Tuesday, October 11, 2011

50 Business Intelligence Failures

1. Inadequate integration knowledge on Business Intelligence and Decision Support System Level. Dashboards are not good solutions.
2. Not seeing complete Business Intelligence picture.
3. No vision of what is final goal of BI tool. What are final outcomes and how will they be used for company prosperity and competitive advantage.
4. Demands are created faster than Management Information Systems can absorb, implement and stabilize = Generating to big cumulated requests.
5. No or little relation between financial statements and non financial key performance indicators.
6. Lack of integrative systems in Legacy level.
7. No or inadequate Mater Data Management solutions.
8. No or inadequate Data Quality processes.
9. Minimizing data quality issues by customer.
10. Too high expectations from Business Intelligence. It is not an Expert System.
11. Business Intelligence solutions are specialised. Can not cover everything.
12. Out of the box solutions cover minor part of your current process and required functionalities. Be prepared to change processes more then to customize out of the box solution.
13. Too dynamic complex market, like telecommunications.
14. No internal technical knowledge on management and on expert level
15. No internal dedicated team of experts to cooperate with vendors.
16. No internal resources to handle knowledge generation.
17. No internal structure to handle development of Business Intelligence layer.
18. Wrong project lead.
19. Too many internal enemies.
20. Too ambitious management.
21. Too naive management and project lead.
22. Wicked management. Political games can destroy any project.
23. Political games on management level. It is easy to declare "not needed any more".
24. Lack or inadequate internal marketing.
25. Wrong scheduling of modules. What comes first, next and at the end matters.
26. Too long implementation time. Many other projects with significant impact happen during implementation time. Each impact could mean new code writing and starting from beginning in certain segments.
27. Wrong vendor selection.
28. Selection of wrong software solutions.
29. Not enough resources to finish project. Especially in cases of additional costs that can double or triple initial investment.
30. Lack of data definition or poor methodology.
31. Too little integration with comptrolling. Comptrollers stay in their world with separated solutions.
32. Internally made solution, more likely to fail than vendors solution.
33. There is no one stop shop vendor solution. Each solution should be compared with leader in particular segment.
34. Business Intelligence project should evolve.
35. Can outsource the whole thing. Company must avoid the temptation of outsourcing everything and only things that are not core competences.
36. Just give me the dashboard. Companies must have a solid and stable BI infrastructure before implementing dashboards.
37. Information chasm between financial and non financial systems is too big.
38. Inadequate consultancy.
39. Trusting consultants too much.
40. Start with internal development without asking users what they need.
41. Mega requesting appetite. Data Warehouse/BI must cover all requests without any exemptions.
42. Include all departments and business units, especially call on workshops as many people as possible. Nobody should say later that was not informed. Some call it spam with project but don't believe to gossips.
43. Jumping into BI and Data Warehouse project without own IT administrators. Why should you have them since it is out of the box solution.
44. Leave all operative work to consultants.
45. Believing without any doubt to presented models and presentations of vendors.
46. Scheduling and starting in parallel major upgrades of legacy systems.
47. Not giving up from processes and not willing to change them.
48. Modifying project scope and making false promises just to enter into company.
49. Insufficient customer specification and lack of vendors notification about it.
50. Staying too tight to signed specification. Not allowing single change.