Recently, I had the chance to build an awesome predictive model for a client who wants to showcase the expected inquiries and new business their client can expect to see before actual data comes in. It’s kind of like peeking at a corner of a wrapped present at Christmas for us researchers.
We built the predictive model with the following components:
1. Field Staff Data
Impeccable Field Staff Data – our client has a great reporting system that makes it easy to access attendance, interactions, and leads data (their key metric which is captured electronically).
Also, a solid exit survey that has four components:
- Experience Metric – What is the consumer’s past experience with the brand? Are they a current customer or to they currently have loyalty to a competitor? This may come into play with the value that you assign an inquiry, new business, lead, etc.
- Qualification Metric – Are interactions even in the market to shop for the product?
- Action Metric – Are they going to contact the brand they just had an experience with?
- ‘Authority’ Metric – Are they the person who can make the decisions? This is important because you can talk until the cows come home, but if you don’t talk to the person with the power, you are talking to yourself.
2. Post-Event Data
A post-event survey that measures the actual action taken as some point in the future.
This is incredibly important because it gives you a filter to predict how many of your “super excited and engaged consumers who are ready to take on the world after your experience,” actually decide to inquire on your brand.
After a couple of edits here and there, the client and I were happy with the product. Now comes the wait… how does our predictive model match up to ACTUAL behavior?!
I will be sure to update and let you know because there is NOTHING better than a predictive model matching up to actual consumer behavior.
How do I know? Because I’ve done it for the same client and a different brand. We were on high horses that day. Guess they had such a great experience with us they decided to give it another go.
Need a predictive model? Get in touch with us.
So far we have a pretty good track record with them because we put A LOT of method to our madness (and it shows).
Photo Source: https://www.flickr.com/photos/smudge9000/