Where’s the Art in Data-Driven Marketing?

Marketing today tends to be more science than art, though it’s not at the scientific level of software development and IT. At least that’s the view of more than 600 participants to a study fielded by Chief Marketing Technologist and Third Door Media. The results make some sense given that around one-third of the respondents identified as hybrid marketing/IT professionals as opposed to marketers. Moreover, a strong majority (72%) reported having ever done computer programming in some capacity, even as a beginner/amateur.

Not surprisingly, a solid majority (63%) of the study’s respondents believe that data-driven decision-making is more of a science than an art, though about one-third (33%) find it equal parts science and art.

So what elements of data-driven decision-making are more art than science?

Participants, asked this question, seemed clear about one thing: more than 8 in 10 feel that’s it more of an art than a science to craft a narrative from the data to persuade others. This speaks to the importance of storytelling and the difficulties that some have with data storytelling. It’s also telling (pun intended) that the market research community is prioritizing storytelling as a skill.

A majority (65%) of respondents likewise believe that it’s an art to come up with the right questions to ask. This likely relates to interpretation of data but also the reason for conducting analyses in the first place, bringing to mind a study in which most decision-makers said that human insights should precede hard analytics when making decisions.

It’s also more art than science to visualize the data to better understand its implications (57%) and to take into account what isn’t included in the data (57%), per the study’s participants.

In sum, these results show that the science of data-driven decision-making lies in the numerical analyses and validation, while the art form rests with the interpretation and communication of the data.

Frequency of Experimentation

Delving further into the scientific aspect of marketing activities, the report indicates that virtually all (94%) of respondents run at least a few of their marketing activities explicitly as experiments. As such, this seems to transcend the respondent’s identification of being a marketing technologist (hybrid marketing/IT) or marketer.

There’s less consensus surrounding how those experiments are run, though. Around 4 in 10 either follow the scientific method (“observations, questions, hypotheses, testable predictions, and the acquisition of relevant data to support or refute those hypotheses”) frequently or always, whereas the remainder do so sometimes/loosely (48%) or never (13%).

Interpretation rears its head again here: among the most difficult parts of experimentation cited by respondents is making decisions indirectly related to the test. The report offers as an example the interpretation of what a winning landing page in an A/B test teaches about an audience.

In sum, and in the author’s words, “if marketing is becoming more scientific, it is — at best — a soft science, connected with fields such as psychology, sociology, economics, and anthropology… The truth is that there’s an art to designing and running great experiments, even as — or because — the underlying process demands scientific rigor. Great science is incredibly creative.”

About the Data: The results are based on a survey of 637 participants who self-identified as belonging to the following professions: marketing (59%); hybrid marketing/IT – a marketing technologist (32.2%); software development/IT (1.3%); and other (7.6%).