What you need for data-driven marketing and how to do it


What you need for data-driven marketing and how to do it

  I explained the meaning of "data driven" and its significance in current marketing.

As a review, data-driven means a business style that determines strategies such as marketing and individual actions (approach to customers, etc.) based on various customer data.

This time, we will talk about what is necessary to realize data-driven marketing, and what kind of data and how to collect it.

Data-driven marketing implementation procedure

Improvements in computer hardware capabilities and software analysis technology have made it possible to use data efficiently and effectively. I introduced last time that it has become possible to realize marketing that works based on data, without relying on the intuition and experience of human power as in the past .

So what kind of elements should be prepared and what steps should be taken to actually execute data-driven marketing? Broadly speaking, the flow is "data collection"-> "data visualization (processing)"-> "data analysis"-> "formulation of measures and action plans"-> "implementation of measures and action plans"-> "measurement of effects" . We will proceed and turn PDCA.

Let's take a closer look below.

First of all, data collection is a major premise

To implement a data-driven business style and be successful, you must first and foremost collect data. The data that can be used in marketing varies depending on the industry and business, but in general, the customer's purchase history, the process leading up to the purchase, usage satisfaction, repeat rate, family structure, friendship, interests, etc. Can be mentioned.

Such data includes not only the history directly linked to the purchase, but also behaviors such as browsing and withdrawal on the website, customer preferences and relationships that can be grasped from postings on SNS, and opinions and requests collected using Web forms. Questionnaire etc. are also included.

In addition, it is necessary to manage this information by associating it with an individual. In order to uniquely link the purchases and behaviors of omni-channel customers, it is also important to develop member IDs and CRMs across service contact points.

These saucers can be used if they already exist, but if they do not exist, they must be prepared. In addition, in companies that are gradually preparing for IT, there are many cases where various customer-related information is distributed in various systems. In this case, you should consider introducing a data warehouse (DWH) or DMP (Data Management Platform) for centrally managing data and building a system for centralized management of collected data.

Visualize data

The next step is to process the collected data to make it easier to analyze.

Big data, which is the basis of data collection, is not always the one that can be effectively used immediately in business. Rather, it is a mass of data that contains various formats and contents, and its utility value is not clear unless it is organized, just like the chaos of cobblestone mixing. From there, it takes time and effort to process the data in order to make it usable. This process can be said to be a procedure for visualizing chaotic big data as something that can be used in-house.

Tools are needed for visualization work. Of course, it is not realistic to process a huge amount of data only by human power. Data-driven tools include web analytics tools and BI tools. I will explain these details next time.

Analyze and utilize data

And the data that has been visualized (processed) is finally the target of analysis.

You can analyze based on the theme, and as a result of further analysis processing, you can find issues, hypotheses, measures, and use them for actual marketing measure planning and action plan formulation.

In order to perform such data processing work in-house, it is necessary to have the skills and know-how to analyze and derive useful elements from it. The human resources responsible for this work are "data scientists" and "data analysts" who are experts in data analysis and utilization.

Data scientists and data analysts need knowledge of database operation, data processing / analysis skills, statistics, etc., as well as business and marketing knowledge to draw appropriate conclusions and proposals according to the content of the required work. is. If your company does not have such specialized skills, you may need to hire or develop them yourself.

However, at present, there is a shortage of human resources with such skills, and it takes time and cost to train them in-house. Needless to say, it is not a skill that you can acquire overnight. Therefore, data-driven marketing requires a certain amount of time to prepare human resources.

Now that you've reached this point, you'll consider the elements derived from data-driven and devise a concrete action plan by utilizing the skills of specialized human resources such as data scientists and data analysts. We will incorporate the plan into actual marketing activities such as advertising and promotion, measure the effect of executing the plan, and further improve by turning PDCA.


In order to execute data-driven marketing, it is necessary to have a mechanism to support policy planning by collecting, visualizing, and analyzing data and deriving elements that are useful for actual marketing activities.

As mentioned in the text, in addition to human resources specializing in data processing, appropriate tools such as DMP and Web analysis tools are also essential for this mechanism. Next time, I will focus on what kind of tools actually play what role.

Post a Comment