Interview with Michael Haenlein, Professor of Marketing at ESCP Europe, Associate Dean of the ESCP Europe Executive PhD Program and Director of the ESCP Europe Research Center on Big Data.

The ways big data is impacting today’s business climate may be too numerous to count. The ability to access oceans of data relatively easily is a fairly new but massive phenomenon. And while many are excited about the capability, few are seriously examining how to leverage this data to exercise the core function of any business – sell a product or service in the marketplace. Few areas are as greatly impacted by this development as marketing and customer relationship management. But how does having so much data about customers improve marketing decisions for companies?

There are currently strides being made in answering these questions. Certainly, cloud-based solutions are available that can help businesses leverage data to lower their Customer Acquisition Cost (CAC), for example.

Studying these emerging issues quite intensely is expert Michael Haenlein, Professor of Marketing at the business school ESCP Europe, Associate Dean of the ESCP Europe Executive PhD Program and the Scientific Director of the ESCP Europe Research Center on Big Data. Professor Haenlein’s research expertise lies in customer relationship management (CRM), social media and marketing research. Specifically, he analyses the relationship between social networks and customer profitability.

Tharawat Magazine sat down with Professor Haenlein to discuss the emergence of big data, how it is currently being used both correctly and incorrectly, and what businesses can do to turn information into profitable marketing strategies.

What are the greatest changes big data has brought about in Marketing?

It has really very fundamentally shifted the way marketing managers make decisions or can make decisions. Historically, the job of a marketing manager was to make decisions essentially within a vacuum of information. This is why some managers were better than others simply because they were better than others in making decisions intuitively. This fuelled an accelerated the whole market research industry which used to be very slow-moving because gathering information was such an arduous task.

But today our decisions can find support in so much data. Instead of deciding which single piece of information you need, it is about deciding which piece of information you should focus on out of this sea of data that you have available to you.

But, how do you narrow the scope? Where do we start?

Let me tell you what you should not start with. Don’t start by saying, “This is big data. This is a big thing. I have to do something with it.” Because it is to some extent a very fuzzy cloud of concepts that are summarized under big data. And just thinking you have to do it one way or the other is not the key.

The right approach would involve tackling an actual problem that you currently face. Maybe some form of a real decision that you think can be aided with some element of data and then deal with that specifically.

For example, you may own a restaurant or a hotel, and you may observe that Trip Advisor is massively important in determining how customers frequent or how they react with your establishment. So, you may decide that you want to systematically analyze Trip Advisor for you and for other restaurants and hotels in the same area. You can start getting into how to download this data in large scale, how to analyze it using text mining for instance. This can help you to determine how much one additional review is worth and how strongly you may want to incentivize such reviews. Always with a very strong focus on the bottom line. Because the point is not to get as many likes as possible or as many shares as possible. The point is to earn more money.

In your experience, how easily do people actually accept what the data has to say?

I think there are three separate problems. The first one is a basic psychological problem. That we always have the bias to believe that what we know is true, and the new information needs to be challenged. This is known in psychology as the confirmation bias. A second issue is that marketing managers historically were not really used to working with a lot of data. Mostly because that data was not available. So, there needs to be a move from your gut feeling towards some data-based decision making.

The third one involves our reluctance to trust machine learning. For example, deep learning or AI can be used in medical research in order to automatically screen x-rays to identify whether people have cancer or not. Now the problem that comes with this is that patients are naturally extremely reluctant to trust the diagnosis of a computer that nobody understands. In the business world, the challenge that marketing managers now have is to make decisions more based on data. But on the other side get a very solid understanding of what this black box means and up to which point they think they can trust it.

So as a business owner I have a marketing decision that requires more big data analytics, what kind of people do I need on my team?

The very first thing you need to do is to get a very basic understanding of big data yourself. I’m not saying that you need to go into the statistics and understand how a deep learning network works. But, if you take the example of Trip Advisor I gave earlier, probably just spend a day browsing through the posts and get a feeling of what type of reviews people post, what type of comments do people make to reviews that other people post. And it just needs a day or two to get immersed in it.

Step two would be to actually do the analysis. And for that, most of the time you will need, at the beginning at least, I guess some form of consultant. In many cases it doesn’t make sense to build these very specific skills in your firm if you don’t even know if you will ever use them again. Some of the work that I personally do deals with managing those pilots in marketing decision-making. Identifying the best method that could be used and teach specific people in the firm how to do it themselves.

But, the third step that many companies often forget is there needs to be a translation between the output of a data analysis and the link to an actual decision that you can act upon. Because big data output is complex. I can analyze one million Trip Advisor reviews for you, and the outcome, a cloud of words, would be intimidatingly complex to anybody who is not an expert in the field. At some point, you need someone in the firm or a managerially oriented consultant that steps back from all this data and all this information and just extracts what it actually means. Remember, it’s not about the findings or about the data. Ultimately, it’s about the decision. Another part of my personal work deals exactly with translating statistical outcomes into a language managers can understand.

We’ve spoken about how big data can help us in assessing, feeding our decisions and driving our decision making. How can it help us in measuring the impact of what we’re doing?

I think at this stage most firms are actually struggling with the whole issue. This is not a family business struggle, even the largest firms still struggle with that. And in that struggle, where many companies start is by measuring things that can easily be measured. They focus on what usually you could call an engagement measure. A ‘like’, a ‘share’, a comment, a retweet, whatever there is. But, ultimately, no firm is interested in engagement. Ultimately, firms are interested in revenue and profit. There are approaches that you can use. And I’ve developed some of them myself of how you can try at least to assess which type of profit impact certain types of marketing efforts can have.