"Data-Driven Thinking" is written by members of the media community and contains fresh ideas on the digital revolution in media.
Today’s column is written by Justin Petty, vice president of global media and partnerships at dunnhumby.
The big data revolution helped to create a new job title: the data scientist. The title, which Harvard Business Review called the “sexiest job of the 21st century,” – became so popular that the ambiguity of the description grew to include everything from statistical analysis to database management – two very distinct and different roles.
It has already been suggested that the title be killed off for this reason. Instead, I believe what the marketing industry really needs is the a different title: “insights artist.”
The ability to amass large amounts of data from various sources is important – and best left to data management specialists. The ability to organize data and use it for targeting and measurement purposes is equally important – and best left to statistical analysts. But making sense of the world – in terms of what consumers are doing, what they want and how they want it, in a way that can then turn all of that into a marketing strategy – requires a special set of skills that isn’t easy to find.
More Art Than Science
Analyzing data is a science. I learned all about it in graduate school, and I can still do linear regression by hand with paper and pencil if needed. What they don’t teach you in school is how to apply all of the critical thinking and statistical analysis to the data to make it come alive. That involves turning data into customer insights, from which a strategy can be built. That requires more art than science, and lots of experience.
A “data artist” can turn data and metrics into a visually appealing graph or chart that makes the data easier to interpret. Thus, infographics have become extremely popular for conveying complex ideas in an easier-to-understand medium. An insights artist combines the skills of an analyst and a data artist, with the strategy of a marketer to go a few steps further and turn it into a brilliant business plan.
Insights artists must understand the business needs and what exactly needs to be solved. They need to see the whole picture from the the data, business and customer perspectives, from the business application perspective and – most importantly – from the customer perspective. This is where all good data analysis begins. Just having massive amounts of data is not enough; it doesn’t automatically turn into insights that can be used. Once the business problems are understood, there is always a matter of choosing the right tools or statistical methods, etc. When the analysis is complete and there are compelling results to share, they have to be boiled down into a few meaningful charts and graphs for the rest of the world to understand.
Metrics: A Key Piece Of The Puzzle
After the logical thinking and investigative work of the analyst is applied to the data, and the informative yet easy-to-understand graphic development of the data artist is applied to the results, a story begins to emerge. This is where an insights artist separates from the rest. Too often, an analyst will do brilliant work only to fall flat when it's presented. While it may be on point and statistically sound, it doesn’t resonate with the audience, which is trying to understand how the data is supposed to guide decision-making.
Knowing which metrics to share is a critical piece of the puzzle. Analysts always want to show how much work they did, because it is a lot of work. This often leads to an endless PowerPoint presentation that has too much “What?” and not enough “So what?” or “Now what?”
Do you need to present ROI as raw dollar uplift, incremental net sales without margin or incremental units sold? These are just different mathematical formulas. What comes after the assimilation of results is more crucial to developing a successful marketing plan. Why was the ROI so low or so high? Did one creative asset work better than another? Most importantly, an insights artist should answer the question: What should we do differently next time to make the campaign even more successful?
How the metrics are presented is part of the art that is applied after the science. An analyst, for example, would say that in the last six months, the number of lapsed buyers of a brand totaled 4.4 million customers. A data artist would show a map that equates that number to the population of Kentucky. An insights artist shows the same information but also decomposes the reasons behind the lapsed customers, along with insight into: Were they brand-loyal or brand-switchers? Which brand did they migrate to, or are they out of the market altogether? Which segments can potentially be won back, and which tactics have previously worked in the past to bring customers back?
Searching For Insights Artists
Of course, finding someone with all of these traits is not easy, especially when the single most important factor to being an insights artist is experience.
In a perfect world, every organization searching for a data scientist would be hiring a former analyst who has spent the last 10 years as a marketing manager. But those people aren’t easy to find. An alternative would be to stop looking for one Renaissance man or woman to do it all – and instead solve this skills/experience combination problem by assembling the right mix of people. Often, the best solutions are derived from collaboration.
When an analyst, data artist, and marketing manager can all come together and understand the value each brings to the equation, success is more easily achieved.
So I propose we kill the data scientist title in favor of the insights artist. In reality, this is most likely a team of traditional roles that organizations have been hiring for years: analysts and marketing managers. The key is to stop isolating them in departmental silos and bring them together, so that each understands the problem, the potential solutions, and the best way to convert massive amounts of data into actionable insights.