The Cutting Edge of Marketing Data Analysis. The Real Reason Marketing Precision Isn't Improving? (Part 2)

• The Cutting Edge of Marketing Data Analysis: The Real Reason Marketing Precision Isn't Improving (Part 1)

In recent years, companies across all industries have increasingly turned to data analysis for their marketing efforts. While the stage of implementation varies by company, a clear gap in results is emerging between "companies that are effectively leveraging it" and "companies that are not seeing results."

To explore the reasons behind this, we interviewed Hiroyuki Sato of DENTSU CROSS BRAIN INC., a data scientist with extensive experience. We heard about the latest trends and challenges in marketing data analysis from the field.

Even with infrastructure and systems in place, failure to deliver results stems from "inability to grasp data correctly"

Q. In the first part, you shared several reasons why companies struggle to achieve results with data marketing. Many struggling companies suffer from "narrow vision," potentially failing to "view metrics holistically." Furthermore, you mentioned cases where companies misinterpret data because they don't understand the meaning or nature of the metrics they handle.

Because of these reasons, a company offering subscription services misinterpreted the cause of a decrease in cancellation rates when the pandemic hit...

Sato: The same principle applies to website conversion rates. In Google Analytics, the conversion rate uses sessions as the denominator and conversions as the numerator. For example, consider an e-commerce site. Even if a new user visits that site, it's rare for them to make a purchase immediately. Typically, they'll browse other sites, compare prices and related products, and only then decide what to buy and where to buy it. So, while an increase in new users boosts session counts, the number of conversions doesn't rise proportionally. Consequently, the conversion rate actually decreases. Therefore, focusing solely on conversion rate fluctuations and devising countermeasures based on that won't lead to effective strategies.

Q. I see. Even if you chase the data, if your interpretation is wrong, it naturally won't lead to improvement. Before speaking with Mr. Sato, I thought the "big data analysis" trend had run its course a bit, and companies implementing it had accumulated experience, perhaps entering a second phase. But is it actually the case that "properly capturing data" is still insufficient in many instances?

Sato: It's true that progress has been made in "starting to track data" and "building the infrastructure and systems to collect necessary data." However, when it comes to whether companies are truly capturing the data they're looking at correctly... I suspect many are still questionable. When we receive a consultation from a company, we start by re-examining the given conditions and the identified challenges.

Essential "Qualities" and "Attitudes" for Driving Data-Driven Marketing

Q. So, what qualities and attitudes do you think are required of marketing personnel?

Sato: Well, this is purely my personal opinion, but simply put, I think it's about having a logical way of thinking, combined with curiosity. Above all, I believe it's crucial to have an attitude of genuinely wanting to understand things. Taking a genuine interest in questions like, "What exactly is the conversion rate in Google Analytics?" or "What does that number actually mean?" And then being able to think logically about it.

But this isn't limited to big data analysis or digital marketing, is it? To use the example from Part 1: when you see changes in churn rates, isn't it crucial to think logically about "What factors could be driving this change?" and "Why did this happen?" For instance, suppose the click-through rate for an email newsletter improves. Do you interpret this as "This newsletter is becoming more effective"? Or do you consider "Maybe it's just taking traffic away from other channels?" and check the overall traffic flow? That difference is significant, isn't it?

Q. I see. So it's about how much imagination you can apply when looking at the numbers. In marketing, storytelling ability is also said to be important, but it comes down to whether you can imagine what developments might arise from there.

Sato: Hmm, calling it imagination feels slightly off. To reiterate, it's about "whether you're thinking logically." When a phenomenon occurs, do you thoroughly verify all possible causes? There are actually numerous correlations where "if you look at this metric, you must also examine this other metric or this movement." By verifying each one, you naturally branch out into various scenarios.

When someone without prior know-how stands at the threshold of seriously starting data analysis, they often take courses like statistics or programming for analysis. But no matter how much they learn, if they don't cultivate the ability to think logically, I think it's difficult to truly master using data.

By consciously aiming for "logical" thinking, marketing insights become visible

Q. I'm starting to wonder if the meaning of "logical" I had in mind is different from what you're talking about, Mr. Sato. My understanding of "logical" meant having objective criteria for judgment—deciding based on "because many people support it" or "because it has higher ratings," rather than "because I think so" or "because this looks cooler." But your "logical" seems to go deeper.

Sato: I'm just saying whatever comes to mind (laughs), but I do constantly think about what kind of attitude "logical" entails. For me, "logical" means

1. All stakeholders present have a clear understanding of the definitions of the terms used.
2. When advancing a discussion, stakeholders clearly understand the premises being set.
3. It means aligning definitions, premises, and common sense with the results of the current data analysis, then organizing everything so there are no contradictions.

That's what I think it means. And honestly, I suspect many people surprisingly neglect this. What exactly is the definition of the terms we're discussing right now? And what are the premises of the discussion? If you proceed with the discussion neglecting those points, it can never truly be "logical."

Discussing Google Analytics conversion rates without knowing their definition can never be truly logical. We often use the term "churn rate," but there are actually multiple ways to calculate it. When someone says "churn rate is decreasing," does everyone in the room assume the same calculation method? Are we proceeding with discussions based on the premise that "churn rate is decreasing" without aligning that understanding? In some cases, the very way the numbers are constructed could be wrong.

Of course, no matter how "logically" we think, we can't possibly understand everything. Marketing really is incredibly difficult, isn't it? Because you're dealing with "human hearts," it's tremendously challenging. Making marketing decisions under such circumstances must be extremely unsettling. My desire is to provide even a little reassurance in that space. On top of such data analysis, we apply imagination, crafting stories and scenarios. I believe data scientists like myself are ultimately just the foundation supporting that story-building, imagination, and even wild speculation.

We spoke with Mr. Sato across two parts. It's said that the spread of big data analysis has made "things previously unknown" understandable. However, if there are situations where the interpretation of numbers is mistaken, or where the same figures are interpreted differently by different people, then no matter how much data analysis is done, it leaves significant doubt about whether the "next move" conceived from it is truly the right strategy.

Marketers engaged in data-driven marketing might consider taking a fresh look at the fundamental question: "What exactly is the data I'm looking at right now?" Could it be that we're chasing numbers without a clear definition? Carefully examining this origin point could be a crucial step toward enhancing the precision of your company's marketing.