Throughout my marketing career, I’ve noticed a question that marketers, both on the client and agency side, frequently face: “Does that generate numbers? Does it increase revenue? Where’s the proof?” Marketers typically respond: marketing is about touching insights, being creative. It can’t be purely about numbers.
With the rise of digital marketing, e-commerce, and the beginning of the digital transformation wave, the use of data for decision-making and performance evaluation, especially for marketing departments, has shifted to a higher, more detailed level. While companies previously relied on market research, brand health studies, or social listening tools typically on a yearly, quarterly, or monthly basis, data usage now happens daily, with more diverse metrics for measurement and optimization. I’ve been applying data to planning, execution, and optimization across multiple time scales (real-time to weekly, monthly), and I’ve found that data usage has significantly contributed to improving processes, quality, and effectiveness, even in areas traditionally hard to quantify, like content creation.
However, using data for decision-making also has its obstacles. The data we have is typically historical (even real-time data has some lag), and we use this past information to understand the current situation (Descriptive data analysis), to analyze and find formulas for success or reasons for failure (Diagnostic analysis), and perhaps at the highest level, to predict future trends (Predictive analysis). Yet in complex, ambiguous situations, no single predictive model can encompass all qualitative and quantitative factors to produce accurate predictions. Meanwhile, our intuition, which naturally works as a predictive mechanism and can incorporate both qualitative and quantitative factors, can be a remarkably capable prediction model for such ambiguous situations. Of course, intuition has its own problems with bias, data legitimacy, etc., and most importantly, we don’t fully understand or control how this model is built.
In an Advanced Data Analytics course I’m taking, the Google Chief Data Scientist had advice for us during exams: “After finishing, review your answers. But if you’re still not sure after reviewing, the answer you chose first is likely correct, because our intuition is usually right.”
But in a podcast on the YouTube Channel “Diary of a CEO,” I heard a former CIA officer discussing human behavior advise: don’t trust your intuition immediately. Rely on the information you’ve gathered instead.
These two opinions seem contradictory, but looking closer, I noticed a common thread: both are very cautious about the information source that appears to be opposite to their strength. The deep data professional like the Chief Data Scientist still leaves room for intuition when data isn’t sufficient for a decision. The behavioral psychology expert who works primarily with intuition clearly recognizes the potential for bias in his own data source and carefully cross-checks with objective data.
So the right question isn’t “follow your heart or follow your head?” Rather, I think we need to consider that intuition itself is a source of data, but data that’s been processed and even modeled through methods we may not fully understand.
Then the answer becomes another question: “How do we effectively combine data and intuition?” A big and fascinating question that I’m still working through, practicing, learning from experience, and learning from friends and colleagues. I hope to be able to share more soon for discussion.
