- Measure – Find out how bad your data is, so when you’re fixed it, you’ll know much better it is. Without measuring, you’re guessing.
- Cleanse – Perform a Deep and Full Cleanse. This process solves all your data issues in one project. Your aim is have the cleanest and most accurate data as is humanly possible. You have to have a starting point for very clean data, as it’s a challenge keeping it clean from now on. Measure it after the initial cleanse and compare to step 1, you should see a big improvement.
- Sustain – Now incorporate technologies, define processes and appoint suppliers that help keep your data to that high-level you created in step 2. By having on-going measurements, you will know exactly how good the data is or is not.
- My Data Standard – Define how each element of your data should look. This will help everyone know what’s expected.
- My Data Entry Guide – Most errors occur through data entry, so having a specific guide educating users on how to enter data is important. This is not just about how you enter an address or a phone number; it’s also about what you do when you find a duplicate contact, company, invoice, product, etc.
Every time you hire a new employee make sure data standards is a topic during their induction training. Creating the right mind set from the beginning will go a long way to data quality being good.