Artificial Intelligence

Five steps to building a successful BI strategy. Part 1

Pinterest LinkedIn Tumblr

BI isn’t enough, a strategy is needed

Most experts have long argued that learning how to derive value from data is essential to running a successful business. The market has gone so far that no one will argue that data is at the heart of any company’s business. This means that any organization in any industry must have a BI tool and a separate strategy for effective use of this tool.

At Sisense, I lead a team of BI consultants. Based on my experience, I can say that projects using Business Intelligence with a strategy have been much more successful than those without a strategy. The strategy played an important role in all phases of using the BI platform: during installation, and during implementation, and during use.

How important is the BI strategy in the modern world?

BI strategy is what allows a business to measure its performance, find its competitive advantage, benefit from data mining and statistical methods.

Any organization has been generating data for a long time. The only question is how she uses them, how effectively she builds her BI strategy. Below are five steps to building a strategy that will maximize your business value.

The first step to successful BI: understand where you are and where you are going

A successful company development strategy involves moving from descriptive analytics (what happened) to diagnostic analytics (why it happened), then to predictive (what will happen) and prescriptive (what needs to be done to make it happen). All these stages of evolution are distinguished by Gartner. If we talk about the distant future, it will be dominated by cognitive analytics (analytics that repeat the patterns of the human brain).

To know where you are going, you need to understand what stage of analytics you are in right now. Most companies focus on descriptive analytics and diagnostics. Most likely, you already have reports automatically generated in one form or another. Reports help to understand what’s new in the data, to visualize it to a certain extent. They are likely already providing significant business value, but the next steps in analytics can be even more beneficial. These steps are relatively easy to navigate if you build the right BI strategy.

The second step to successful BI: define your goals and create a data map to achieve them

At this stage, you need to decide what business goal the data should help you achieve as you implement your strategy. The more carefully you plan your business goals, the more successful you will be. It’s important that your data answers a specific business question or can solve a specific problem. Do you want to be more efficient? Find a bottleneck in the R&D process? Do you want to share your knowledge with clients? Looking to add value to your solution? Data can help with all of these tasks.

Predictive and prescriptive analytics will allow you to answer a very wide range of questions. Will you be able to meet your goals by the end of the year? What market segment is best for your organization to target? What set of products should you offer on the market? Who in the company contributes most to the sale of a product or service? Is there a set of parameters that largely determines the success of your business, which is important to take into account. 

In my experience with a wide range of clients, it is best to start with financial and sales data. Finding insights in this area that are useful to your business can have a significant impact on the development of your organization. Take profit growth, for example. In order to analyze it, you first need to collect all the relevant information. This can be data from Salesforce, an ERP system (for example, Oracle), from any marketing tools that store customer data.

Large companies have convoluted sets of applications with many customer touchpoints. These applications collect very large amounts of information. All this information will need to be evaluated in order to understand which technologies will be relevant for solving specific problems. This same granular approach will allow you to determine which datasets are needed for your tasks. It is also important to keep in mind what types of data are missing. Lack of information can prevent you from getting a more complete picture of the situation. Moreover, in the future, it may prevent your company from moving from descriptive and diagnostic analytics to predictive ones. At the end of this phase, you should define the KPIs that you plan to strive for as you implement your BI strategy. In addition, you need to get a clear map of the data sources that you plan to use in this case.

Feature Image credit