Nowadays Business Intelligence systems become a real must-have for any business, regardless of the field of activity and industry. At the same time, the opinion that these systems do not bring any added value, that the profitability and efficiency of this investment are rather difficult to measure is quite frequent. In some cases, the anxiety of the system’s complexity and the inconvenience of its use turn into a stumbling block.
However, we believe that the process of decision making should always follow logic principles, common sense, and business needs.
In this article, we would like to outline a few key points you have to take into account before starting the BI implementation project. With that in mind, you will succeed in solution development for a particular business at a particular moment taking into account the further growth and development of the company.
According to Gartner analysts, about 70-80% (!) of implemented Business Intelligence projects fail. This is primarily due to the fact that the "smart system" gives an overall picture based on figures. The decision making is still the prerogative of a human.
Another factor that influences the success of the entire project depends on the quality of the input data, as well as on their processing quality at each stage of the process. Usually, the functionality of existing systems with their standard set of reports (ERP, CRM, and others) runs its course and becomes insufficient for qualitative analysis of data as far as data amount grows, number of data sources increases. The number of employees working with large data sets to prepare analytical reporting, to create presentations, etc. significantly influences the stability of the system. Therefore, the accurate structuring of the process is almost such important as the implementation of a Business Intelligence system itself.
Based on our work experience on various implementation projects, we`ve gathered a number of factors directly influencing the success/failure of the project.
1) The implementation initiative should come from the top management of the company because these people have enough knowledge and experience to formulate goals and objectives for the implementation project. Also, the opinion of employees directly working with the data generated by the system is highly important. Usually, the end user is exactly the person who fully understands all the benefits and advantages of the implemented corporate reporting system.
2) Goal Alignment. The lack of a clearly stated implementation goal can be easily set to a failed project. To get necessary data in the result you have to clearly formulate data requirements already at the planning stage. Detailed goals investigation will help determine the best possible tools to achieve them. At this stage, it is also important to take into account the number of the system’s end users. How will this number change in a year, two, or five years? BI systems are often assuming 10-20 users. After a while, it turns out that dozens or even hundreds of other employees also use the system or are interested in using it. A huge number of requests and a continuous array of information in the system lead to its failure and, therefore, to the inefficient operation. Thus, the investment of time and money spent on the project is just wasted.
Therefore, a responsible approach while estimating the system’s scalability becomes one of the prior tasks during the implementation phase.
It is also worth to notice that a clear understanding of the goals and their competent statement will help create not only an effective analytics solution for business data but also a user-friendly product. Goals and objectives dictate how the data will be presented: required set of reports and dashboards, their priority, data sources of each report, how the reports and dashboards will look like.
3) The next important condition for a successful BI project is the availability of reliable and high-quality transaction data in the accounting system. Only successful use experience of an ERP system (or another system providing data for the BI system), as well as the proper regulation of the accounting process can enable the successful implementation of a Business Intelligence system.
That is why it is quite easy to predict the "boom" of BI implementations if the implementation of ERP systems boosted a year or two before that.
4) We would like to notice that the process of quality control should be carried out as close to the initial data as possible. It should be continuously conducted at all stages of data gathering and processing. To identify errors in an already prepared report is a very complex, time and resource consuming process.
5) The understanding by each responsible employee of his role in the overall process plays a significant role as well. Only competent work and precise fulfillment of all conditions at each stage of data collection, cleaning, and specification guarantee an effective and qualitative result. They are fundamental components of each BI system.
6) Iterative system creation
As a rule, it takes about 6-9 months to develop a working BI-solution. That is why the “step-by-step implementation”, starting from the most understandable and obvious elements is one of the most preferable methods. After implementation has been completed, you can scale the system capturing more and more input data, performing a deeper and broader analytics.
The implementation of Business Intelligence implies not only new software. Such changes involve the development of new work processes, new work positions or entire services (business analysts, financial analysts, etc.). Therefore, it is a long process, in the context of implementation itself, but also in the context of changes "acceptance" within the company.
However, there are cases when the company simply does not have any other choice but to launch the entire system at once. If it is a requirement of business, then the main task of management is to work out any possible problems and to organize the process as efficient as possible.