skip to main content

Extracting Value from Data

  • Joy Baker

    Joy Baker

Created at July 11th, 2019

Extracting Value from Data

At Advertising Week Latin America, a recent marketing industry conference, I visited with several retailers interested in sharing data insights associated with their customers to other companies or key strategic partners. As an example, information received from consumers who sign up for marketing promotions or who join loyalty clubs is not only valuable for customer acquisition and improved customer relationship management but, if gathered appropriately, can be monetized to drive even more relevant marketing for consumers across the marketing ecosystem.  These data arrangements can enable monetization, analytics and reporting partnerships for companies seeking to leverage data as an asset.  To do so, however, there are a few things a retailer must first consider.

1. Is the data collected in a privacy-compliant manner?

Privacy laws, regulations, and best practices vary by country. Therefore, data buyers are looking for companies that not only know these local regulations but also ensure that their data is collected in a manner that is compliant, including appropriate documentation. 

Best Practices: 

  • Document the data collection processes.
  • Keep an indicator on each individual record regarding how and when the individuals consented to allow their data to be collected and used.

2. How is the data sourced?

Data buyers want to understand how the data is sourced to gain insights regarding data consistency, potential bias, and accuracy. These insights help determine the appropriate uses of the data along with how to best combine it with other data sources.

Best Practices:

  • Document the original purpose of the data collection.
  • Clearly articulate information related to consistency of the data. 
    • Are all data inputs collected in the same manner and for the same purpose, and what are they?
    • Are all the fields collected at the same data aggregation level?

3. How current is the data?

Knowing the currency of the data provides valuable insights to data buyers. Buyers often purchase multiple sources of data to improve insights, and it is valuable when combining them to know which records are most current and which records are older. Data sets that include accurate collection and confirmed dates on each record are more valuable.

Best Practices:

  • Share whether the data is regularly updated or if it is a static data set.
  • Store a minimum of two dates on each individual record – one for when the individual originally provided his or her contact information and accepted appropriate privacy options. The second is the most recent date the individual’s contact data was confirmed. A third date related to the most recent date the data values were updated would elevate your file above the rest.

4. Is the data quality and coverage good?

Data that is incorrect, full of random characters, or not consistent isn’t very valuable. Data providers are expected to maintain high quality in the processing of their data assets. This includes removing records that contain characters that are not appropriate for the field and ensuring the data individuals have provided makes sense for the data field. 

Data coverage is also very important. Data buyers are looking for data sets that contain a large number of non-default, valid values. 

Most data buyers will not purchase a data set without receiving a sample. The evaluation will include looking for inconsistencies, incorrect data, low data coverage percentages, and whether the same individual is included in the file multiple times without being indicated as the same person.

Best Practices:

  • Include basic data cleansing in the data preparation processes
    • Remove or flag deceased individuals 
    • Remove or flag duplicate individuals 
    • Remove incorrect data (i.e. the year of birth value for the individual = ‘1832’)
    • Remove incorrect or garbage characters (i.e. an email address that does not contain the @ character)
  • Standardize common data field values (i.e., for state name field, choose either ‘Mexico City’ or ‘DCMX’ or ‘MX – DF’ as the standardized field value)
  • Ensure the data field coverage estimates provided to potential data buyers represent the non-default, valid, positive values. (i.e., coverage counts for a field should not include the field value of ‘No Reportado’)
  • Provide information about how different data inputs were combined as well as a summary of data cleansing and validation processes. 

5. Is the data unique?

Interesting and unique data about individuals and households that is privacy-compliant, current and with good coverage is in high demand. Marketers are seeking to understand consumers’ attitudes, interests, and behaviors to improve their overall marketing strategies. Additional options to drive monetization revenue from data are for use in analytics or reporting.

Best Practices:

  • Highlight elements that are interesting or unique.
  • Ensure the interesting and unique elements are refreshed as often as possible.

The monetization of consumer data if sourced, processed and distributed ethically and appropriately can benefit both consumers and marketers as it drives more relevant marketing.   For more information on the monetization of data, email