Sign up for our newsletter and get the latest big data news and analysis.

The Business Costs of Bad Data

Tara Kelly_SplicesSoftwareIn this special guest feature, Tara Kelly of SPLICE Software discusses the link between Data Hygiene and Customer Retention and also how to gauge where your data stands so you can clean it up and reap the rewards. Founder, President & CEO of SPLICE Software, Tara Kelly, has a passion for enabling clients to engage in a meaningful, Data Driven Dialog(TM) with their customers. As a serial entrepreneur who has developed three companies including one outside the technology field, Tara’s expertise is multidimensional but focused on creating businesses that use technology to enhance operations, service and the customer experience.

You understand how important data is to the success of the organization, but can you always trust the data itself? It’s vital for marketing, sales, customer communications that are important to their life-cycle, and high-level strategic decisions, so data needs to be accurate, timely, and properly organized to provide actual value. Poor data can result in missed opportunities – causing your business to fall behind competitor launches because your data didn’t show the trend.

“Poor data quality has a ripple effect. It creates competitive disadvantage, bad strategy, lost productivity, customer relationship and financial loss. However, by focusing on data hygiene – maintaining your good customer records, and fixing or appending incorrect ones – all businesses can see big results,” said Joe D’Agostino, Audience Solutions, at Acxiom, a leading a enterprise data, analytics and software as a service (SaaS) company.

Messy Data Results in Hard Costs

TDWI reports poor customer data costs $611 billion each year for U.S. firms, a staggering amount of money that impacts companies on multiple levels. This number is likely much bigger than most executives imagine. It also presents an optimal time to use the 1-10-100 rule to examine how impactful bad data can be over time, and the compounding effect of the errors.

The 1-10-100 encourages businesses to always look at data as a hard asset, one that can swing the fortunes of a company either up or down.  This rule is used when dealing with situations where the quality of something is in question and you need to judge the cost of fixing the problem. When used in the context of data, it shows the need to keep high data standards over the entire lifecycle – essentially when the business is formed all the way to the end.

For data, the rule means verifying the quality of a record incurs a cost of $1, cleaning that record costs $10, and then working with a record that has never been cleaned means a cost of $100. It underscores the problem with delaying data cleansing and the need for high quality initial data.

Data Obsolescence Occurs in an Instant

Cleaning data is not a “one and down” type of situation, even if the cleaning is performed flawlessly. This is because the incoming data is not static for most businesses, it is rapidly changing and coming in to the organization from multiple channels.

For an example, let’s consider the demographics of the United States. Consider for the argument that you hold a database that contains all of the adult population of the country. The data includes employment, address, marital status, and other related information. In just one hour, 5,759 people will change jobs, and in one year 1.4 million will retire from work. Five hundred and fifteen people will get married in an hour (potentially changing their name and address) while, in a year, 4.7 million individuals will tie the knot. Also, in just an hour, you’ll see 186 people declare bankruptcy and 263 go through divorce. In a year, 2.04 million will buy their first home, and a staggering 43 million will move.

The exact data points in this example aren’t important, they’re simply stated to show the scale of changes that will occur in a database at both short and long timescales. Allowing the data to stagnate means missed opportunities, misguided strategies, and the potential for losing established customer relationships. The solution? Focus on data hygiene, but leverage efforts to get the most out of quality customer records and efficiently work to fix the bad data.

Instituting Data Hygiene

You need to establish a baseline, by instituting an assessment on the health of the information. The assessment phase is intended to give you a realistic picture and force you to admit if previous best practices were not optimal when it comes to the quality of the data.

As the picture develops about the strength of your information, it’s time to quickly take action by cleaning the data immediately. You want to verify information, remove duplicates, and fix larger-scale formatting problems so you have a pared down and more usable set of information. Ideally, you want to engage in cleaning before you desperately need the data.

Remember the hourly and yearly timeframe shifts in U.S. demographics, and the similar changes that occur in myriad other data sets? That should push you to clean records frequently, setting protocols in place, and schedules to combat data decay.

If you enlist a third-party solution provider to cleanse the data, beware of many hygiene firms that are very expensive compared to the results. Rely on a well-reviewed and experienced firm that will give you a competitive edge through clean and current data.

 

Sign up for the free insideBIGDATA newsletter.

Leave a Comment

*

Resource Links: