- Did you know that 10 to 25% of customer information is incorrect?
- And in a year almost 25% of your data goes stale
- According to MarketingSherpa, an average of 2.1% of contact details change every month!
Data is plays an integral role in key decision making in any organization. With it almost doubling every year, it is essential to ensure that quality is kept intact and accuracy is improved. B2B marketers need to realize that correcting incorrect information is a costlier affair.
Whether it is required to find updated and targeted sales leads, or to increase ROI from marketing investments or to screen verified candidates while headhunting, data is the core essence.
With defined strategies, planned processes and an ongoing action plan, it is possible for organizations to ensure data quality is constantly maintained.
B2B managers popularly use the ‘1-10-100’ rule:
$1 to verify a contact as it is entered.
$10 to clean existing data and manage duplicates.
$100 per record to correct, if you do nothing to clean bad data.
One the most important data quality metrics is DATA VERIFICATION.
What is Data Verification?
It is a process that ensures that data is correct and error free. It is entirely factual and does not focus on the significance of the result. This is where validation is important, which is a process that ensures data is logical and reasonable. Verification is conducted to ensure data entered is true to the original source. Some of the methods used include:-
- Proof Reading – data is checked by manually comparing it with the original document.
- Double Entry Checks – data is reentered and the two copies are compared for discrepancies.
- Other methods also include phone and email verification.
While manual methods may be time consuming and expensive, new software technologies have been developed to automate the data verification process, proving to be cost effective as well as faster.
‘High Volume’ Data Sources Are Not As Clean As You Think!
- Online Data Sources – while using data collected from web forms, web registrations, lists, spreadsheets and other online data collection tools; do not directly dump it into the database. More often than not people do not enter information correctly online. Therefore this data needs to be scrubbed cleaned of errors before contaminating an existing database.
- Crowdsourcing – is a community driven concept data hk wherein data is contributed by users. Various groups and people provide information such as email addresses, phone numbers and social media profiles. This source provides unlimited amount of data and information, making it possible build databases that could effectively utilized.
However, crowd-sourced data has to be verified and validated before being added to the database. Encourage users to update details to ensure data is as complete, correct and current.
Data Quality Best Practices
Verification is integral in maintaining data quality, accuracy, relevancy and currency. Data needs to be collected, compiled, organized and maintained in a way that it provides insightful information. While new processes and technologies have been developed to maintain ‘clean and green’ databases, a simple method to begin with could be:
- Formalize strategies – Companies need to define their objectives to ensure data management is organized and databases built in direction with the company’s strategic goals.
- Fine tune processes – To ensure strategies are implemented efficiently, processes need to be streamlined. Creating a map is a great of identifying different stages of the data cleansing process.
- Focus on data integrity – Once data cleansing programs are in place, it is important to integrate clean data with existing databases. To avoid clutter and muddling of old data, with new data or bad data, it is important to ensure integration processes are tight.
- Formulate an ongoing data cleansing program – Cleansing data once is not going to clean new data that pours in on a daily basis. An ongoing program needs to be formulated to maintain quality of new information being introduced.