Better Data Quality, Better Business Results

In this acronym-rich industry of ours there’s still one acronym that I hold above all others because its truth is undeniable: GIGO. For those of you who might not remember this one, it stands for “garbage in,  garbage out” – and it describes what we all know too-well: the poorer the quality of the data in our ERP  application, the less value we’re going to get out of it.

GIGO has been with us practically since the dawn of the computer software age; and yet, in our excitement over new products, new gadgets, and (most of all) new catch-phrases, GIGO has to a great degree been pushed off to the side. Organizations have been too-accustomed (and too forgiving) in accepting “bad data” in their software applications, and when the economy falters and organizations downsize their staff, the incremental increase in bad data is simply accepted as the cost of doing business with fewer staff to do it.

Don’t you believe it.

Chalking up the increase in bad data to a reduction in staffing is an all-too convenient excuse, and – if anything – staff downsizing should bring the subject of bad data to the forefront of a business and make addressing it a corporate imperative.

But here’s the thing – most organizations think of data quality in terms of data entry; making sure that required fields aren’t left blank, making sure that validated fields have legitimate responses, and seeing to it that date fields contain date values, numeric fields contain numeric values, and so on. Now all of these things do improve the quality of data, but they are not all you can do to improve it.

Consider the following:

• Making sure that phone numbers have the correct number of digits
• Making sure that email addresses include the ‘@’ symbol
• Making sure that “follow-up” or “service” dates are no more than ‘x’ days in the future
• Making sure item on-hand quantities do not go negative
• Making sure that order discounts do not exceed ‘x’ percent (and that an order’s gross profit is greater than ‘y’ percent)
• Making sure that the customer data in your ERP application matches the corresponding data in your CRM application

The above is just a small sample of the types of “data integrity” conditions that you can watch for in your ERP system; generally speaking, these conditions resolve themselves into the following:

• Checking for missing data (e.g., an order without a “require date”)
• Checking for incorrectly-formatted data (phone numbers, tax IDs, email addresses, etc.)
• Checking for conditionally-missing data (e.g., if field ‘a’ = ‘x’, then field ‘b’ must have a value)
• Checking for conditionally-invalid data (e.g., if field ‘a’ = ‘x’, field ‘b’ cannot have values ‘y’ or ‘z’)
• Confirming that date, time, & numeric values fall within acceptable ranges (e.g., a “lease renewal” date cannot be more than 365 days in the future, or an “order discount” cannot exceed 15%)
• Performing cross-application data validation (e.g., a client in your ERP system should have the same email address as their record in your CRM system)

As mentioned earlier, most organizations’ attempts at improving data quality usually begin and end with only the first item above – making sure that required data entry fields are not left blank.

So – do yourself, your staff, and your customers and partners a favor; the next time you go into your ERP application and you see something that makes you want to say “How did that happen?” – don’t just chalk it up as the price of doing business. Implement a data-quality system that continuously checks for (and responds to) that “bad data” that can occur at any time – not just when data is initially entered.

After all, everyone makes mistakes; it’s having to live with them that we can all do without.