How to Identify and Clean up Bad Data

How to Identify and Clean up Bad Data

cost of bad dataThe implications of bad data are huge, as we discussed in this post. It causes not just direct financial losses to the tune of $600 billion annually in the US alone, but it could destroy a business. Businesses need to understand what causes bad data and take proactive steps to remediate the causes.

1. Automate: The number one reason for bad data is clerical errors as data is entered manually into systems. Automating manual steps such as moving figures and summations would reduce bad data significantly.

2. Shun the obsession with more: The hype on big data notwithstanding, collecting excessive data may become counterproductive, leaving no time for the marketer to undertake any meaningful analysis or evaluation of the collected data. Also, it could lead to data overload, increasing the chances of errors or wrong analysis. The lack of use of information leaves little incentive to improve data quality. A marketer is better off trying to put the available data to better use rather than seek more and more data and dilute its usage.

3. Seek out consistency: Fragmented information systems and blurred controls could lead to significant duplicate entries. Data scattered across disparate databases makes it difficult to mine that data for useful information, or use such data to build and improve customer relationships. Investing in a robust system and procedure upfront with regards to data collection would lead to better data consistency and eliminate the menace of duplicates. Some ways to do this include making changes in data collection forms, harmonizing methods, and instituting routine checks on data quality.

4. Undertake Data Audits: The best of precautions notwithstanding, bad data invariably creeps in due to unavoidable manual entries, technological glitches, deliberate wrong inputs by customers, and other reasons. The solution is a periodic data audit, to review key customer touch points and audit the customer experience. These audits should ideally review how data is collected and how the collected data is being used. Specific points for data audits include:

  • Reviewing the data collection forms and eliminating the collection of superfluous or unnecessary information.
  • Ensuring that data collection is consistent across the various touch points, including websites, ads, point of sales terminals, social media and more.
  • Using tools such as the ones offered by Business Objects and IBM to clean up inaccurate or contradictory data.
  • Place an expiry date on data, to update or delete stale records.
  • Enrich data with outside sources, to update stale records, such as a contact’s change of job
  • Break down data silos by ensuring that the various data related efforts go into a centralized data mart
  • Ensure that there are proper definitions and validations in place, so that the data entered is of high quality in the first place. One important check is searching for duplicates before entering new data, to avoid reinventing the wheel.
  • Ensure that data confirms to standard conventions and definitions, to eliminate misunderstandings.

Adopting data quality best practices could boost revenues by as much as 66%, and in most cases more than compensates for the time and money invested.

One way to improve data is through system integrations. ReadyTalk integrates directly with Eloqua, Marketo, Salesforce and other platforms to automatically move your webinar and conferencing data between platforms. By eliminating manual processes, the quality of the data is greatly improved. Learn more about ReadyTalk integrations.




Share this: