Only 40% of Email Addresses can be Validated so ask all these big data providers if they model their email addresses or if they come from a verified source. Meaning do they guess at the naming convention based on LinkedIn profiles / email domains then use SMTP negotiation for validation.
This approach has many challenges – first of which is that only 40% of US based companies validate email addresses. The other 60% just validate if the domain is accurate.
Take my name for example. I go by M. Scott Miller on LinkedIn so how would someone fabricate my email address? Would I go by M or Scott or Mscott? Take my wife Liz as an example – what do they do if an Elizabeth on LinkedIn goes by Liz for her company email. The most egregious however is the larger companies (10,000+ employees) that have many common names like Scott and Elizabeth, so they use a middle initial or numbers to provide unique email addresses. Rarely do these large companies provide email validation– they just accept all email addresses and don’t give you any other information otherwise.
If your current vendor uses email modeling then you are only getting 40% of a database – the other 60% is a guess. Add to the fact that B2B data expires at a 3% rate per month then that guess becomes pretty risky as it ages. Using LinkedIn alone is the biggest hurdle because rarely do LinkedIn users update their profile if they didn’t leave their employer on their own terms. Unemployed job seekers don’t tell the world they are unemployed – they wait to update their profile on LinkedIn once they have a job.
Do yourself a favor – ask your contact vendor these three questions
- Where does your data originate? If it is the web alone then you are at risk of soliciting the unemployed or missing out on executives that don’t want the attention.
- How often do you update your data? B2B data expires at a 3% rate per month. If they respond every 3 months then just add 18% inaccuracy from the start.
- Do you model your email addresses? Only 40% of modeled email addresses can be validated so 60% of their data is guess at best.