Table of contents
- What is a BI tool?
- BI tool concept
- Reasons why BI tools are attracting attention
- Six main functions of BI tools
- 1. Data collection function
- 2. Data processing function
- 3. Data visualization function
- 4. Data search function
- 5.Job management function
- 6. Sharing function
- BI tools and AI
- 5 BI tools to watch in 2021
- “Power BI” for companies using Microsoft products
- License fee free first choice “Google Data Portal”
- Royal road BI tool “Tableau”
- Recommended data search BI for marketers “Qlik Sense”
- Engineer-friendly next-generation BI tool “Looker”
- How to introduce BI tools
- Step 1 Selection of candidate products
- Step 2 try
- Step 3 Initial introduction
- Supplement.Preferences/Data Connection
- Extra How to effectively utilize BI tools
- 1. Align your requirements
- 2. Make it a habit
- 3. Monitor regularly
- Usage examples of BI tools
- 1. Sales analysis of EC sites, etc.
- 2. Service usage monitoring
- 3. Talent management
- 4. Utilization of IoT data
What is a BI tool?
BI tool concept
BI Tool (Business Intelligence) tools are tools for aggregating various types of data accumulated inside and outside the company, visualizing them in tables and graphs, and utilizing them for business decision-making .
The effective use of data through BI tools can be useful in management and on-site decision-making, further expanding the possibilities of business.
Reasons why BI tools are attracting attention
BI tools play the role of a compass when promoting digital transformation (DX).
In order to advance from STEP2 to STEP3 when the maturity steps of DX are divided into three stages, it is necessary to quantitatively judge the state of the company based on the data.
By visualizing the usage status of your company’s products and services with BI tools, you can “update products and services based on the needs of customers and society.” will be able to make decisions .
Due to the simple system of accumulating and visualizing data, many companies are using BI tools.
Six main functions of BI tools
There are six main functions of BI tools.
- Data collection function
- Data processing aggregation function
- Data visualization function
- Data exploration function
- Job management function
- Sharing function
1. Data collection function
This is a function to acquire the data held by the company with a BI tool.
Data types include relational databases such as Amazon’s Redshift, file data such as Excel and CSV, and SNS data such as Twitter and Facebook. All tools can be used to acquire major data, but we recommend that you check whether your company’s data can be handled.
2. Data processing function
This function processes the collected data so that it can be easily visualized.
This processing function is important in BI tools. The first point is whether or not to process without code. Recent BI tools can process data with GUI operations (no code) such as mouse clicks so that anyone can easily process data. There are two things to keep in mind when processing data in no-code.
One is that it is difficult to reuse in teams. It is often easier to see the code in one place when checking whether it is correct when reviewing multiple processes.
The second is where to process. BI tools have a high degree of freedom and can be processed in many places before the data is visualized. Conversely, this becomes a bottleneck when checking specifications. It is recommended that you decide where and what kind of processing is to be done. The following is an example, so please refer to it.
|Acquisition location||when to use||The reason is|
|Before acquiring BI tools||It is better to process the customer’s total number of purchases, product classification, and other general-purpose classification information in the company on the database side before acquiring the BI tool.||When writing calculations on the BI tool side, calculation formulas differ depending on the person, and it becomes necessary to maintain them later.|
|When collecting data for BI tools||If you want to quickly check the data for personal use, it is faster to process the data when the BI tool collects it.||When using the BI tool with multiple people, it is better not to process it here because it can reduce the time and effort of confirming the data processing process. A way of describing data collection and processing separately.|
|When writing formulas for BI tools||It is good to process the data necessary for the output of the BI tool.||Because it becomes difficult to organize the database if data processing is performed before acquiring a BI tool.|
|When creating graphs for BI tools||It is used when narrowing down the aggregation period and attributes of data.||When checking the data processing, if you look at the graph creation part, it will be in good condition, so maintainability is good.|
If the department that uses the BI tool and the department that creates the database are different, it will be necessary to process the data on the BI tool side. If possible, it would be a good idea to clarify the system and roles as described above.
3. Data visualization function
This is the main function of BI tools.
The data visualization function is generally more detailed in paid BI tools. It is important to determine how much expressiveness you want. Operability is also important. Let’s check if you can operate without stress because you will be doing the same operation many times.
4. Data search function
It is a function that manipulates data based on the visualized graph.
After creating a group of graphs, the BI tool does not just look at them, but rather allows the viewer to check the data in the way they want to see it and in the order they want to see it, and make business decisions.
For example, in self-service BI tools such as Tableau and Qlik Sense, there is a function to filter all related graphs by clicking on a graph. For companies with multiple services, clicking on a customer can easily filter all service usage for that customer.
5.Job management function
The job management function in BI tools is a function that schedules data updates.
The frequency of updates varies for each company’s data. In addition, there are various requests from users, such as weekly updates and near-real-time data. The job management function provides data update settings that meet such requirements. There are also BI tools that have error checking and a chain function that runs the next job according to the execution result of a certain job.
6. Sharing function
This is a function to distribute the created graph group by e-mail.
It is possible to expand the distribution of information, and by placing it in places where referees usually make decisions, such as Slack, URL links, and other applications, decision-making by BI tools will be made naturally. The ability to share is very important.
BI tools and AI
A BI tool called Exploratory can easily handle machine learning methods such as decision trees and random forests.
However, many tools are limited to providing simple statistical methods. Instead of implementing machine learning functions, they often provide a ” connecting ” function that displays the results of R, Python, etc. on BI tools.
5 BI tools to watch in 2021
There are many BI tools, but we will introduce 5 recommended BI tools that have a lot of introduction results.
“Power BI” for companies using Microsoft products
- Companies introducing BI tools for the first time
- Companies that have introduced Microsoft 365
Among paid BI tools, the license usage fee per user is cheap (several thousand yen per month), and depending on the Microsoft 365 plan, the usage fee for Power BI may be included, so the company will charge an additional usage fee. It can also be handled without
Since it is a Microsoft product, you can intuitively understand it with an interface like Excel. In addition, cooperation with Microsoft 365 apps is smooth. Many companies classify and store data in Excel, so this is a great strength of the product.
Data can be stored in cloud storage “Azure Data Lake Storage” and analysis can be performed quickly on the cloud. If your company can handle it without additional usage fees, we recommend trying Power BI first.
The advantage of Microsoft, which has a long track record in Excel, is that there are many powerful functions. Compared to other BI tools, it has more calculation functions around dates, and it is also attractive that more than 250 functions called Data Analysis Expression (DAX) functions are available.
License fee free first choice “Google Data Portal”
- Companies introducing BI tools for the first time
- Companies that have introduced Google services such as BigQuery
If you have a Google account used with Gmail etc., you can use “Google Data Portal” for free. It is also attractive that you can start using it on the browser without any special setup.
It is widely used in IT start-up companies. Although it can be used without a license fee, it has all the basic functions of collecting, aggregating, visualizing, and sharing data.
As a feature, you can add shapes and text like Google Slides. There are many BI tools that cannot add figures, so the point is to create a design that is easy to convey.
I think that you can grasp what kind of visualization expression can be done by looking at the template .
We have published a book on how to use GoogleDataPotal and how to use BI tools, so if you are interested, please read it.
Royal road BI tool “Tableau”
- Companies that want to use BI tools for a long time
- A company that wants to introduce a tool with a wealth of functions and a proven track record of implementation
Except for personal use and free use, Tableau is the most used BI tool in Japan. (Questionnaire results of BI Tool Research Institute)
The most important feature is the variety of visualization expressiveness. Other BI tools can also represent basic graphs such as bar graphs and line graphs, but the same bar graph can express multiple axes, trend lines, grouping, etc. to understand data.
The speed of searching data can also be changed at a very high speed by switching viewpoints using a technology called “VizQL”.
I think Tableau is the only choice for companies that are particular about data visualization. I created a Udemy course for Tableau beginners ! I would be happy if you could watch it for those who are going to introduce it!
Recommended data search BI for marketers “Qlik Sense”
- Retail/EC companies with rich customer data
- Companies with marketers who are good at data exploration
Qlik Sense is an unrivaled tool for multifaceted analysis of user information. There is a patented technology called “associative technology”, and by connecting data in various databases with IDs, all data is linked and aggregated.
Data can be aggregated under any conditions simply by clicking on the screen of the tool. It is recommended for marketers who want to dig deeper into user conditions under various conditions.
If you’re interested in what Qlik Sense has to offer, check out our demo video !
Engineer-friendly next-generation BI tool “Looker”
- Companies that want to realize high-speed decision-making
- Companies that want to centrally manage and visualize company-wide data
Looker is a latecomer among BI tools. Among the products introduced, the initial license fee is high. Although it is relatively difficult to introduce, it is a necessary tool for achieving high-speed decision-making and efficient and highly reusable data aggregation operations, aiming for a data-driven company .
Looker uses its own language, LookML, to describe aggregations and data relationships within SQL. This LookML improves the reusability of SQL because version control, inheritance, object definition, etc. can be performed using Git repositories like software development.
Of course, since it is a Google product, it is also compatible with Google Analytics and Google Ads.
It is an essential tool for realizing fast and efficient decision making with BI tools like software development.
Recently, the concept of Headless BI, which has evolved further and centrally manages data aggregation definitions and passes them to applications, has also appeared.
How to introduce BI tools
Since I introduced the recommended tools, I will also introduce how to introduce the tools. In this section, we will divide the steps required to utilize BI tools into three steps. * I will omit the details such as organizing the requirements.
Step 1 Selection of candidate products
First, compare the tools and select the product.
If you find a product that interests you after reading this article, please contact the manufacturer. It is also recommended to select “Google DataPortal” as a trial in the sense that there is no account usage fee. Before introducing the product, you can try it out to see if it can be used as expected. Let’s make sure that it can be used by trial use instead of introducing it suddenly.
Step 2 try
After selecting a product, we will try it out to see if it can be used as expected.
Common check items are as follows.
|item||things to do||point|
|visualization expressiveness||Check the expressive power of graphs and tables. Since recent BI tools cover the main expression methods, it is good to see the layout settings of frequently used graphs.||There are many companies in Japan that are particular about the layout of the table. There are many things that cannot be done, such as cell merging and total display, so I will check how far I can do it and explain it to the manager in advance.|
|Account structure||There are many things that are divided into developers and referrers. Make sure that what the referrer can do is what you expect.||If it is discovered after introduction that you cannot do what you want to do with reference, the cost plan will exceed the plan.|
|Ease of development||Let’s check how efficiently we can develop data visualization work because there are many repetitive settings.||There are many cases where it is urgent to create a graph. There is a risk of making a mistake if the development efficiency is poor when working in a hurry, or if there are many extra steps.|
|governance||Let’s see how finely you can define whether to show data or not.||After the introduction, if another department wants to use the BI tool, it would be nice to be able to set the settings to show only the department at the file level or data level.|
Step 3 Initial introduction
After confirming that the BI tool can be used as expected in the trial, it is time to introduce it.
First, make an implementation plan. I would like you to be aware of the following points. How many weeks does each item require? Also, consider whether there are any concerns.
|Preferences/Data Connection||Get your data ready for BI tools.||If you have some knowledge of SaaS type products, you will be able to connect to data and visualize data in less than an hour. * For details, refer to “Supplement. Preferences/Data Connection” below.|
|graphing||Create the graph group you want to create.||It is difficult to create a perfect output in one go. The appearance of graphs and conditional expressions may be blurred. Therefore, it is a good idea to present the conditional expression in a simple way such as a bar graph and review it with the referrer.|
|graph review||Check the appearance of the graph and conditional expressions.||The point is to create a “loved graph” that can be used many times. Be conscious of how you use it and write supplementary comments and be careful.|
Configure settings to use BI tools in your company’s environment.
In the case of a SaaS type product that can be used immediately with a web browser such as Google Chrome, it is very easy and can be used just by setting it so that the company’s DB can be recognized from the BI tool.
On the other hand, for on-premises products that build BI tools on your own server, you need to install it yourself or ask a company specializing in installation for a fee. Anyone can easily use it if it is only text data such as Excel and csv.
Extra How to effectively utilize BI tools
BI tools are used by various stakeholders within the company, so it is necessary to create output based on everyone’s opinions. Here are three points.
1. Align your requirements
Communication between graph developers and users is important for BI tools. There are often cases where the developer cannot understand the user’s purpose or request well, or the user cannot convey the request well. I’ve created a document about organizing requirements for developers, so I’d be happy if you could refer to it.
2. Make it a habit
The output of graphs created with BI tools is not the end of creation. The goal is to be used in decision making. Therefore, let’s ask the user what the purpose of use is. And try to incorporate it into situations where decision-making is required.
For example, if there is a regular decision-making meeting, it would be a good idea to add the URL of the BI tool to the regular meeting minutes. Also, if you decide to make decisions based on fixed conditions and timing, add a scheduling function that determines the conditions and delivery timing to communication tools such as email and Slack so that you can make decisions habitually.
3. Monitor regularly
Monitor the operating status of your BI tools. BI tools can visualize the access status, etc. If there are graphs that have not been used for more than half a year, it would be a good idea to abolish them or directly ask the users why they are no longer used. If the initial requirements are no longer needed or if the output is unsatisfactory, we will continue to make revisions and use it.
Usage examples of BI tools
1. Sales analysis of EC sites, etc.
The best way to use it is to visualize your company’s earnings such as sales.
In the case of e-commerce sites and POS data for retailers, which often contain product and user data, RFM analysis such as purchase amount, purchase frequency, last purchase date, etc., and user classification and visualization are performed. We often conduct area analysis for each prefecture at our sales offices and customers’ purchase stores.
2. Service usage monitoring
The next most common use is to monitor the usage of your company’s services and products.
Visualize data to derive product improvement proposals from user behavior and optimize customer behavioral experiences within services. Quantitative data alone makes it difficult to identify why a particular state exists, so it is also important to conduct qualitative interviews directly with users in conjunction with quantitative data analysis.
3. Talent management
BI tools are not limited to specific purposes, and can be used for various purposes as long as you have data.
There is also an increasing need for talent management, which digitizes the skills, abilities, and work performance of the company’s employees and conducts appropriate personnel assignments and recruitment and training activities. It is necessary to carefully collect data and reassign personnel, but by managing based on certain data-based standards, it is possible to implement fair personnel measures.
4. Utilization of IoT data
With the spread of IoT devices, various things can be handled as data.
For example, there are cases where the usage status of in-house printers is aggregated with BI tools to optimize usage. By integrating printer usage data and employee ID data, the General Affairs Department will be able to ascertain which department is printing how much. You can also create a mechanism that notifies each department if the color ratio is high, or notifies that double-sided printing is required if the print volume is high.
This time, we introduced an overview of BI tools and how to use them.
Visualize data and make data-driven business decisions by referring to the introduction methods and usage examples of the BI tools introduced this time.