How to stop bad data entering your CRM system

There are 3 primary ways the data in your CRM degrades in quality

Marketing enters incorrect, incomplete, or denormalized data, or corrupts good data
Sales people enter in duplicate, invalid, or incomplete data
The data just naturally decay in quality over time

The first source listed above is now largely fixable and is the subject of this post. The second is controllable with reports that find the issues, and management support to cajole the sales people into better behaviors. The third is fixable with regular data appends and cleanup.

There is no excuse anymore for marketing in the entering of bad or incomplete data. Marketing Automation Systems have the capability to validate and standardize data, rejecting data that does not measure up. So what are the functions of a good response management funnel as it relates to maintaining data quality standards? If you have simple forms that use a web2lead like interface to jam data straight into your CRM, then you have a highway to data decay.

Here is a 90 second video on the Market2Lead Response Management process. Also here is a good second video, 3.5 minutes in length, on all the data validation and rejection features for keeping your data clean.

To maximize data quality, and richness over time, you will want your on-form capabilities to include:

The combination of City, State, and Postal code is validated
Email format validation (has to have a “@” and a “.” for instance)
Phone number format validation
Postal code format validation
Rejection of common junk (asdf, xxx, abc, qwert, etc) in names
Rejection of public domain emails for high value business to business offers

Once the form is accepted and submitted, there is still plenty to do. You are likely to de-duplicate with your database on email or some combination of name and company name and components of the address. Only then do you send them to the fulfillment page (or confirmation page).

You still have more processing on the response however. At Market2Lead we do another round of rejection, this time against a list of 600+ expletives and other junk words people enter into the name fields. The interesting part here is that this record is still sent into the database, but with a status of “Rejected”. The reason you want to send “rejected” responses to the database is that it allows the marketer to identify common sources of junk submissions and also check for “false positives”. The rejected submissions are easily deleted at any time because of the “rejected” status.

The next step is augmentation. This is a fantastic idea, and highly undervalued. You can use it to do all sorts of stuff like assigning sales territories to leads, or converting state abbreviations into full length state names. You could create a flag to indicate if the person used a public domain in their email address, or automatically tag registrants as competitors based on their email domain. You can recognize new leads from existing customers and flag them so they are sent to the account owner without delay.

After all the data rejection, augmentation, and suppression is done, then it is time to score the leads, and decide where to route them. With strict data standards, documented process, and automation to prevent bad data from getting into the system, your marketing campaigns will be more successful, and your sales people will be more appreciative of the data they get from marketing.

-Kevin