesting one, and it’s a way to kind of uniquely identify a certain individual acting in a certain way. Now, why is that useful? Well, think about segmentation. If you can understand at an individual level how someone is responding, it makes it a much easier thing to be able to determine what you want to be doing next with.
• Virtual shopping: This concept around virtual shopping is extremely important as well, because it gives you the ability to kind of test someone in a virtual environment about how they may react in reality. It takes you as far and as close to reality as you can when you’re trying to get these insights is going to be extremely valuable over the long-term.

Considerations

Let’s now start to think about data collection and thinking specifically about some of the key considerations we need to make when collecting data.

• Sample size: Why is this extremely important? If we get the sample size wrong – perhaps too small – the results don’t become generalizable, and if they don’t become generalizable some of the key insights may be amiss. And in reality, if you’re going to make decisions based on the data or insights you collect, you could actually be straying further away from what your customers want, then in reality without doing anything.
• Operational issues: We’ve all experienced the difference between putting a plan down on page and what happens in reality. We need to think about quite robustly how we overcome these operational issues. How do we collect the data in a way that we’re not creating an error in biases? That we’re not misleading people, that we’re asking them at the right time and in a way that we have permission to do so. Largely if we film in public areas, we may not have the permissions to do so, for example, or if we ask questionnaires in certain areas. Again, you may not get the best responses.
• Segmentation: You need to be clear at the beginning of the process exactly who you’re looking to target because, again, if you end up doing a survey on school kids, when in reality your target audience was professionals, you’re going to miss a trick. So thinking about the target audience set up, to begin with, is going to be a very important thing to do.
• Accuracy of the information: This is very vitally important. We’ve all done it before in surveys, a little bit quick or not reading the questions too fully, and we’ve ended up doing it in a haphazard way. Unfortunately, the people analyzing the data have then troubles to distinguish between what was realistic and true with that of people simply just filling it in for the sake of it, so we’ve got to be very conscious and start to think about the anomalies within that data set but it’s always better to actually try and get accurate information at the outset then try and post-rationalize some of the skewed results you may have because of the accuracy element.
• Bias: No data collection process is without its biases and its inherent biases. We need to, as a consequence of that, try and overcome them as best we can. And we need to recognize that they exist and overlay the humanistic or intuitive element on to the data. We will do a bit of a “make sense check” on this, and if it doesn’t make sense, you may have to redo that study to ensure that your accuracy is absolutely spot on. Triangulation becomes very important at this stage as well. So do think about that.