Marketing analytics very often refers to the set of metrics related to digital marketing. Alternatively, the term can also refer to using data and analytics for marketing purposes, that is to explain and showcase a service or product’s value proposition. The latter approach is discussed in an excellent paper by Dr. DJ Patil: Building Data Science Teams and it is the path I will follow in this article. By using a number of examples, let us see how data and analytics can boost the external image of an organization.

If you have a data-focused job, this article will provide you with useful ideas to put into practice and help your team be more visible from the outside of your organization. To this end, I have tried to cast as wide a net as possible and select examples spanning different business contexts across several industries. Hopefully, some of these ideas will be applicable to your own type of analytics activity.


It is both easy and hard at the same time. Easy because most people can relate to the topic at hand, and hard because most questions have likely been answered before. Consequently, it can be difficult to come up with anything new. The first example is taken from the aforementioned article by Dr. Patil – the OK Cupid blog, which uses data collected from the platform’s users to answer all sorts of questions related to dating (What makes a good profile picture, flirting trends in 2016, etc). Several elements make this blog very interesting to read, such as the unique dataset, the question-based approach to analysis along with a touch of humor to top it off.


This can be equally captivating as it sheds light on things and activities you would not normally have access to. The 2005 book Freakonomics by Steven Levitt and Stephen Dubner discusses subjects as varied as the economy of drug dealing or how data can be used to uncover cheating in sumo wrestling. The book and its sequels made it to the New York Times best-seller list, and the Freakonomics websitecontinues to tackle off-beat topics – which is in part what keeps the Freakonomics brand fresh and interesting to read.


Science is another realm which is accessible only to a small segment of the global population. Five years ago, the discovery of the “God Particle” – the Higgs boson – received a wave of well-deserved attention for a number of reasons. First, it elucidated a long-standing question about why things in the Universe have mass. Second, on the technical side, building a multi-billion-dollar machine is no “piece of cake”, and it took an impressive amount of effort and coordination to realize this feat. The same can be said about the data science work behind this discovery: the CERN laboratory is among the leading Big Data powerhouses worldwide.


Nate Silver’s FiveThirtyEight blog takes a data-driven approach to cover sports, politics, culture, science and economics. A project that greatly helped to spread its popularity was the quasi-real-time coverage of the US presidential elections. The forecasts were updated daily as new polls were made available. By the end of last year, nearly one thousand articles covered the 2016 election.


This notion is not to be underestimated. People crave for fresh news and love to see the action as it happens. As an example, flightradar24 shows the passenger planes in the sky 24/7. As I am writing this, there are more than 10,000 planes tracked live on the world map. And while I have no real use for this knowledge, it feels quite interesting nonetheless! Beyond transportation industry, there are plenty of other websites that track real-time changes in financial markets, economics, society or environment.


An old quote says that “a picture is worth a thousand words”. Indeed, finding the right picture can make a significant difference in how your story is perceived. Then, data visualization can be a powerful enabler for Marketing Analytics. Consider the distorted maps at worldmapper which take a different approach to comparing country metrics around globally. A surprising effect is produced whenever a metric comparison goes against the country size ratio which is somewhat hardwired in our brains. Countless new ways of visualizing data appear every day, fueled by a combination of new technologies and creativity. For inspiration, check out David McCandless’ informationisbeautiful website or simply browse through the extensive d3.js gallery.?


This notion also contributes to the visual appeal of data. An interactive chart allows the user to see only the minimum necessary and proceed to explore the data in his or her own way. For example, the Observatory of Economic Complexity at MIT enables you to track international trade between countries. Just in case if you were wondering how much wine France exports, look no further!


Bringing in text analytics can further increase the “buzz” effect. Popular examples include the analysis of the State of the Union speeches, tweets, historical texts or even popular song titles. In another example of social network analysis, Facebook Likes have been shown to predict several personal attributes of the users: from gender to religion, political views or relationship status.


More recently, neural networks have become increasingly popular and are applied to a variety of use cases. Facebook research team maintains a blog with regular posts on Artificial Intelligence (AI) and machine learning topics, and Berkeley AI team showcases progress in image recognition.

The examples presented so far illustrate how data activities can support Marketing Analytics. They are by no means exhaustive, and many other options are available to your organization. If you plan to include analytics in your Marketing strategy, here are some interrelated questions to consider:

◊  Are you sitting on a unique dataset – which nobody else in the market has? If yes, any type of analysis (complex or not) can yield interesting results. If not, all is not lost: several of the previous illustrations use publicly-available data.

◊  Does your dataset directly capture unique behavioral elements of your consumers? (think for example of wearable devices or last year’s Pokémon Go craze). Knowing how and where people spend their time, or what they buy, when, and why creates stories which are very interesting to the wider public.

◊  Is your data covering a less-familiar angle of our world? If yes, there is a novelty factor which could potentially kick in and help your story propagate wider.

◊  Is your data fast, i.e. does it change in nearly real-time? This may be an opportunity to provide “fresh news” on a continuous basis through real-time analytics.

◊  How do you plan to visualize your data? Doing this interactively allows users to feel they are part of the story, which stands to generate a higher level of user engagement.

◊  Will you be using best-in-class tools and techniques to analyze your data? Text, image or video analysis via machine learning techniques can generate stories which were simply impossible to conceive five or ten years ago.

Finally, what about cases in which data volume and variety are minimal (the so-called “non-interesting” datasets)? The traditional marketing techniques still apply: you can, for example, run a survey to understand how people feel about your product or the needs it attempts to address. The advent of social networks has made it very simple to reach hundreds or thousands of people, and the barrier to collect valuable data is lower than it has ever been.

In conclusion, Marketing Analytics can be a valuable avenue to pursue anyone working with data. Any such exercise should, of course, be under full agreement with all associated ethical and legal considerations – such as the user’s privacy rights, data protection, etc. When properly thought through and executed, Marketing Analytics has the potential to generate a significant buzz and substantially increase the visibility of your organization.