by Emma Richardson, Operations Manager, Sales-Link Inc.
More big data companies are pushing for automation for standardizing and validating data. For a company with a database of over 500k contacts, automation will become a necessity. At Sales-Link, almost all of our data has been validated manually up until this past year. We have shifted our validation process to a Google tool that searches to validate our information. There are 3 things we must ask before we shift to full automation:
1. What are the threats of relying solely on automation?
a. Reliability of the tool - there are many reasons why automation tools stop working. Servers crash; code becomes corrupt. These are obvious threats that would complicate the automation process.
b. Automation may only collect raw data - a tool used for obtaining information is only half the picture; you may need resources to further abstract, parse the information into a standardized format.
c. Exact matches are hard to come by - a tool can locate potential prospects, otherwise known as “scraping”, where you try to match something in your database to either add to, modify, and/or validate. While this will speed up the task, you can end up with a lot of misspellings, abbreviations, and many other nuances that causes problems with matching, or aligning information.
d. False Positives – in the same way the tool can fail to match the search with the prospect and fail to verify, it can wrongly verify information for another contact. There can be prospects at the same company with the same title that have similar names or the same initials. They are different people, but the tool will read it as a match, and wrongly validate.
2. What can we put in place to protect us from these threats?
a. First, don’t ever trust automation 100% and have checks and balances in place.
b. Regularly track numbers-amount of tries and successes of your validation tool. If you see a drop in executions, you know something went wrong with the tool.
c. Do not rely on one resource to validate information in your database. Use sites where the prospects opt-in to giving their information updated by themselves instead of websites that harvest information.
d. Don’t be afraid to employ risk-based verification by your quality control (QC) team.by setting up a report of prospects that failed to verify by the tool. You can then have employees manually verify to see if you get similar results.
e. If there are a lot of reported false positives, consider developing your tool more. You must believe in continuous improvement.
3. What still needs to be done manually?
a. There are many opportunities for a large database to have missing information. Often, we find prospects missing titles, phone numbers, and cities. The information missing from your data will interfere with automated validation. For those who don’t believe in automation, not to fear, we still need a team of manual workers to find some of the missing information.
b. As mentioned above, some attempts at matching for a validation will fail. You will need a team of manual workers to work through the failures and QC your tool.
See also:
Is Your CRM System Keeping You From Connecting with Buyers?
We Believe Sales Is Undergoing a Major Shift In Our Industry
Pulling Back the Curtain: What’s Behind the Scenes of Lead Generation?