There was a discussion on a users group listserver a few days ago about the pros and cons of centralized vs. decentralized data entry. The question was, should all staff be able to enter and correct data in the database, or should that function be restricted to just a few? Several people weighed in, and the majority seemed to be in favor of concentrating that work to a few people.

Centralizing data entry is the easy choice, which is why most favor it. While I see the attractiveness of concentrating the data entry management into the hands of just a few, I think it’s the wrong choice.

It’s easy because ultimately, data management is about personnel (or staff) management. That is, your data management is only as good as the people who are doing it. So if you have fewer people doing it, that’s fewer people you have to train, keep up-to-date, and check on. Simply, it’s fewer people to manage.

But I think the downside is worse. Here are just a few:

  1. Data is processed more slowly. By definition, a concentration of the data entry processing is a bottleneck. For example, if I learn of an email address change today and have to forward it to the fortunate few who can make data changes, how long will it take for the change to take place? I know it’ll be a lot longer than if I had just made the change myself.
  2. Non-data entry staff stop caring about the data. Ownership of the data, by definition, now belongs to the chosen few, which means those who are not responsible for data entry will eventually stop caring about the data. Or worse yet, when there are problems with the data (there always will be), guess who will get blamed?
  3. Too many eggs in one basket. If you have only one or two people handling all the data entry, what do you do when those people are out sick, or worse yet, leave the company? There is little “bench strength” when data entry is centralized.

As I said, centralizing data entry is “easier” than decentralized, because in order for decentralized data entry to be effective, you have to have good documentation of business processes, good and continuous trainin, and diligent remediation with staff that aren’t “getting it.” You also have to have very clear rules about who can change what data.

I won’t kid you: It’s a lot of work. But the long-term benefits (e.g., cleaner data updated more quickly and more accurately, because fewer hands have touched it) make it worth it.