Data quality issues affect a high percentage of datasets across tens of thousands of businesses. But many businesses fail to spot the danger signs right away. The longer the problems go undetected, and the longer people fail to act, the more expensive it will be to fix the data quality issue.
And it’s not just a question of cost, as we will see.
Small businesses think data quality can’t be a concern, since they know most of their customers on first name terms.
Medium-sized businesses see their CRM building nicely and don’t realise how quickly it’s decaying.
Large businesses dip in and out of data without assigning anyone a lead role in caring for it.
But data quality is critical to every business in the list, since all are vulnerable to performance and efficiency problems.
The small business could offend a customer or client by getting the details wrong, and the medium-sized business may fail to capitalise on growth once its database starts to age. Large businesses may be sitting on a data warehouse full of inaccurate, useless data; by the time anyone does anything about it, it may have generated thousands of hours of extra work.
So how do you spot a data quality problem? What are the warning signs?
Breaking It Down
Different data stakeholders interact with data in different ways. The feedback you get from one department will be different from another, but everyone has their own way of detecting data quality problems.
Here are some of the red flags to look for, across all departments:
It takes a long time to deal with customers when they call on the phone
When using a CRM system, staff proclaim, “I have no idea what I am supposed to type in this field.”
Customers get frustrated because you can’t see previous conversations they’ve had with other team members.
You have to audit all of your contact records before using them.
You can’t get any of your reports to match up.
Every time you change a record, someone or something changes it back.
Your email marketing campaigns constantly generate a high bounce rate.
You can’t find the information you need on the system you expect to use.
Your customers are complaining that they cannot unsubscribe from marketing messages.
There are no required fields in your forms so you just skip over everything but the name.
When you search for a customer in your Customer Relationship Management (CRM) system, you see the person two or three times and you don’t know which one to use.
You spend half of your day trying to get the computer systems to do what you need them to do.
People think you are spamming them.
Some contact records won’t save because the entries are invalid, but you can’t see anything wrong.
You don’t know how to type an international telephone number in a field.
If you have two versions of the same person, you know which one to use because someone has typed NEW or DUPLICATE on the end of their surname.
The CRM is asking for a state, but your customers are not in the USA.
You find it difficult to train new employees because there are so many learned workarounds to facilitate data entry.
Your company is getting huge amounts of returned mail after each mail marketing campaign.
Sales teams say the CRM system is giving them false leads and wasting their time.
You have invalid entries in contact records: dates that aren’t dates, genders that aren’t genders, or strange characters in fields.
There’s no point ringing anyone because all of your phone numbers are wrong.
You use your own spreadsheet to capture data because the system doesn’t work.
If you hear your team mention any of these problems, you have a data quality problem. And this is by no means a definitive list.
What About Security?
We think of data quality as being a convenience and cost problem, but it can spark all kinds of unwanted chain reactions.
In our list, the last point – number 23 – is the one that will worry your CIO.
If employees are not using the tools you have provided, you are paying for systems that don’t work, and you are paying each person to come up with alternatives. So business data may be saved randomly, with no security, on multiple devices and potentially outside of the business’ control.
They might be recording customers’ details in draft emails, in a notes application, on an unencrypted mobile device, on a memory stick, in an excel spreadsheet on their desktop – the possibilities are endless, and the consequences disastrous.
Own the Problem
Any business can have a data quality problem, and the biggest indicators are employees. They are the ones using the data; they are the ones who stand to gain the most when data quality is maintained. As quality drops, their jobs get harder, and morale crashes.
While maintaining data is a joint effort, the business must take ownership itself. It is the CEO, the managers and the board who are responsible for owning the data and managing it effectively. These same people must make it fit for purpose.
The first step is to implement a data quality solution that meshes with existing systems. For example, a tool that integrates with the CRM will clean data without the need for importing and exporting. Cleansing, deduplicating, matching and enhancing records will help to get the problem under control. Control over data entry – such as checking form values in a field – will help to maintain a better standard of data going forward.
Once the data is clean and nurtured, there is no reason for any employee to take that data into their own hands. The business benefits from a leaner, more accurate dataset, and improved efficiency that directly benefits its bottom line.
This blog was written by Martin Doyle, CEO and Founder of data quality software company DQ Global. From time to time guest contributors write on the Workbooks Blog – have something to say? Email the Workbooks Marketing Team on email@example.com