Adding a New Lead Scoring Model Based on Outside Data

Our company is quite unique compared to most other Marketing Automation Companies. We're a manufacturing firm that supplies a building product. The nature of our business leads us to converse as both B2B as well as B2C - the B2C being Homeowners. With a long buying cycle, our B2C marketing works out to be very similar to B2B.

Our homeowners segment was being neglected, we have emails/campaigns based on their buying cycle, but that's about it. We weren't proactively reaching out to homeowners. This was partly due to the fact that we did not have good qualifying model. This problem was deepened by the very large chunk of our web traffic/activity is from Homeowners. The combination of the two had made it difficult for us to reach out to homeowners.

Our basic goal for this project was to start with some kind of lead scoring model, based on outside data that we would be importing in, that will help us proactively reach out. The goal wasn't something major, out of reach, but to be able to have a working model - which we can refine later. This was a project starting from 0, so there was no baseline - the current state was none.

Here were some of the awesome Eloqua University classes that influenced this project - and helped out. And as always, Topliners posts helped greatly!

Eloqua 10: Data Cleansing

Eloqua 10: Lead Scoring

Eloqua 10: System Integration

Project Steps

The use of outside data was to become one of the key scoring entities for a Homeowners explicit score. The challenge is this data was based on the Zip Code of the homeowner, so we would have to build a Lookup Table. Given that this is a Luminary post, I have kept a lot of the steps short - as most of the items were detailed out in the various Eloqua University classes. Should anyone have questions, please feel free to post.

Step 1 - Prep for Import

Prepare the Excel Look up file for import.

Step 2 - Upload Lookup Table/Create Update Rule

1. Before getting to this step, I created new fields, in which the new data would live.
2. Using the knowledge from the awesome Data Cleansing class, import the data for the lookup. Go to Contacts -> Data Tools -> Data Tools Drop Down -> New Lookup Table
3. After entering the data, upload the Lookup Table Entries
4. Once the Lookup Table was ready, I was ready for creating the Update Rule Set.

Step 3 - Build Program/Feeder

1. Once the Update rule was ready - it was time to make the program to run the update rules. A simple program that entered a contact, ran the update rule and pushed them to the CRM to update their new lead score:
   
   
2. The feeder was a simple one - once the initial run was set up, we created a shared filter of all homeowners that didn't have the field populated yet - and made sure they had a ZIP filled out
   
   

Step 4 - Lead Score Model

1. Now that we had the fields populated, after waiting a night for the program to run - we were ready to add this to our lead scoring model for Homeowners

Results

We are seeing a much better distribution of homeowners in terms of the explicit scores after adding these additional criteria in the scoring models. Our next step is to create a better way to push over the score to our Salesforce - as we have multiple models already being pushed. Prior, our homeowners distribution had a large number of D's and C's - but now we are seeing many more A's and B's. As soon as we refine the Lead Score Model and push it to SFDC we will work with our LDR team to proactively reach out to these leads and continue to refine the model.

The entire program did take some additional configuration, as we had a requirement to have the new fields to be numeric, so that we may set intervals to score off of. This caused a challenge, as you cannot overwrite value from Lookup Table Field if the field to be updated is a numeric field:

Our workaround to this challenge was to create 2 versions of the field, a text and numeric. We created our update rules to update the text field -> which pushed to a numeric SFDC field -> Eloqua then pulled the new SFDC field into the numeric field, which we were able to then properly score off of in intervals.

We hope to show some solid results on how this helped our LDR team to create new business via proactive calls to homeowners!