I'm very concerned about the amount of email TrueDelta sends. So I do my best to limit it to the minimum amount necessary to carry out top quality research. But it might not be clear why each email I send is necessary. An explanation might help.
1. Monthly Updates
TrueDelta hasn't always sent out a monthly update. As a result, many people who signed up to participate in the reliability research early on probably forgot all
about it. Also, without visiting the site they had no way of learning about new site features or the progress being made with the research. So TrueDelta
started sending a single monthly update to keep people aware of changes and, above all else, to let them know the site was alive and progressing.
Because these updates only go out once a month, they can be somewhat lengthy. I've considered that multiple shorter emails would probably be more thoroughly read, but feel it's more important not to overload members' inboxes.
2. Survey Emails
When the minimum sample size is reached for a model / model year, enrolled owners start to also receive survey emails with encoded links to a pair of
surveys. I use this email method to avoid the hassle of a sign-in system for the reliability surveys. The Trip to the Shop Survey is filled out the month after any trip to the shop. The Quarterly Odometer Reading Survey (included in the other survey) is filled out the first
month an owner participates, then every January, April, July, and October after that. So a response is only required four times a year.
Why, then, ask about repair trips every month? Because I want people to be able to answer off the top of their heads, without digging up any paperwork. If it's been much more than a month, most people won't be able to reliably and accurately do this.
To minimize the amount of time these monthly emails require, their body never changes and their subject specifies whether a response is required.
3. Follow-up Survey Emails
Follow-up emails are sent to people who did not respond to the initial email. I've been experimenting with the number and timing of follow-up emails. The problem is, the more I send, the higher the response rate gets. Consider the number of responses to the required October survey:
- Initial email: 491 (391 in the first 24 hours)
- First follow-up: 146
- Second follow-up: 121
- Deadline notification: 53
The initial email goes out the first week of the month. The follow-ups go out one and two weeks later, respectively. The deadline notification, first tested in October 2006, goes
out the next-to-the-last day of the month (or the last Thursday of the month if this day falls on a weekend). It's for those people (like me) who wait till the last minute and then need a reminder that it is the last minute.
In each case, response drops to near zero within a couple days of when the email is sent.
To minimize the amount of email, the first follow-up is only sent in months when everyone is required to respond.
To further minimize the amount of email, in December 2006 TrueDelta will stop sending follow-ups in months that don't require a response to those members who consistently respond before the follow-up when a response is required. Based on the October numbers, over half of participating panel members will no longer receive follow-up emails.
Again, I want to clutter everyone's inboxes as little as possible, so if there's a way to reduce the number of emails without harming the quality of
the research results, I'll do it. If you have a suggestion, please send it to me.
Thanks for reading.
Michael Karesh, TrueDelta
First posted: November 30, 2006
Last updated: November 30, 2006