Business Intelligence versus Big Data: Intelligent Information

In his last notes, Christophe Dumoulin, Director at Business & Decision, laid the basic principles of Big Data by formalizing key issues and providing benchmarks. This raised a new question about the use of data for them to have a significant impact the markets. This third article addresses choosing between the use of Business Intelligence technologies or Big Data for processing information.

The data processing, a constant dilemma for businesses:

Whether Web Analyst, Data Scientist, the simple user or manager, everyone is trying to understand the operation of all the available data and determine the real benefits for the company. The volume of information has gone from scarce to superabundant in the last few years. Among the most important challenges expressed by the “Chief Marketing Officer” Four to note: the explosion of information, increased trade on social networks, the proliferation of information, and consultation terminals changing demographics.

This brings new opportunities but also raises many questions about the use of traditional technologies used to exploit this massive amount of data. This new paradigm can be summarized in one sentence: a wealth of data with no real explanation and without context makes it difficult to transform it into actionable information.

All examples that could be cited in the explosion of data show that the data generation is done at a faster and faster speed. It therefore becomes important to know how to process this information to derive trends in new business. This specific pertains to perspectives such as fighting crime, rearrange cities, improve customer knowledge, innovate faster in the life sciences, promote Collaborative economy, etc.

Fundamental Reminder: Business Intelligence versus Big Data

Before entering the heart of the subject, start with a reminder of the fundamentals of business intelligence. With 25 years of experience in the field, I myself will try to come up with a synthetic definition. BI is a set of tools and techniques to gather, cleanse and enrich structured or semi-structured data for storage in various forms of SQL type database.

The data will be managed in standardized formats to facilitate access to information and processing speeds. The goal of BI is to produce performance indicators to understand the past and analyze the present to extrapolate a long-term vision and define future competitive advantages of the company. BI is used by a large number of internal and external users to support the operational activities of the company using strategic monitoring.

Let us try to understand the Big Data around the traditional definition of 4V. A customer database contains the following information: name, gender, age, occupation, status, etc. All this information is stored in a traditional data warehouse. If the definition of 4V is applied to the application and you migrate to a Big Data infrastructure, the answer would be negative.

The volume of data is no longer a problem in itself, we can now talk wide Data Warehouse. The variety of sources is taken into account with the new technologies and low cost integration of additional sources. The velocity is maintained by the application data bus allowing an increased volume of data per time unit. The veracity of the data is finally an immutable theorem in the analysis of data regardless of the infrastructure.

Two different analysis methodologies

Explore further and more in-depth data by introducing new dimensions of analysis: event detection, the chronology of events in the collection of information, the time between events or situations or contexts that qualify events that occurred.

The demonstration can be done by example:

• Case 1: a consumer looks at an advertisement, then he visited the site two days later and calls an adviser. The next day he makes a purchase.
• Case 2: a consumer buys a product, the same day he visits the website. Then three months later he calls an advisor. The following month he looks at advertising.

These two cases show the need to understand the events and the sequence. Although in these two examples the customer purchased the same product, analysis of the customer experience and his career are radically different.

Now consider a client that addresses a counselor after-sales service.

• 1st case: he visits the site twice in the day and in the evening he called a counselor.
• Case 2: He visits the site twice in the day and found the answer to his question without contact.

The interpretation of the information will be different though in both cases because the customer has got the right answer to his question regardless.

In these two examples we can easily measure the difference in Business Intelligence and Big Data. In the first example, marketing implements precise sequences to capture and lock the customer in a course defined according to business rules. The volatile client is spontaneous, hybrid and indecisive. They are constantly breaking rules, pre-established routes and marketing processes. The volatile client, spontaneous, hybrid and indecisive constantly break the rules, the pre-established routes and marketing processes in and out.

Big Data technologies used to store the same data, but in different contexts, apply different treatments and series of differentiated algorithms (NoSQL and other appropriate technologies, graphs, etc.). One can also start learning operations with data without preconceived ideas as well as observation phases to detect weak signals famous (partial or limited information provided by the environment). All information, degrees of customization, or types of recommendation collected will be replicable to be modeled. The knowledge obtained infer strategy, organizations, people and business processes.

I concluded, and it is a personal reflection, there is no direct linkage between BI and Big Data. Analysis techniques are radically different, practiced with know-how and new technologies. The new paradigm is out with current thought patterns and tends to revolutionize the same approach to data analysis. The issue is well beyond the technological debate on SQL databases, no SQL column, Memory, and any other variant.

The interest of Big Data resides less in treated subjects than in how to understand and solve problems in transverse areas (marketing, logistics, risk management …) or in specialized areas (health, energy, distribution … ). This is the heart of the challenge of Big Data: knowing human activity, understand its context, establish relationships between the activity data to provide, at a given time, an individualized and personalized real-time service.

My next post will focus on the analysis of a technical specialist and methods used by big data users in the information value creation process.