A member of the PRSPCT-L listserve (for prospect researchers) asked "What do other institutions do to regulate or “police” data entry work, specifically when you have multiple users with authority to make updates?" 

It's a tough problem.  Here are a few approaches (other than locking down data entry):

  1. Policies. You need clear data entry standards.
  2. Security. You need to control who can make changes and what they can change.
  3. Training. No one gets access to change data without training on your policies. Training should be targeted to role. You don’t need a week of training to enter a contact report.
  4. Monitoring. Someone (or multiple someones) needs to review new and changed entries.
  5. Retraining. When people make mistakes, bring it to their attention. Some organizations make that person fix the mistake. If they continue to make mistakes, you need to retrain them and be prepared to take away their data entry access.
  6. Automation. You should at least run reports on a regular to look for errors. You may be able to automate some fixes. You might also use vendors to perform cleanups like merge/purges on duplicates.
  7. Acceptance. Some things have to be fixed by your office or by Development Services. No one is going to fire an effective fundraiser because she’s bad at data entry (although they might get her an assistant who will handle the entry). Donors will give online and misspell their own names, or leave the caps lock on.

Decentralized data entry is efficient and is the only practical approach at some organizations. But the more people you have entering data the more problems you’ll have with quality control. If no one's in charge of quality control you're going to have a database full of garbage.  For more on this, see my posts:

New Year’s Resolution: Be Good To Your Data

Best Practices for Managing a Database