So you just updated your database and added some shiny, new features. They look sleek and should allow you to collect all sorts of useful information, but do they work?

Testing new features is an essential step to ensure valid data and an intuitive user interface.

Why test?

Data is an asset to businesses, so it is worth investing in its validity. While you can sometimes catch errors just by noticing mis-formatted data, it is more efficient to complete rigorous testing to find problems before they arise.

What to test?

First of all, when a user inputs information into your database, you want to ensure that it is landing in the right bin – make sure that phone numbers aren’t mapped to the home address column by mistake. If a user goes back and updates or deletes their information, that change should be reflected in the behind-the-scenes spreadsheet. It’s also key to note whether once saved, data stays saved. A computer crash shouldn’t lose all of your hard-earned information. 

Next, data should only save if it’s correctly formatted. An entry,  like inputting a new hire, should only save  if all of the key information (e.g. name, contact information, and position) is present.  If someone tries to save an incomplete entry, an error message should appear.

How to test?

To test your database, create test use cases or sample things a user would do. Test actions you know should work and ones you know shouldn’t to see if the database responds how you would expect. For instance, the database shouldn’t allow a four digit phone number to save, but it shouldn’t have a problem with a ten digit one. Also take note if deleting data that is mapped to multiple tables causes errors. 

Even correct data becomes useless if it isn’t processed correctly. To ensure that a formula is doing its job or that a “save” button works, try inputting clean data into a procedure when you know the expected outcome. Ask yourself whether your new database function is fulfilling its purpose, getting the anticipated result, and meeting the user’s needs. 

On a final note, make sure to test sooner rather than later, so you don’t have thousands of lines of data to fix. Now go keep those databases clean and running smoothly!