Simon Chadwick

We talk to Marco Vriens about The Insights Advantage

SC: Your latest book, The Insights Advantage: Knowing How to Win, talks about insights. How would you define “insight”
MV: There are different definitions: an insight can be a fact, a thought, or an analysis of facts, but it must increase your understanding of a particular situation and do so in a way that creates some type of energy to act. So, for example, I have an insight about a customer need and decide I will make a particular product or realise I have thought about this whole product area in completely the wrong way. Even if you’re not putting a new product on the market or making a price change, if you start thinking differently about a situation, that has behavioural consequences – you will look at different data or talk to different people. And that energy to act is an essential component.

How do you actually behave to discover an insight?
To produce more and better insights, we need people, tools and processes. I outline a process for interactive discovery, and that process can drive behaviour. The process starts with a problem definition. The next step can be a specific market research project. The process, though, is focused on discovering insights, not on finishing one particular project. So, following the market research project, the process directs you to look at all the available data, you analyse the data across multiple data sets and apply advanced analytics. Interpretation happens throughout the whole cycle, but after all analyses you must ask yourself what it means, how does it all come together, and the “so what?” and “now what?” questions.

And then you enter the last part of the discovery process: checking to what degree the insights are validated, asking if you know enough to act, are the insights compelling, and is more research and work needed. Many research departments are very project driven; they identify a project, get the funding, scope it out, hire a vendor, the vendor does the work, and produces the data which must be presented, and then they move on to the next project. But you need to spend time looking across studies and data sources, as it may not be feasible to solve a particular problem with one data source or one project. People are so project driven that they don’t have the time and patience to look across projects, which is really important. Most success cases that I discuss in my book – where market research had a huge financial impact on the performance of the firm – focused on identifying a set of insights that really changed how the firm competed in the market. In many cases, this took several connected projects, multiple parts of analyses and validation, and it was not unusual for it to take several years before the breakthrough insights were discovered and implemented.

Increasingly, it seems that clients are asking researchers to consider multiple data sets, historical and secondary data. Now we have the phenomenon of big data. Where do we fit into all of that?
Big data can mean different things to different people. I make the distinction between two big data situations.

One is where you have many different data sources, all of which can give you some insight into a particular problem area, and this situation has really become much more common over the past ten years. I recall early in my career that doing a market research project generally meant doing just one survey. Now that scenario is not that common at all. In addition to the survey, we may look at other survey, qualitative and, of course, increasingly big data, and very often data from Google search. At GE Healthcare, I had a situation where I could look at 100 cross-data sources on top of all the surveys to provide insights around an area, and it was quite usual to look at four or five data sources for any particular project: internal interviews or quality data, a market research paper and secondary reports from healthcare market research suppliers.

Big data is challenging, because decision makers and marketing directors have to make decisions – but when there’s 50 or 100 plus data sources, where do they begin? Plus, now I think there is data available that many don’t even know about, or don’t know how to pull or use it. It’s overwhelming, and often neither they nor the market research people have time for that. There is so much data available now that it really makes sense to look at these before anyone scopes out a new survey. One of my clients called this “secondary first.” And that’s a great idea, because it may reduce wasteful spending and it will lead to more and better insights.

The second aspect of big data is where you get enormous amounts of data, often from online sources. Clickstream data are usually described in size and in velocity, because clickstream data, Google search insights, Twitter tweets and Facebook mentions create an ongoing stream of real-time data – as opposed to a survey, where you get your insights a couple of months down the road.

From my research, I think there are a couple of areas where big data offers big advantages to marketing research. One is marketing mix modeling, which has probably benefited the most. The second is trend spotting, which previously was not really feasible for marketing researchers and was often done by domain experts who used online and media information and their domain expertise to identify trends. Now, with big data and Google search insights, they can pull vital data online and analyse it. It does require advanced analytics, but if you are willing and capable, this can help find significant trends in a particular industry, and that is a gigantic advantage to anyone in market research.

A third area is analysing Twitter, Facebook and other social media data, looking for customer complaints, new product ideas, anything connected with a crisis. You can rapidly see what people are saying online via text analytics, and that’s another important element for market research.

What does this mean for the skill sets and type of people we’ll need in the future?
Big organisations with a sizable market research department have the resources to have a mix of people and specialists in certain areas. But there will be more demand for what I call the ‘versatilist.’ Versatilists are not super-specialists, but have a sub-level of expertise in different areas – the traditional market research areas, psychology, advanced analytics, IT, etc. I know many people working in market research who have little understanding of analytics and rely on their vendors, but if they’re going to use analytics more, they need to be more knowledgeable – otherwise they’ll be sold inadequate products and will make significant  mistakes. On top of that, they will not see some opportunities, because they’ve never thought about it. If they’ve never looked at marketing mix modeling as an interesting area, how will they know what to ask their vendor? Even if they ask for a fact sheet proposal, they won’t know whether or not it’s correct.

In your book, you underline that you need additional soft enabling skills to translate the insight into action.
Softer skills such as building trust between the stakeholders and the research department are vital, and this plays out on the individual as well as corporate level. If you build trust with stakeholders and other teams, they will share information more freely, be more likely to act on your recommendations and help you to identify the right problems. Building trust is also driven by hard skills. Solid statistical and marketing research skills provide some credibility, but also, if I take you to a meeting, I need to trust you will not embarrass me and you will provide the right advice so I’ll make better decisions. A lack of trust can derail a lot of opportunities.

Communication and follow-up are very important. If you want people to act on your insights, it’s not enough to present and disseminate them. For selected insights, you have to keep engaging with your internal decision makers and ask, ‘What are we going to do? What decisions are we going to make?’ You need to get them to commit to something, and then check if they are doing it. And that is time consuming.  

Even if people are happy with a concept or presentation, they go back to their day-to-day activities and other things come up. You have to be aware of the issues, mitigate and tackle them. You need to be an evangelist. If you believe that your insight is true and has value, you have to be committed, take the time to convince people and understand the political environment. It’s not always a straight line from insight to action – sometimes there is a political reason why something cannot be acted upon immediately, so you may need advice from your superiors or people in your line management. You have to pursue all the avenues to get the organisation where you think it needs to go. 

Dr. Marco Vriens is managing director strategic analytics and SVP methodology at The Modellers. He has led analytics, research and insights teams at Microsoft, GE Healthcare and various marketing research firms. His latest book is The Insights Advantage: Knowing How to Win