DataStage consultants How to separate the good from the bad.

[kduke]

Ray just posted 10 things he likes to see implemented in viewtopic.php?t=107469&highlight= .


Talents needed to deliver a good data warehouse:

1. DBA
2. Data Architect
3. ETL Architect
4. ETL development
5. Database expert
6. Database developer
7. Performance monitoring
8. Performance tuning (if they do not know how to monitor then they cannot tune)
9. UNIX admin
10. Network admin
11. SQL expert
12. Scripting expert
13. Security expert
14. Technical writer

Signs you have a good consultant:

1. A good consultant should obsolete himself by best practices and training in-house staff.

2. A good consultant should deliver changes within the first 30 days. If not get rid of the consultant or consulting firm. He should at least change the way you think about your data warehouse.

3. Documentation should be simple and direct and not overwhelm the client. Beware of complex documentation from big firms which may take years to deliver something simple. It is rare a good developer can write good document. Good developers are rare enough. Don't expect them to be good at everything.

4. Canned solutions lack the business edge to deliver serious corporate changes.

5. Canned techniques are more powerful in delivering custom solutions.

6. Good consultants listen well and do NOT talk down to the clients. They are not arrogant.

7. Good consultants tend to be stubborn but not abrasive, strong willed and have definite ideas. If you need a good independent consultant then you need to change something. If your in-house people are not good listeners and not willing to change then a good consultant will walk away from you. Your company is a no win situation for the consultant. He will listen to what your people are willing to adapt to and know that he cannot fit into your situation.

8. All good solutions involve risk. Follow your gut. If it seems not possible or too complex then doubt the consultant or consulting firm. When something is over engineered and deadlines become tight expect quality and deliverables to look different than described in original documents.

9. Quality Assurance should never look at staging tables or intermediate results, tables or any staging files or datasets.

10. If a solution is too complex for in-house people to maintain or modify then you are tied to the consultant or consulting firm forever. If you have a small staff then this maybe okay.

11. There is no substitute for experience. Big projects with lots of people are not much different than little projects with fewer people. The solutions tend to be just as complex and so are the people problems. The difference in big and little tends to be the amount of buzzwords used. If the amount of buzzwords is significant then I tend to doubt the quality of programming skills. Highly technical people tend to be better with technology and not so skilled with words and people. What do you need may not be what you want. If you need someone to manage then are talking people or tracking deadlines. Non-technical people maybe can track projects and maybe keep it on schedule but can't fix a project that is out of control to save their life.

12. Fixing an out of control project may require firing people. A project out of control is like a cancer. You may need to cut it out and then some kind of long-term attitude changes. Are you and your people willing to make those changes? Firing the wrong people will not fix anything. If you have poor management with unrealistic expectations then your project may just rotate consultants in and out like a revolving door. If good talent comes and goes then maybe management is the problem.

13. If you have a project which is going poorly or high turnover then being hard on the talent that has survived is not the solution. You better be nice to the in-house people they have all the business knowledge.

14. Stress is a killer. All of things just described are deadly to your soul and body. Being a consultant especially the lead consultant is not easy. Some can never fire anyone. Some avoid making serious changes trying to be nice. You reputation may suffer when you get paid to make the tough decisions. Try to be nice to some of the people your stress level will go down. I will post some stress relievers. When you are under stress then you may gain weight, eat poorly, sleep poorly, have headaches, have stomach aches and many more signs that you are sacrificing your health for your job. Money is not a substitute for health. Trust me if you do this too long you may never recover what you have lost.

I have heard of people lying on resumes lately and even interviewing on the phone for someone else so they can get their friend a job. Resumes can be misleading. Phone interviews can be too. Avoid long-term contracts so you can back out of a bad situation. If you outsource a solution complete with tight deadlines then these almost always start out one way and end differently. Throwing money and people at a poorly managed project may not change anything and quality may suffer.

[ray.wurlod]

15. Cynical (paranoid?) so as to generate effective test data.