What is data-driven marketing? A clear explanation of its meaning, implementation, KPIs, and success stories.

Data-driven marketing is a method of making marketing decisions based on objective data analysis results, rather than intuition or experience. As customer purchasing behavior becomes more complex and traditional effectiveness measurement becomes difficult due to privacy regulations, the importance of data-driven decision-making is increasing year by year.

This article explains the definition of data-driven marketing, the reasons why it's gaining attention, the benefits it offers, KPIs for measuring results, five steps to implementation, common pitfalls, and real-world company case studies.

What is data-driven marketing?

Data-driven marketing is a method of marketing that involves collecting and analyzing objective data such as customer data and campaign data, and then using the results as a basis for planning, executing, and improving marketing strategies. Its characteristic feature is that it places the factual data, rather than the intuition of the person in charge or past customs, as the starting point for decision-making.

Meaning of "data-driven"

Data-driven, literally meaning "driven by data," refers to a business approach where decisions are based on data. This approach is applied not only to marketing but also to all areas of business, including management, sales, and human resources. In the field of marketing, data such as customer attributes and behavior, advertising volume and costs, and product and service usage are analyzed, and strategies, from formulation to implementation and evaluation of measures are consistently based on data.

Differences from traditional marketing

The biggest difference between traditional marketing and data-driven marketing lies in the basis for decision-making. Traditional marketing relied primarily on tacit knowledge, such as the experience and intuition of the person in charge, while data-driven marketing relies on verifiable data.

However, this does not mean that experience and intuition become unnecessary. Field knowledge is an important asset when formulating hypotheses, and data is used to objectively verify those hypotheses and increase their accuracy.

The difference between "data analysis" and "data-driven"

Data analysis is merely the starting point of data-driven marketing. Data-driven marketing encompasses a series of processes, from making decisions based on analysis results, implementing strategies, verifying those results, and using that information to inform future actions. It's important to note that simply setting up a dashboard and "looking" at the data does not constitute practicing data-driven marketing.

The background behind the growing attention to data-driven marketing

The rise of data-driven marketing stems from three environmental changes: the increasing complexity of purchasing behavior, stricter privacy regulations, and advancements in analytical technology.

The increasing complexity of customer purchasing behavior and channels.

With the widespread use of smartphones and social media, the path customers take from becoming aware of a product to making a purchase has become more diverse and complex. While previously, methods that relied on a uniform purchasing behavior model, such as AIDMA, and delivered the same message to a large, undifferentiated audience worked, now each customer interacts through different channels and their decision-making processes differ. To accurately capture this complex customer behavior, objective data analysis is essential.

Strengthening privacy regulations and reviewing effectiveness measurement

Due to browser restrictions on third-party cookies and stricter laws and regulations regarding personal data protection, targeting and measuring effectiveness by tracking individual behavior is becoming increasingly difficult each year. As it is no longer possible to rely solely on measurement methods that track clicks and conversions on an individual basis, there is growing interest in utilizing existing first-party data and measuring effectiveness through statistical approaches (such as MMM, which will be discussed later).

The evolution of AI and analytical technologies, and the increasing volume of data.

The widespread adoption of cloud computing, BI tools, and AI has significantly lowered the barriers to data collection and visualization. However, as the amount of data that can be handled increases, the importance of interpretation and design—specifically, "what to read from which data and how to use it to inform decision-making"—has also increased. Simply introducing tools is no longer enough to guarantee results; we have entered a stage where the ability to transform analysis into decision-making is paramount.

The benefits of data-driven marketing

The four main benefits of data-driven marketing are: a deeper understanding of customers, improved reproducibility of strategies, improved ROI, and the establishment of a continuous improvement cycle.

Deepening customer understanding

By analyzing data on customer attributes, behavior, and attitudes, you can objectively understand customer needs and purchasing decisions. For example, if you can identify through data which customer segments are responding to what values ​​and choosing your company, you can design optimized communication for each target group, leading to improved customer loyalty.

Improving the reproducibility of policies

By identifying the factors influencing the success or failure of initiatives based on data, success can be prevented from being a matter of chance and instead accumulated as reproducible know-how within the organization. Because it becomes possible to quantitatively explain "why it worked" and "why it failed," it becomes easier to maintain the quality of decision-making even when personnel changes or organizational restructuring occur.

Improving ROI (Return on Investment)

By visualizing how much each initiative contributes to results, you can reduce investment in low-performing initiatives and reallocate budget to high-performing ones. In maximizing results within a limited marketing budget, optimizing budget allocation based on data is one of the most direct ways to achieve returns.

Establishing a continuous improvement cycle (PDCA)

By verifying the results of implemented measures with data and incorporating improvements into subsequent measures, a cycle can be established. Quantitative verification, rather than subjective reflection, clarifies the direction of improvement, continuously enhancing the accuracy of overall marketing activities.

Key performance indicators (KPIs)

In data-driven marketing, KPIs (Key Performance Indicators) are set according to the objectives, and the results of the measures are measured quantitatively. Typical indicators can be broadly divided into those that measure the efficiency of the measures and those that measure the relationship with customers.

Metrics for measuring the effectiveness of advertising and campaigns

Typical indicators for measuring the efficiency of advertising and marketing campaigns include the following:

Metrics for measuring customer relationships

The following are some metrics used to measure customer acquisition efficiency and long-term profitability:

When assessing the soundness of a business, it is important to consider not only the CPA (Cost Per Acquisition) for a single year, but also the balance between LTV (Loan-to-Value) and CAC (Customer Acquisition Cost).

A method for measuring effectiveness that integrates online and offline activities.

Measuring solely based on clicks and conversions does not provide a complete picture of marketing activities. This is because offline measures such as TV commercials, transit advertising, and in-store promotions cannot be measured at the individual click level, and even with online measures, the medium- to long-term effects such as awareness and brand image formation are not easily reflected in conversion data.

This is where a method for statistically estimating the sales contribution of each initiative, whether online or offline, becomes necessary. A representative method for this is MMM (Marketing Mix Modeling), which will be explained in the following chapters.

How to implement data-driven marketing [5 steps]

Data-driven marketing proceeds in five steps: setting objectives, collecting data, analyzing data, formulating and executing strategies, and evaluating their effectiveness.

STEP 1. Setting Objectives and KGI/KPIs

The first thing you should do is define "why" you will use the data. Clearly define your ultimate goals (KGIs), such as sales, profit, and market share, and then work backward to set the KPIs that should be measured. If you start collecting and analyzing data without a clear purpose, the analysis itself becomes the goal, and the effort tends to fail to lead to decision-making.

TEP2. Data Collection, Integration, and Storage

We collect the necessary data according to the objective, and integrate and store it in a format that allows for analysis. This includes a wide range of data, such as website and app behavioral data, advertising volume and costs, sales performance, and customer survey results. For many companies, the first hurdle is that data is scattered across different departmental systems. Furthermore, for data containing personal information, establishing rules for acquisition, storage, and use that comply with relevant laws and regulations is essential.

For more information on the approach to collecting and accumulating marketing data, see the XICA Analysis Insight report "Best practices for data collection: What we learned from working with clients"Please also see.

STEP 3. Data Visualization and Analysis

We visualize the collected data and interpret patterns and trends. Beyond simple aggregation and graphing, understanding the causal structure—specifically, "how much each measure contributes to the results"—often requires expertise in statistics and data science. If there are no in-house specialists, utilizing external experts is an option.

STEP 4. Strategy formulation and implementation of measures

Based on the challenges and opportunities identified through analysis, we formulate a marketing strategy and translate it into actionable measures. Specifically, this involves selecting target audiences, determining the message to convey, and allocating budgets to each channel. The key to this step is to transform the analysis results, which should not end as a "report," into a budget and actionable steps.

STEP 5. Verification of effectiveness and improvement

We verify the effectiveness of implemented measures using data and reflect improvements in the next cycle. By incorporating the verification frequency into the operational cycle in advance, the speed of improvement changes significantly. Regarding the selection of effectiveness measurement methods,What is the optimal analysis method for visualizing marketing effects? Explanation of important points regarding analysis methods that companies should chooseThis is explained in detail in "[...]".

Common challenges and failure patterns

Many of the reasons why data-driven marketing fails to take hold are not due to a lack of tools, but rather to data fragmentation, analysis becoming an end in itself, a shortage of personnel, and a lack of understanding within organizations.

Data is scattered and siloed within the company.

In many cases, data is fragmented across departments: sales data is held by the sales department, advertising data by the marketing department, and customer data by the customer service department. Without the ability to integrate data across departments, it's impossible to accurately understand the relationship between strategies and results. A practical approach is to first inventory the locations of the data necessary for your objectives and then integrate the highest priority data first.

Analysis becomes an end in itself, failing to lead to effective policies.

This is a common mistake where the goal becomes simply setting up the dashboard or creating the report. There is a significant gap between making data "visible" and making "decisions" based on that data. It is crucial to define the connection to decision-making—specifically, "what changes will be made based on these results"—from the analysis design stage.

Shortage of analytical personnel and skills

Talent in statistics and data science is scarce in the job market, and training such individuals in-house takes time. Rather than trying to do everything in-house and becoming stagnant, it is more effective to build a system in stages, such as aiming to internalize core areas while utilizing external partners for advanced analysis.

We cannot gain the understanding of management and other departments.

The shift to a data-driven approach involves a transformation of the decision-making process, and therefore will not take hold without the acceptance of both frontline staff and management. An effective strategy is to create small but clear success stories early on and spread the experience of data functioning as a "common language" throughout the organization. As seen in the MeganeTop case study discussed later, even organizations that have emphasized "experience and intuition" can change when objective data provides the foundation for discussions.

MMM (Marketing Mix Modeling) is an analytical method that supports data-driven marketing.

MMM (Marketing Mix Modeling)This is an analytical method that uses statistical models to break down and estimate how much each marketing measure, such as TV commercials, digital advertising, and sales promotions, contributed to sales. Because it does not rely on tracking individual behavior, it has recently attracted renewed attention as an effective measurement method in an environment where privacy regulations are becoming stricter.

What can be achieved with MMM

With MMM, the following are the main things you can do:

• Integrated online and offline effectiveness measurementThis allows you to compare the sales contribution of all marketing strategies, including TV commercials and OOH (out-of-home advertising), which cannot be measured by clicks, using the same metrics.
• Optimizing budget allocation: Based on the effectiveness and efficiency of each measure, it becomes possible to simulate "how much to invest in each measure to maximize results."
• Understanding medium- to long-term effects (brand effect): This approach allows you to capture not only the short-term effects immediately after advertising is launched, but also the long-term effects that emerge through the accumulation of brand engagement.

XICA's approach: The two pillars of investment: "quantity" and "quality" of strategy

XICA Co., Ltd. has developed the MMM analysis platform "MAGELLANIn addition to optimizing the "quantity" of marketing investment through "", the CMM (Consumer Mix Modeling) analysis platform "COMPASSThis service helps unravel the mechanisms by which consumers choose brands and supports the optimization of the "quality" of strategies. A key feature is that data science experts and consultants work alongside clients from analysis to implementation into decision-making.

Success stories of data-driven marketing

Here, we introduce three examples of data-driven marketing practices implemented by companies supported by XICA. In each case, these initiatives improved the accuracy and speed of decision-making by combining "intuition and experience" with objective data.

Case Study 1 [Communications] KDDI: Reduced communication costs by approximately 2% through monthly MMM analysis

KDDI utilizes MMM (Marketing Management Modeling) to optimize marketing investments in the telecommunications market, where the customer acquisition process is complex. Its key features include incorporating weekly consumer awareness surveys into the model to clarify the "measure → awareness → results" structure, and a rapid PDCA cycle of monthly analysis and optimization. At UQ mobile, optimizing monthly marketing investments resulted in approximately a 2% reduction in communication costs per subscriber compared to before optimization (FY2023 results).

→ 詳細：Support case study of KDDI Corporation

Case Study 2 [Food Manufacturer] Prima Ham: Optimizing both the "quantity" of investment and the "quality" of brand strategy.

Prima Ham faced a challenge with its flagship brand, "Kaoru® Arabiki Pork," where it couldn't directly compare the ROAS of each initiative, leading to a reliance on past practices and external proposals for budget formulation. By quantifying the effectiveness and efficiency of each initiative using MMM, it was discovered that TV commercials have a "long-term cumulative effect" that influences results in the medium to long term, even after airing. As a result, the budget for the following year was shifted to a system where "maximizing ROAS" was the criterion. Furthermore, by identifying the drivers of customer loyalty using CMM, the company is proceeding with the reconstruction of its brand strategy that does not rely on price competition. Another significant achievement is that data has become a "common language" across departments, enabling logical explanations to management and briefings to agencies.

→ 詳細：Support case study of Prima Ham Co., Ltd.

Case Study 3 [Retail] Megane Top: Data-driven identification of key factors in brand switching and customer loyalty for "Megane Ichiba"

MeganeTop, which operates "Megane Ichiba," the top-selling eyewear chain in Japan, focused on acquiring younger customers and used CMM (Customer Relationship Management) to analyze the key drivers of "brand switching (new customer acquisition)" and "customer loyalty (customer retention)." The results showed that the key to acquiring younger customers was not price advantage, but "experiential value" in the purchasing process, which was supported by data. Identifying key drivers has improved the speed of decision-making, and has led to objective, data-driven discussions in an organization that previously relied heavily on "experience and intuition."

→ 詳細：Support case study of Megane Top Co., Ltd.

Tools that can be used for data-driven marketing

The tools that support data-driven marketing are selected and organized according to their purpose, ranging from data collection and storage to analysis and visualization, and automation of strategies.

The important thing is not to make tool implementation an end in itself. Based on the objectives and KPIs defined in STEP 1, select the tools by working backward from "what data and functions are necessary for that decision-making."

Frequently Asked Questions (FAQ) about Data-Driven Marketing

What is Data-Driven Marketing?

Data-driven marketing is a method of marketing that uses data-based analysis to adopt a more scientific approach to strategy.

What skills do you need to implement data-driven marketing?

Data-driven marketing requires data analysis skills and knowledge of statistics. It is also important to have a business perspective and knowledge of marketing.

What steps do I need to take to implement data-driven marketing?

To implement data-driven marketing, you must first collect data and analyze it. Next, it is important to develop and implement a marketing strategy based on the results of the analysis. It is also necessary to analyze the data periodically and review the strategy.

What tools are used for data-driven marketing?

There are many tools and services that can be used to realize data-driven marketing. Below are some examples of data collection, storage, organization, conversion, analysis, and visualization.

• Data warehouse (DWH) for storing data
• A data management platform (DMP) that converts data so that it can be used with other tools
• Web Analysis Tools
• CRM tool for managing customers
• BI tools for visualizing data
• Tools to analyze marketing effectiveness and optimize budget allocation
• MA tools that can automate some of your marketing tasks

What should you pay attention to when implementing data-driven marketing?

When implementing data-driven marketing, there are a few things to keep in mind:

• It is important to check the accuracy of your data. If you base your strategy on inaccurate data, it may lead to results that are different from what you intended.
• It is important to analyze data from a business perspective. Simply collecting data is not enough; it is important to link that data to business strategies.
• It is important to hire people with the necessary skills to analyze data or outsource to external experts. Data analysis requires expertise, so having the right team on board is key to success.

Summary

Data-driven marketing is a method of making marketing decisions based on data and repeatedly implementing and evaluating strategies. In practice, it is fundamental to steadily go through five steps, starting with setting objectives and KPIs, followed by data collection and integration, analysis, transformation into strategy, and effectiveness evaluation.

Many of the stumbling blocks are not the tools or the analysis itself,"Designing a system that connects analysis to decision-making" and "Building organizational acceptance"As the examples of KDDI, Prima Ham, and Megane Top introduced in this article demonstrate, when data begins to function as a common language across departments, data-driven marketing becomes a mechanism that enhances the organization's execution capabilities themselves.

The valuable assets of experience and intuition are backed by objective data. As a first step, try identifying one of your company's decision-making processes that continues despite having an unclear rationale.

▼Related articles for those who want to delve deeper into data-driven marketing

Introducing the mechanism and benefits of data-driven marketing
The first step towards a data-driven marketing organization
How to make data-driven decisions
Bringing marketing analytics in-house: Advantages and disadvantages explained.

Author information

XICA Co., Ltd. Marketing Team