We spoke to Dr. Hsin-Hsuan Meg Lee, a professor here at ESCP Business School specialising in self-presentation and digital identity. She is currently teaching a class called Creative Analytics. Her degrees in zoology and animal behaviour formed the bases for her scientific research training. Indeed, when she made the transition to marketing, she carried the habit and love of observation with her. She maintains that this is also why observation is still her favourite research method, a method on the rise since the emergence of the digital culture.

Tell me a bit about your Creative Analytics course? Can marketing analytics really be creative?

The Creative Analytics module is not about the debate and paradox between creativity and analytics as is often portrayed in popular media.

Creativity is a crucial element in marketing analytics. We don’t have to think of it as either the creative route or the analytical route; the two work hand in hand. So in order to make analytics work to achieve your goals, you’ll need to seek out a creative approach to gather relevant information, organize data, develop the research model, interpret the results, and to finally come up with the implementation plan. At every step of the way, managers need to be creative in finding the best research tools and making decisions in prioritizing these tactics (as analytics often involve compromise).

While the statistical process itself can also be really creative, the Creative Analytics course really emphasises the managerial decisions taken along the way. It’s basically an analytics course (almost) without statistics.

When and why do companies need to consider marketing analytics? Are smaller companies or startups who may not have the resources really losing out by not conducting more research?

To answer this question we need to define “analytics. “For me, marketing analytics is the process of gathering, organising, transforming and interpreting data in a way that can assist companies to make better marketing decisions. It’s important for all companies to consider marketing analytics, if they value marketing.

That said, analytics vary greatly in their scope and scale, and companies may have different levels of capacity. Indeed, some companies may be concerned about the resources required to take on analytics. Companies shouldn’t conduct research for the sake of doing it, but if data is just lying around, it’s a waste not to use it.

In your opinion, is there such a thing as conducting too much research? Is it okay to take a risk even though the research to back up the idea isn’t there?

I believe all analytics and research involve a human element. Managers eventually are the ones to decide what goes into a research assignment, a model, a survey, or an experiment. What should be measured and how should the data be interpreted and implemented are all determined by managers. There aren’t any faults in science, as data is just data.

You can never have too much data. Information is the key to any business. That said, too much information can potentially jeopardise the decision-making process. I think there is danger in being overly data-driven and ignoring the fact that all analytics should be managerially relevant.

Consumer research should be complementary to creative ideas within a firm. Insights often come from observations and reflections. Observation is a form of research, whether it’s done scientifically or casually. The reliability of the results may be different–the risk level might differ, but observation should be taken into account all the same as it speaks to the creativity of companies.

Do analytics always translate to real-life consumer needs? How do you balance what analytics are telling you and common sense, if these two are vastly different?

I don’t believe analytics always translate to real-life consumers’ needs. The quality and the reliability of the results are influenced by the rigorousness of the process. Most of the analytics are based on samples. Big data–large volume of data–expands the sample size to such a degree that we could consider it the closest to the reality we’ll ever get.

However, from raw data to analytics results, there are still several steps to take, such as organising the data and constructing the models. Every single decision made in the process should influence the results. From here, analysts will indicate how certain they are about the results, which can never reach 100% unless they were able to get the whole population in their analysis.

So yes, there will be times when the results may seem to contradict what you believe to be true, which all comes down to the managers’ call. You can choose to trust the predictions or not, it’s all about taking risks.

Do you think you should come up with the idea first and then use analytics to support it, or should the analytics drive the idea? Or both?

Blindly diving into data is risky, if not suicidal. I’ve seen many results be manipulated to fit a conclusion. Just like any decision made in business, you should never start projects without knowing the objectives.

If the purpose is to explore, then indeed keep an open mind and let the analytics drive some ideas. These exploratory studies are important from time to time. If managers already know everything, then there would be no failing businesses, right?

But marketers need to be careful with how they use exploratory studies and make sure they conduct this kind of research if it fits the goals of the research. If there are any assumptions established from the start, the objective of the research should just be set accordingly.

Too often marketers start a research project with a pre-existing idea that they are seeking to prove and use the wrong research tactics to test this idea. Typically, in these scenarios, they will use exploratory methods, which end up being totally inefficient. For example, rather than choosing to run A/B testing to prove a certain idea/campaign might work best, a manager might choose to run an exploratory study instead. Managers will then mold these exploratory analytical results to fit their pre-existing idea, or just ignore the results (which is a complete waste of time).

How have web analytics impacted consumer research? What are the most important metrics to look out for?

Web analytics provide a great opportunity for companies to observe consumers. There are of course other ways to understand consumers, and while web analytics should play an important role (depending on the business), they should never substitute other legitimate methods.

I don’t think there’s one single metric to look for. What is important for companies to measure greatly depends on their purpose, the objective and the capacity of the companies. I think believing that there’s a magic formula is a dangerous belief.

Can you share an example where conducting research has led to an impossibly creative marketing idea?

The most well-known successful example is probably Netflix. And the most famous failing example is probably Google Plus.

I think real success often lies in the decisions made along the way. It’s those daily decisions where you can see the most obvious use of analytics. These decisions include whether or not to drop the print catalogues, which customer groups to focus on, which keywords to include in campaigns, which headlines to use in the newsletter etc. Creativity often hides within these baby steps.

What are your favourite free analytics tools?

I don’t really have one favourite free analytics tool. For different purposes, I use different things, but most of them are rather academic oriented. I’m slowly moving towards using R for most of my own academic analytics, which is free but may not exactly be so user-friendly.

Interested in how to become a manager who understands how to implement creativity in marketing analytics? Check out how ESCP Business School and its Marketing and Creativity programmes, which will equip you with just the tools you’ll need!