2014/10/28 07:00:00

Implementation of BI. Typical errors

According to analysts of Technology Evaluation Centers (TEC), only about 30% of implementation projects of BI systems are successful. The main problem is even not the release of the project for a framework of the budget or the taken-away temporary borders, and mismatch of results with expectations.

The directory of BI solutions and projects is available on TAdviser.

The analysis of others errors, as we know, is capable to help competent heads to avoid own. There is a number of factors which with guarantee lead to failures of BI projects. Therefore to implement implementation it is successful, it is necessary to consider and provide these factors to them counteraction.

Business need for use of BI

Prior to the beginning of the project it is necessary to define absolutely precisely what benefits will be brought to business by implementation of BI. Often implementation of a BI system it is caused by the growing competition and need to give the answer to market calls. Sometimes implementation of BI happens relevant at merger of other company and it is essential increase in scales of business: in that case, BI tools I allow to provide its transparency.

The BI project should be initiated and supported by business users, but is not imposed by IT specialists. It should not become one of standard, standard steps of development of corporate IT infrastructure just because "it is so accepted", "it is prestigious", etc. Probability that implementation of BI will change for itself the solution absolutely of other tasks, such as business performance management, data integration, ensuring the best interaction of divisions and users, business process optimization is in that case high.

Goal setting and project tasks

The first and necessary step at implementation of BI is study of the purposes and tasks. Strict determination of the purposes and tasks will help to select best tools for their solution and will provide understanding from end users.

It is necessary to think over and estimate scales of implementation of the BI solution at the organizations also. Sometimes, that initially a BI system is created counting upon 10-20 users, and it turns out that hundreds of employees are interested in use of its opportunities (on extremely measure, regarding report generation). In this case a system will not be able to cope with a flow of requests and will work insufficiently effectively that will undermine in users confidence in its quality and stability and, perhaps, will force them to return to former tools for the business analysis (spreadsheets and routine handwork). Thus, all efforts and means spent for implementation will be vain.

Having estimated efficiency of implementation of a BI system in the next division, heads can demand to implement at themselves the same solution. However the tasks solved by this division and also volume and efficiency of the analyzed information can strongly differ. Therefore it is necessary to treat scaling with care and to regard it as the separate independent project.

Business support

Both actual, and sponsorship of the project from the top management have, perhaps, the most significant role for a project success. It is very important that management took personal part in the project and had clear idea of specific benefits of implementation of BI. It is ideal option, especially if in the project progress there will be any difficulties. On the other hand, when several managers work on the project at the same time, it can turn back a situation when each of them tries to keep the project in the direction. The situation can warn this, in advance having approved the purposes, project tasks between them and also having defined what divisions of business and in what degree will be affected by the project.

Ensuring collecting and quality of data

Separate task within the BI project is problem solving of collecting and processing of initial data. It is necessary to think over what information will be used for request processing and report generation now what data can demand further with what frequency it is necessary to synchronize data and also as detailing, consolidation, data scrubbing will be performed.

Now very few organizations using BI carefully think over and provide necessary quality of data. However the correctness of data is a necessary condition for acceptance of right and timely management decisions that finally is a criterion of quality of the created BI solution. Often it is considered that the problem of data scrubbing should be solved by means of a BI system by default.

Occurs as well opposite approach when the problem of ensuring quality of data is completely shifted to shoulders of staff of the divisions delivering initial data or using BI. Meanwhile, it is a task complex. A part of BI platforms incorporate tools for quality control of data, however even them not always happens enough. Sometimes implementation of separate products for integration and data scrubbing is required, and the solution of this task develops into the independent project.

Choosing a Solution

You should not select such solution which exceeds requirements of the company essentially. If only 30-40% of systems capabilities are involved in the project, so a real opportunity to achieve ROI smaller costs is missed.

Also widespread mistake which is made by many companies implementing new technologies is Perekastomization of a system. You should not forget that providing exact and reliable information on most rationale should remain a main goal of all efforts. Sending reports for the smartphone with the inline images can be configured and later.

Management of the BI project

The problem of project management is key for most the organizations implementing BI, especially when it comes to IT resources. BI technologies often are beyond the approach to implementation of business applications which settled in the company. The matter is that information resources which are located by different divisions are necessary for work of a BI system. In the course of extraction of these data and their delivery in storage there can be difficulties. To avoid them, it is necessary to follow such recommendations:

Consider the possibility of appointment of a certain employee with a certain level of responsibility for all BI initiative who will receive the reporting from IT specialists and business of sponsors concerning all activity on the project
Found the special program for support of this organizational initiative
Appoint the selected specialist responsible for the analytical reporting, in each functional area which will control compliance of reports to the plan and process of their production. Degree of its competence should be happy is high, but can be reduced in process of "ripening" of the BI environment
Discuss the prospects of development of a BI system and its support with the top management
Initialize collecting of analytical information with the help of both vertical, and horizontal communication schemes.

User aspect of Business Intelligence

How a BI system in itself was good, the project success is provided by users. That the staff of the company "accepted" a new system, they should be involved, first, at the initial stages in processes of formulation of tasks and purposes, assessment and the choice of the software product and the implemented tools, and secondly to train in all opportunities of the created solution.

The choice of the correct moment for carrying out training – one of necessary components of successful implementation. If to provide training too early long before a system is put into operation, then users will not be able to correlate knowledge acquired by them to the solution of the specific routine tasks, but also will manage to forget all explanations by the time of start of a system. If to provide training too late, then users will brake and sabotage implementation unclear and frightening them by the novelty and complexity of the software product. The optimal moment for carrying out training, perhaps, is the period just before start or at the very beginning of operation of a system.

Problems of adaptation of BI systems are not in area of competent data structure and technologies, and in the field of personnel management and overcoming cultural contradictions. Any changes, as a rule, originally cause an unacceptance, and time that users got used is necessary, adapted to new processes. Besides, it is necessary to create and control observance of a system of rules which will not allow users to return to the former, usual modes of work bypassing a new system.

System efficiency evaluation

Assessment of payback of BI solutions – a difficult task and not having for today the standard conventional solutions. There is an opinion that to count return of investments from implementation of BI during the initial stage of use impracticablly as main advantages will become obvious only in the long term.

For assessment of potential success of the BI project experts suggest to study similar implementations in other companies and to compare their results to own purposes and tasks.