retains value to your business. Consider doing some of the following:
—Keep past and present data in a consistent format. If you decide you need to change a format (e.g. changing a column name from “Age” to “Birth Year”), go back and update the past data, if possible. If not possible, create a new field so that it is obvious something changed, as opposed to changing the meaning of an existing field. Document what was done and when so that past data can be integrated into future analysis more easily.
—Run test queries on a regular basis on the new data so that changes or bad data can be detected as early as possible.
—Back up your data regularly, both on-site and off-site. You want to be prepared for user error, disk failures, theft, and acts of nature. A backup stored near the original only protects from the first two. Test your backups periodically—they are of no use if they aren’t being done correctly.
—Worry about privacy and security from the beginning. Having to tell all your customers that their personal information was compromised is not good for business. Something as simple as a stolen laptop can be a reportable breach. Encryption and not storing unnecessary personal information are straightforward means of protecting your business. Using third-party vendors with advanced security for things like credit card transactions can off-load this worry.
There are many more things you can (and will) worry about, but if you keep some of these things in mind at the start, you can be prepared to begin using your data much sooner.
Most of what I recommend above can be automated, and with little or no extra effort beyond setting things up correctly, you can be in a great place to start doing analytics as soon as possible.
