Using the Right Metrics for the Right Job
Luke Glasner, Principal, Glasner Consulting
Adam Holden-Bache, CEO, Mass Transmit
Ryan Hofmann, Sr. Consultant, Responsys
Dela Quist, CEO, Alchemy Worx
- EEC S.A.M.E. (Support Adoption of Metrics for Email) Project
- Accepted (instead of "Delivered") - any email not rejected (same metric, better name)
- Render (instead of "Open") - number of times HTML was displayed (with images on)
- Inbox Placement Rate
- Total Confirmed Opens = Render + Counts from Clicks (to capture any text or image-off opens)
- Need to map the lifecycle of your customer, with regards to email
- Subscribers are most engaged in the first 90-days
- "Emotionally Unengaged" = unresponsive
- Are people receiving your email and then doing something online? (Even if they don't open/click?) Third time I'd heard that in the same day in different presentations.
- Lifecycle = opt-in date through last engagement (click, open or non-email, like signin)
- Find mean and standard deviation
- 95% of database within range (mean + 2 standard deviations)
- If this range is too large, we need to look at confidence ( a |---*---| b )
- (a) is when you've sent email before they've become unengaged
- (b) is when 95% of responders are lost forever
- What kind of lift do you get from being in the inbox even if they don't open? (concentric circles - last click, email clickers, email openers, email subscribers)
- From Gilt (earlier): when the template was optimized for mobile, mobile conversions went up. But so did desktop conversations. (If they could understand the email on mobile, they acted on mobile, or saved to later access on the desktop. But, if they couldn't understand it on mobile, they just deleted it.)
- More engaged (first responders) - less valuable. They open everything. In this sender's experience, 25% of revenue was consistently found in emails more then 21 days old. The emails were not working in a vacuum, but influencing each other. 50% of revenue came from emails over a week old.
- Suggests forgetting open/click rate and thinking about open/click reach (that until you've told them not to send, you should be sending and working on improving open/click)
- You want to grow this list? Great! How much? Uh...
- Benchmark: 3% of monthly unique visitors to your website should be signing up for your newsletter.
- Expecting opens and clicks from every email by every reader: If you sell refrigerators and they open every email, they are useless. But, if they buy a refrigerator, it was all worth-while. Can't treat each message as its own little thing like a message in a bottle tossed into the ocean.
- If it keeps getting successfully delivered but not opened, it's still branding reinforcement and top-of-mind.
- Most people will only sign-up for one vertical's email - keeping them on your list could prevent them from signing up for your competitor's email.
- Want to show the value of email to your organization? Stop sending any email for a year. Caution: You (and the rest of the company) may be unemployed.
- Don't ask anyone to sell on your behalf. If it does, great. But you should do your job yourself. Social Media is great for letting your fans talk about how great you are, but they will never sell enough to be valuable. So give them good material to work with, don't try to sell to them in the social space.
- If you don't have a preference center, get one. (Five or fewer total checkboxes.)
- Use progressive profiling to learn more over time.
- Map out the lifecycle, use it to figure out when people are dropping off, when it's too late, and how to adjust the lifecycle. (see math above)
- Make sure your definition of engagement is multi-channel. Go beyond opens/clicks/direct revenue.
- Give them good content to share socially. And make it easy to share.
- Consider the totality of your email program and how emails work with one another.
- Make sure your email is mobile-friendly. Even if it's complex and needs to be acted upon on a larger screen, make sure it doesn't get deleted before they get to the larger screen.