TellMyCell.com Launches SMS Drip Campaigns

I classify my email messaging programs into 3 categories:

promotional blasts (1 message out to a whole list or portion of a list)
transactional messages (administrative messages whose primary purpose is not promotional)
drip campaigns (automated messages that are sent repeatedly to specific cohorts as they meet the qualifying criteria)

Super-easy-to-use bulk SMS supplier TellMyCell.com is currently beta testing drip campaigns for users of their text messaging services.

Currently, the configuration options are limited, with triggers based off of time passed since join date.  I've got my fingers crossed that they'll introduce greater control and more options over the triggers.

In addition to drip campaigns, the service offers both shared short-code and web-form list-building options, as well as shared short codes.  You can quickly set up an account with them and test the service for free at www.tellmycell.com.

How Do You Know When You've Lost a Customer?

Your customers don't typically bid you goodbye before defecting to a competitor.

 How do you know when they're gone? Is there a way to predict that they might be headed for the door in time to win them back? 

We worked on developing an approach based on a Poisson distribution.

 Basically the model looks at your historical customer data, and maps out: based on the number of days since my prospect's last visit or purchase, what is the likelihood that they will ever return?

 We looked at several years of data. We mapped out, of all the people who visited on any single day, how many returned the next day? Of the people whose last visit was 1 day ago, how many returned the next day, and so on.

 As you might expect, the longer it had been since the last time our prospects had been to the site, the less likely that they would return the next day.

 Based on this model we were able to determine a target date when our customers passed a critical threshold -- they were more likely NEVER to return than they were to return. Not a perfect model, of course, but a line in the sand that we could use to begin optimizing our way forward.

 We used it to develop a customer winback program targeted only to those customers who had reached the critical threshold -- more than 50% chance that they were gone for good. Preliminary results showed a 30% lift in customer return rates.

 We worked to lock in these gains by setting up an automated communication program that automatically messaged customers as soon as they hit the threshold and set it on autopilot.

 Next step -- we'll split testing the timing to see if we can improve on those 30% gains

 How does your company approach the challenge of figuring out when a customer has moved on to greener pastures?