What is marketing data analysis? Explanation of methods, procedures, and tools for different purposes.

Marketing is "making sales unnecessary." These are the words of Peter Drucker, the father of modern management.[English-Japanese Translation] Definitive Edition of Drucker's Quotations / Diamond Inc.Than).

Data analysis is essential for this type of marketing. Most marketing activities that rely on intuition and experience end in failure and waste costs.

However, even though it is said that data analysis is essential for marketing, many people probably don't even know what kind of data analysis methods are available.

This article systematically explains everything from the basic concepts of marketing data analysis to 13 methods for different purposes, how to proceed in practice, and how to choose the right tools. Whether you're just starting out with data analysis or a marketing professional looking to review your company's analysis system, please use this as a practical reference.

What is marketing data analysis?

Definition of Marketing Data Analysis

Marketing data analysis is the process of collecting, organizing, and analyzing data related to customers, markets, competitors, and strategies, and using that data to inform marketing decision-making. By answering questions such as "Who should we reach, what should we deliver, through which channels, and how much should we deliver?" based on data rather than intuition or experience, it improves the accuracy and reproducibility of marketing investments.

Why is data analysis essential for marketing today?

The importance of data analysis in marketing has long been recognized. However, the need for it has rapidly increased in recent years due to three environmental changes.

① Diversification of digital channels and explosive growth in data volume

Marketing channels such as TV commercials, web advertising, social media, e-commerce sites, and email newsletters are increasing year by year. As the number of channels increases, the performance data for each measure also becomes more complex and voluminous. We have long since exceeded the limits of what humans can intuitively grasp about "which measures are contributing to sales," and it is becoming difficult to aim for overall optimization without a system for systematically handling data.

② Increasing complexity of consumer behavior

The consumer journey has drastically changed from the simple "see a TV commercial and buy it in-store" pattern of the past. Now, a complex path involving awareness on social media, comparison on review sites, price comparison services, and online purchases is commonplace. To properly understand this customer journey, cross-database analysis is essential.

③ The need for speed in decision-making

In today's rapidly changing market environment, quickly iterating through the cycle of planning, executing, and evaluating marketing strategies is crucial for gaining a competitive advantage. Companies that have a system in place to quickly determine "what is working and what is not" through data analysis will see a significant difference in the results they achieve, even with the same budget, compared to companies that do not.

The difference between "general data analysis" and "marketing data analysis"

The term "data analysis" is used broadly and applies to all business areas, including operational efficiency, financial analysis, quality control, and demand forecasting. Marketing data analysis is one area within this broader scope, but its distinguishing feature is that the objectives and target data are clearly defined.

In other words, "having data analysis skills" and "being able to effectively use marketing data analysis in practice" are two different things. The latter requires not only statistical knowledge but also an understanding of marketing-specific objectives, metrics, and data structures.

This article provides a systematic overview of marketing data analysis. First, it introduces 13 analytical methods tailored to specific objectives. Next, it covers how to apply the analysis to practical business situations, how to choose the right tools, and the process of translating analysis results into actionable strategies. Use this as a guide to select the method best suited to your company's challenges.

[By purpose] 13 data analysis methods to support marketing

Now, let's take a look at 13 representative data analysis methods that directly lead to improved marketing results.

The main scenarios in which data analysis is used are as follows:

• Create a marketing strategy(Market/competitive analysis)
• Analyze customer data(CRM/LTV analysis)
• Uncover trends in your data(aggregation and trend analysis)
• Uncover data relationships(Correlation and causal inference)
• Identify data classifications(segmentation)
• Discover hidden details in your data(Insight discovery)

We will introduce important data analysis for each scenario, so please use this as a reference when selecting the data analysis you think is necessary for your company's marketing.

There are many other approaches to data analysis besides the methods introduced here, but data analysis can be broadly divided into four stages, from "what happened (descriptive)" to "what should be done (prescriptive)." If you would like to get a more bird's-eye view of your company's data utilization phases, please also refer to this explanatory article.

·Related article: What are the four categories of data analysis? The role and use of "descriptive," "diagnostic," "predictive," and "prescriptive" analytics to answer marketers' questions

Create a marketing strategy

1. 4P Analysis

4P analysis is a data analysis method for developing marketing strategies by organizing information along the four axes of "Product," "Price," "Place," and "Promotion."

For example, when releasing a new product or service, we design the functions and performance, appropriate price, sales channels, promotional activities, etc. to suit the target users.

The important thing is to have a balance between the four Ps. A marketing strategy that is biased towards any one of them is likely to fail.

2. STP Analysis

STP analysis is a data analysis method that organizes information along three axes: "Segmentation (segmenting the market)," "Targeting (determining the target market)," and "Positioning (confirming your position in the market)," in order to clarify your position in the market.

Since business is a competition to gain market share, when formulating a marketing strategy it is necessary to understand the position of your company and its products and services.

3. SWOT Analysis

 SWOT analysis is a data analysis method that organizes information along four axes - Strengths, Weaknesses, Opportunities, and Threats - to develop a marketing strategy that is based on reality.

 By organizing information based on SWOT, you will be able to develop an appropriate marketing strategy while taking into account the positive and negative factors of your company and the market.

Analyze customer data

4. Decile Analysis

Decile analysis is a data analysis method that divides customers into 10 groups based on product and service purchasing data, and visualizes purchase ratios, sales composition ratios, etc.

Analyzing customers not only as a whole but also by dividing them into various segments enables you to analyze customer data from a different angle.

This identifies the customer segments in which you should prioritize your marketing investments, allowing you to "select and focus" your investments.

5. RFM Analysis

RFM analysis is a data analysis method that classifies customer groups by scoring and ranking customers based on three pieces of data: "Recency (number of days since the last purchase)," "Frequency (number of purchases and purchase frequency)," and "Monetary (total purchase amount)."

After organizing the data based on RFM, customers are classified as follows:

*Created by our company based on the following:
Source: https://tech.pepabo.com/2017/12/06/tableau-rfm/

This will enable you to identify customers to whom you should actively upsell or cross-sell, and customers to whom you should work to prevent from dropping out.

Uncover trends in your data

6. Cross-tabulation analysis

Cross-tabulation analysis is one of the basic data analysis methods. It uses data obtained through questionnaire surveys to aggregate data by combining multiple axes such as respondent attributes and question items.

Example: An ice cream manufacturer is running a campaign targeting women, and they conduct a cross-tabulation analysis to decide which products to target for the campaign.

① Comparing the sales data for vanilla ice cream and strawberry ice cream, which are candidates for the target product, vanilla ice cream accounts for twice as much sales as strawberry ice cream. If you look at this data alone, you might be tempted to make vanilla ice cream the target product for the campaign.

However, this time the campaign is targeted at women. Looking at the sales data by gender, we can see that men make more purchases than women.

So, let's look at the sales data broken down by gender of the purchaser. We found that 9% of vanilla ice cream purchasers were men, while 8% of strawberry ice cream purchasers were women. Based on these results, we decided to target strawberry ice cream for this campaign aimed at women.

Cross-tabulation analysis involves combining various axes in order to extract the information needed for data analysis.

Source: [Data analysis from scratch #2] Three analysis methods that data analysis beginners should remember | XICA Co., Ltd. https://xica.net/xicaron/data-analysis-for-beginners-2/

Uncover data relationships

7. Multiple Regression Analysis

In statistics, "multiple" means plural and "regression" means causation.

Multiple regression analysis is a data analysis method that understands the relationship between one outcome (target variable) and multiple factors (explanatory variables).

How much should we spend on advertising to reach our sales goals?

"How much effect can we expect from each of Measure A and Measure B?"

"How can we allocate our budget this fiscal year to maximize our marketing results?"

This is an analytical method that is essential in the marketing field, as it can be used for sales forecasting, formulating marketing strategies, and more.

For example, by analyzing the degree of influence of various factors (explanatory variables) that affect sales (objective variable), you can consider the optimal budget allocation for marketing initiatives.

Source: [Data analysis from scratch #2] Three analysis methods that data analysis beginners should remember | XICA Co., Ltd. https://xica.net/xicaron/data-analysis-for-beginners-2/

Free downloads of related materials

A guide to multiple regression analysis in Excel that empowers marketers
~ Understand the correlation between marketing measures and business results ~

Download Now

8. Logistic regression analysis

Logistic regression analysis is an analytical method for predicting and explaining the probability of a binary response variable (outcome) occurring from multiple explanatory variables (factors). Binary means that there are only two response variables, such as "YES" or "NO."

For example, you can use data about a potential customer (sales, number of employees, profit margins, etc.) to predict the likelihood that that customer will purchase your products.

<Difference between multiple regression analysis and logistic regression analysis>

• Multiple regression analysis
  Predict the value of the objective variable from multiple explanatory variables
• Logistic regression analysis
  Predict the probability that the objective variable will be "1" from multiple explanatory variables

Although multiple regression analysis and logistic regression analysis have some similarities, their use cases are clearly different.

9. Association Analysis (Basket Analysis)

Association analysis (basket analysis) is a data analysis technique that uses purchasing data to analyze how often a certain product is purchased together with other products.

By identifying combinations of products that are frequently purchased together, you can use this information in your marketing strategies, such as placing or recommending products that are likely to be purchased together nearby.

*Created by our company based on the following:
Source: What are the benefits and implementation methods of basket analysis that anyone can easily use? | Cross Marketing for research and market research https://www.cross-m.co.jp/column/data_marketing/dtm20220722/

Identify data classifications

10. Cluster analysis

Cluster analysis is a data analysis method that discovers features (variables that provide clues for prediction) for each piece of data from a population of data and identifies groups (clusters) of data according to these features.

Cluster analysis can be applied to not only customer information, but also any other data such as companies, products/services, regions, or survey items.

It also makes it possible to simplify huge amounts of data, and in marketing it is possible to segment consumers and implement appropriate measures for each cluster.

11. Decision Tree Analysis

Decision tree analysis is a data analysis method that classifies data using a tree structure. It is often used in data analysis because it can produce relatively accurate results even with simple analysis and the results are easy to interpret.

This is a data analysis method primarily used to discover potential customers and understand the impact that products and services have on customer satisfaction.

Discover hidden details in your data

12. Factor Analysis

Factor analysis is an analytical method for finding hidden common factors (components) from observed element data.

It is used when you want to understand your customers' essential needs, such as quantifying the user intentions hidden behind survey response results.

13. Correspondence Analysis

Correspondence analysis is a data analysis method in which the results obtained from questionnaire surveys, cross-tabulation analysis, etc. are plotted as a scatter plot.

*Created by our company based on the following:
Source: What is correspondence analysis? | Intage Market Research https://www.intage.co.jp/glossary/400/

You often see this in manufacturers' product descriptions and marketing proposals.

The relative positions of attributes allow you to understand the relationships between them, so you can use this to understand the relationships between competing products, such as "if you have this attribute, you're more likely to have this attribute as well."

The necessity and benefits of data analysis in marketing

Based on the aforementioned environmental changes, we will now outline the specific benefits that can be gained by your company actually engaging in marketing data analysis.

Data-driven decision making

Traditional marketing that relies on intuition and experience is an activity with a strong element of gambling. Marketing activities incur large costs, so it can be said that "if it hits, it's heaven, if it doesn't, it's hell."

Data analysis is essential to eliminate this gambling element.Data-driven decision-making can dramatically increase the success of your marketing efforts.

Enabling personalized marketing

As consumer needs have become more diverse, the need for "personalized marketing" has increased.

"Personalized marketing" refers to the analysis of customer data, prospect data, market data, etc.Implementing different marketing for each segment (target classification).

In order to achieve "personalized marketing," it is essential to deepen understanding of the market and customers by making full use of the analytical methods mentioned above.

Improving marketing through repeated hypothesis and verification

The benefits of data-driven marketing include:You can quickly "implement measures" and "check results"You may.By repeating the hypothesis and verification cycle in detail, you can quickly improve your marketing activities.It looks like

If you can use various data analysis methods depending on your marketing objectives and speed up the hypothesis-verification cycle, you can expect better marketing results.

Steps for conducting marketing data analysis

Marketing data analysis involves the following main steps:

Start with the above steps and follow the flow carefully to proceed with your data analysis. Go back to the step before you feel something is wrong and proceed with the flow again. If you do not follow this flow, your marketing data analysis will most likely fail, so be careful.

For details of each step,[Data analysis starting from zero #1] “8 steps of analysis” that data analysis beginners should know first' is explained in detail in.

This article contains important basic knowledge not only for marketing but also for data analysis in business in general, so please read it together with the main article.

Recommended data analysis tools for marketing

When analyzing marketing data, if the amount of data is small, it is possible to use tools such as Excel. However, if you need to analyze a large amount of data, we recommend using a dedicated data analysis tool.

Here we will introduce four data analysis tools that are recommended for marketing, so please take a look.

MMM (Marketing Mix Modeling) Tool

MMM is a statistical method that comprehensively analyzes marketing-related big data and visualizes the direct and indirect impact that each marketing initiative had on results.

As the number of media and channels for marketing continues to increase, it is impossible to understand accurate marketing results without integrated data analysis. MMM solves this problem and allows you to understand the overall picture of your marketing activities from all data, and the MMM tool is a tool that incorporates this mechanism so that anyone can use it.

The MMM analysis platform provided by XICAMAGELLAN" has been adopted by companies in a wide range of industries and sectors to visualize advertising effectiveness and optimize budget allocation.

MA (Marketing Automation)

MA is a system that enables lead generation and lead nurturing through predefined scenarios and imported lead data.

While analyzing online and offline prospect data, you can automate parts of your marketing initiatives by triggering pre-defined scenarios.

By utilizing big data, we can improve the efficiency of marketing operations, allowing marketers to focus on creative work.

DMP (Digital Management Platform)

A DMP is a data analysis tool for managing and analyzing the vast amounts of data accumulated on the Internet.

DMP providers primarily use audience data (third-party data) collected independently to optimize advertising operations, but they can also combine this data with user data they own to identify likely potential customers.

In addition, by utilizing a DMP, it is possible to deliver advertisements and emails to segmented users, making one-to-one marketing possible.

CRM (Customer Relationship Management)

CRM is a data analysis tool that comprehensively manages user data and allows for customer analysis from various angles. In addition to data analysis, it also allows for cross-departmental use of user data.

Originally, it was a tool for managing user data and building good customer relationships. However, in recent years, data analysis functions have been strengthened in many CRMs, making it an indispensable data analysis tool for customer analysis.

How to use data analysis results in marketing strategies

Simply conducting data analysis will not lead to improved marketing results. What is important is to translate the insights gained from the analysis into concrete marketing strategies and turn them into actual action. Here, we will explain the process and key points for how to utilize the results of data analysis in your strategies.

1. Extract issues and hypotheses from the analysis results

First, we identify current issues, discover new discoveries, and derive hypotheses from the visualized and analyzed data. For example, we organize facts and trends that can be read from the data, such as "the dropout rate of a specific customer group is high" or "certain products are selling well in specific channels."

2. Planning and prioritizing measures

Consider what marketing measures will be effective in response to the issues and hypotheses you have identified. If multiple measures have been proposed, prioritize them and create an implementation plan, taking into account factors such as impact, feasibility, and resources.

3. Implementing measures and measuring their effectiveness

Implement the planned measures and measure their results quantitatively. By setting KPIs (key performance indicators) and evaluation criteria in advance, you can objectively judge the effectiveness of the measures.

4. Continuous improvement through the PDCA cycle

Based on the results of effectiveness measurement, we identify areas for improvement in our measures and use this to guide our next steps. By repeating this PDCA (Plan-Do-Check-Act) cycle, we can improve the accuracy and results of our marketing activities.

5. Developing specialized personnel and systems

To effectively move forward from analysis to policy planning and execution, it is effective to utilize personnel with specialized knowledge, such as data scientists and marketing analysts. If your company lacks specialized personnel, consult with external experts or consider internal training and strengthening your team structure.

In this way, to utilize the results of data analysis in marketing strategies, it is important to go beyond simple analysis and to be aware of the entire process, from identifying issues to planning and implementing strategies, measuring effectiveness, and continuous improvement. By repeatedly making decisions and taking action based on data, you can steadily improve your marketing results.

Deepening Marketing Data Analysis: A Data Science Perspective

As you implement the methods, approaches, and processes for utilizing the measures described so far, you may eventually encounter a wall where "we have the data, but we don't know what it means." The perspective needed to overcome this wall is data science.

In recent years, "data science" has become a rapidly growing field in the marketing industry. Data science is...An approach to analyzing the vast amounts of data held by companies and deriving insights that will benefit the business.Say.

As data analysis has become commonplace in business, the belief that "data can explain everything" has become widespread. This is half true and half false. This is because data itself is a collection of impersonal information, and simply looking at data cannot explain anything. In other words, "data does not give answers."

To gain answers from data, business professionals themselves need to breathe life into the data. This "breathing life into" aspect—using various approaches to derive insights from the data and translate them into subsequent actions—is what data science does. As the need for data utilization in marketing increases, data science has become indispensable.

For data science,"What is Data Science? A Data Professional's Easy-to-Understand Explanation [3 Recommended Books for Beginners]"We've provided a detailed explanation on this, so please take a look.

Frequently Asked Questions: Marketing Data Analysis

Q1. What is the minimum data required to begin marketing data analysis?

The minimum data required varies depending on the purpose of the analysis. If the goal is to understand customers, purchase history and customer attribute data are the starting points; if the goal is to evaluate the effectiveness of a campaign, time-series data on advertising expenditure, costs, and sales are the starting points. It is important to first take stock of the data your company already possesses, define "what question you want to answer," and then identify the necessary data. In practice, it is more realistic to start with the data you have on hand and fill in the gaps as needed, rather than waiting until you have perfect data before starting the analysis.

Q2. Is it possible to perform marketing data analysis even without specialized knowledge?

Many methods such as 4P analysis, STP analysis, SWOT analysis, cross-tabulation analysis, and decile analysis can be performed using Excel or other spreadsheet tools, and can be undertaken without specialized statistical knowledge. On the other hand, methods such as multiple regression analysis, factor analysis, and cluster analysis require a certain level of statistical understanding. Starting with methods that do not require specialized knowledge to build experience and gradually moving on to more advanced methods is a realistic approach to developing an organization's analytical capabilities. When advanced analysis is required, utilizing external experts and tools is also an option.

Q3. Of the 13 methods, which one should I start with?

The methods you should prioritize will vary depending on your company's situation. As a guideline, if you have customer data (purchase history), RFM analysis or decile analysis is a good starting point; if you want to know the cost-effectiveness of your advertising and marketing measures, multiple regression analysis or MMM is a good starting point; and if you want to clarify your company's position in the market, SWOT analysis or STP analysis are good starting points. There is no need to aim to "master everything at once," and choosing one method that is most directly related to your immediate business or marketing challenges and trying it out will lead to results the fastest.

Q4. Why aren't the results of data analysis being used within the company?

The most common problem is that analysis starts with "data first" rather than "problem first." Analysis that begins with the idea of ​​"what can we learn from this data?" often yields results that are difficult to translate into concrete decision-making. To effectively utilize analysis within a company, it is crucial to clearly define "whose decisions, what decisions, and by when" during the analysis design phase. The method of sharing analysis results is also important; instead of simply listing numbers, presenting them along with a bridge to action—such as "therefore, this policy should be changed"—will increase the rate of adoption within the company.

Q5. How long does it take to see results from marketing data analysis?

The time required varies greatly depending on the type and purpose of the analysis. Methods that use existing data, such as cross-tabulation analysis and RFM analysis, can be completed in a few days to a few weeks, and the results can be reflected in measures relatively quickly. On the other hand, methods that require time-series data, such as multiple regression analysis and MMM, can take several months from the start of data collection and preparation. Furthermore, it takes a certain amount of time for the results to be reflected in the numbers after the analysis results are reflected in the measures. It is not realistic to expect sales to increase immediately after starting the analysis, and it is important to understand that accuracy and results accumulate as the PDCA cycle is continuously run.

In conclusion: Taking data analysis to the next level.

Marketing data analysis isn't about knowing the techniques. Techniques like 4P, RFM, and multiple regression analysis are merely tools for answering questions. What's important is formulating the right questions to address your company's challenges, choosing the right techniques, and translating the analysis results into actual strategies.

You don't need to master all 13 techniques introduced in this article at once. The realistic first step towards data-driven marketing is to first identify one marketing challenge for your company and then choose one technique to address that challenge and try it out.

If you're unsure which method to start with, RFM analysis is a good starting point if you have customer data; multiple regression analysis or MMM if you want to understand the contribution of strategies to sales; and SWOT analysis or STP analysis if you want to organize the overall picture of the market and competitors.

For specific details on how to proceed with the analysis, please refer to the related articles below.

• 8 Steps to Data Analysis (Beginner-Friendly)
• Four classifications of data analysis: The role of descriptive to prescriptive analysis

Author information

XICA Co., Ltd. Marketing Team