Recommendations about the choice of BI
Criteria of functionality and system performance should be regarded when choosing the solution for a business intelligence as of paramount importance. Selecting or updating a system for the business analysis, it is necessary to think over methods of storage and data integration, visualization tools and analysts.
The directory of BI solutions and projects is available on TAdviser.
Content |
Selection terms of systems
Many organizations, planning the BI project, carry out the complete assessment of BI systems presented at the market, but not everyone acts this way. First, passing of competitive comparison with the systems of the same class not always for the benefit of vendors therefore they try to avoid it. Besides, often the organizations consider the full choice of a system the expensive and requiring a large number of time project.
For this reason often the companies aim to pass this phase of implementation, and perform the choice of a BI system by the following criteria: whether another is used by software of the same vendor in the company how successful a system is combined with already installed products, management on the previous work is well familiar with a system, the supplier is strategic vendor for the company. But whether these criteria are sufficient?
BARC (Business Application Research Center), 2010
Business Application Research Center[1] demonstrates that those systems which were selected during competitive comparison with other products bring the positive effect significantly exceeding effect of projects where when choosing the solution the functionality only of one system or such assessment was estimated was not made at all. BARC also note that systems capabilities become a decisive factor of the choice whereas the value of vendor and cost of the solution decreases for most the companies now, and this trend accrues since 2002.
As for features of the key solutions presented at the market, certainly, they attract users at the expense of different, inherent only with it advantages. Having analyzed solution selection terms for more than 2 thousand projects, BARC reduced data retrieveds in the form of the tables reflecting technical and commercial attractiveness of the different systems. The main conclusion which is drawn by experts: on any question there is the answer, i.e. in different situations it is possible to pick up that BI system which optimum meets the stated requirements.
For example, according to data of a research for 2012, the greatest satisfaction in group of BI platforms as a result of implementation was got by users Oracle of the Essbase systems, IBM Cognos and QlikView. In this group vendors with income more than 200 million dollars a year or having really international presence are carried. Such platforms are usually used at the industrial level.
Degree of satisfaction of users of large BI platforms
BARC, 2012
In the following group of systems for the corporate reporting products which provide the deployed industrial instruments of preparation of reports are integrated. It is interesting that in this group Arcplan offering flexible tools for reports of any type on the second and third place on degree of satisfaction of users - MicroStrategy and Information Builders is in the lead.
Degree of satisfaction of users of platforms for the reporting
BARC, 2012
In group of systems for creation of deshbord (control panels) the first lines in general appeared behind niche vendors - Yellowfin, Dimensional Insight and besides Arcplan.
Degree of satisfaction of users of the systems of creation of deshbord
BARC, 2012
In group of systems with the developed means of work with OLAP are in the lead, the given BARC for 2012, the solutions BOARD, Targit and Oracle Essbase.
Degree of satisfaction of users of systems on OLAP
BARC, 2012
In group of solutions with the developed visualization tools and data discovery there were five vendors in decreasing order of the user satisfaction it: Dimensional Insights, Phocas, Tableau, Cyberscience, Qlik (QlikTech).
Degree of satisfaction of users of the systems of visualization
BARC, 2012
Analysts of BARC remind that commercial properties of a BI system still are at the head of a corner for most the organizations whereas their functionality is not priority criterion when choosing a system. However use only of one universal criterion – a system performance criterion – could affect effectiveness of projects essentially.
Main selection terms of BI systems
BARC (Business Application Research Center), 2010
Analysts of BARC developed a scale of factors which reflects as the organizations should range selection terms of a BI system and how it really happens. So, the performance criterion should be in the first place (the left column), and actually is on the fourth (the right column). And the criterion of support by the system of various server platforms is underestimated by business on 13 points: it should stand still 5, and occupies only the 13th.
Choice of functionality of a system
Data storage
If the company is faced by a problem of identification of long-term or periodic trends, i.e. users need to analyze the historical data arriving from different divisions within the last 3-5 years, then, most likely, it is necessary to think over more carefully the organization of ETL transactions for data loading to data warehouses.
If the company or any of its divisions needs to analyze information monthly or weekly, then selection and the organization for these purposes (for each of divisions or for the solution of specific objectives) separate data marts, also using ETL tools will be an optimal solution.
If the company is going to analyze operational data in the mode approximate to real time (i.e. updated several times during the day), then, perhaps, it is necessary to refuse the organization of the data warehouse and to pay attention to study of integration tools on the basis of an interface virtual layer of metadata with study of the corresponding interfaces and algorithms (by the principle of EII).
Data integration
As it was already noted above if the purpose of implementation of a BI system is the solution of separate, specific objectives, then it is reasonable to be limited to the organization of data marts. At the same time use of any separate integration algorithms will not be required.
If, on the contrary, BI is implemented for the purpose of receiving a uniform, complete view of shared state of business, then, perhaps, not to do without creation of the centralized data warehouse and, respectively, implementation of necessary ETL tools. Besides, for receiving really adequate picture of business it is necessary to pay special attention on high-quality providing the analyzed data, and implementation of an expanded set of means for their "cleaning" - detection of incomplete or wrong data for this purpose will be required, to the duplicated information, reductions of data from different sources to a uniform format.
If in the company place emphasis on studying of operational data, then it is necessary to think over means of replication and providing access.
Visualization and analytics
Depending on assigned tasks and also from qualification of users, also means for data visualization – control panels, score cards, reports, an OLAP cubes are selected.
For experienced, qualified users an OLAP cubes which will allow them to carry out the deep and unrolled business analysis, with necessary extent of detailing will become a best tool.
Users who in the daily activity are faced the need of acceptance of management decisions and also business efficiency analysis, are interested in the organization of a workplace in the form of a control panel on which in the form of evident scales and indicators the business status in general, with a possibility of switching between separate activities is displayed.
Means are necessary for ordinary managers for the solution of the current tasks, control of the course of accomplishment of separate transaction types and also for control of activity of the employees (each certain worker and a command in general). Besides, for the organization of accurate interaction with adjacent divisions (or regions) it is necessary to have an opportunity to gain an impression about the course of accomplishment of the interconnected tasks.
Vertical or horizontal solution
In the market exist as horizontal BI solutions in which a set of generally applicable tools, and specialized vertical solutions "ground" under specific industries or tasks is implemented. Both those, and others have the merits and demerits.
At advantage of horizontal solutions their capability together with the organization can be considered grow. Such solutions are usually scaled and can cover all activities and all divisions of the large company and also give in to changes easier. A reverse side of such width of opportunities is need of longer and careful setup of solutions, adaptations under specific requirements. Implementation projects become more expensive, and requirements to IT specialists – higher.
Vertical solutions, in return, do not require separate long and labor-consuming setup for the solution of specific tasks and for compliance to requirements of the industry regulating organizations (financial, medical, etc.). However it can turn out that different divisions within one structure will not be able to use single solution, and mastering and integration of several different systems for the business analysis will be required.
Those organizations which both now, and in the future are going to be engaged in the specific activity requiring observance of certain strict regulations will be benefited, most likely, from implementation of vertical solutions. If there is no confidence in such commitment to a certain type of activity in the future, and the probability that specialization of the company considerably will extend is high, then the choice of the vertical BI solution is a certain risk.
See Also
Business Intelligence determination
Standard blocks of the modern BI systems
Classification of the products business intelligence
Advantages of use of a BI system
Implementation of BI. Typical errors
Business Intelligence, BI (world market)
Business Intelligence (Russian market)