Archiv der Kategorie: English

Ready for Big DATA

After more than one hundred years, the processing of data is still the main business of IT. The development and the architecture of IT-systems as well as the world-wide networking were at the center of the attention for a long time. Now the IT-industry rediscovered its original topic, the processing of data. With Big Data the renaissance of data processing has an irresistible title. The topic is interesting, because

  • Data is stored everywhere,
  • Contents are interpretable, and
  • the skillful question is difficult to find.

Big DATA offers remedies, in order to process the giant quantities of data, to facilitate the evaluation of contents and to simplify the formulation of the questions. What do the users have to consider in order being ready for Big DATA?

BigData Lupe

  • Data is everywhere
    Despite long-term consolidation and standardization of the IT, the corporate data is still saved in various systems and formats, on different media, and protected in various ways by passwords. Additionally the master data, like e.g. customer or product data are still stored redundantly in several databases that are controlled by different responsible executives. The data quality (e.g. correctness, completeness, consistency or timeliness) is difficult to assess. Eventually, apples are mixed with oranges and then evaluated. These disadvantages will be amplified by Big DATA – following the slogan: Faster Disaster.
  • Data is interpretable
    The interpretation of contents depends on several aspects. Actually, only the original creators of data know their initial purpose. Over time other employees use this data for daily business. They update the existing and add further contents with their own interpretation. Occasionally, the data will be evaluated, consolidated and prepared in tables and diagrams. The analysts interpret the results and derive new insights from it. In retrospect, the participants will not be able to remember, what their original understanding of the data was. The same will happen better and faster with the new Big DATA.
  • It is difficult to ask skillfully
    New approaches, like Data Mining, offer the possibility to search through data without distinct questions. Thereby, programs find patterns that will be examined afterwards. This unveils regular patterns that result from frequent repetition of fraud. On this basis concrete questions can be formulated, e.g. what did customer XY buy and did not pay. For the effectual use of Big DATA the users must learn to ask SMART questions: Specific, Measurable, Adequate, Relevant and Testable Before you start the examination, the indicators of the answers should be specified. Eventually, the format of the report and the structure of the result are determined. The achieved insights can be used more easily in the business.

Bottom line: It would be a fatal and expensive error to consider Big DATA as an automatism that provides the solutions for the operational tasks. The amount of data doubles itself every two years. Therefore we will have to deal globally by 2020 with 40 zettabytes according to IDC http://ow.ly/Ao5v7 . Now is a good moment to make yourself ready for Big DATA.

Buy a pig in a poke

The price is rarely an indicator for good quality. Some people always buy the most expensive in order to believe that they get the best quality. Others always buy the cheapest in order to believe not to pay too much. The right approach is the best value for the price. For this purpose, it is necessary to look at the proposals in a structured way. Each role in the procurement process has its own view – Purchaser: maximum price reduction; Finance & Control: to pay as little as possible; Supplier: maximum price for minimum effort: Buyer: maximum value-add. But how to evaluate the pig in a poke?

KatzeimSack Piginapoke

On the one hand, the roles need a clear division of work. On the other hand enforces a structured requirements the quality and the comparability. The proposals should contain the complete description of the deliverables, the required contract elements and the prize.
The evaluation will then be done according the following aspects.

  • Selection criteria
    Based on the following criteria the detailed proposals are evaluated: Basic information, like e.g. information about the supplier, project organization, further products and services, costs, Compliance; Functional aspects, like e.g. the available functions, processes, data, interfaces, administration, IT.
  • Weight
    Since not all criteria are equally important, the individual criteria should be weighted beforehand, like e.g. 1=Standard, 2=Important, 3= Mandatory).
  • Evidence
    The traceability of the deliverables creates additional gradation of the proposals. For this purpose, the proven elements are rated 10 and incomprehensible 1.
  • Price expectations
    The internal estimation of the costs sets the limits for the offers. If an offer is considerably under the expected figures, important functions could be missing. If the offer is significantly above the estimated figures, unwanted elements could pollute the proposal.

Bottom line: The approach for the evaluation of the proposals should be clearly descriptive in advance, in order to get an objective evaluation as quick as possible. The clear division of tasks and the standardized requirement specification are the pre-requisite for an effective procurement.