Yesterday I spoke at the Marketo Marketing Nation Summit on a topic entitled “Build a High Performance, High Velocity, Lead-Generating B2B Machine. The room was packed with 500+ people which surprised me, but made it fun.
“SPelfie” – speaker selfie about 10 minutes before I started.
I uploaded the slide deck here, as some people wanted copies of the graphics.Marketo Presentation- SEMPLE
Despite my best efforts to shock the world with philosophy points like those below……..
- Measure what you can, not what you want
- Use sampling, but low sample sizes drive wrong conclusions
- Dashboards make you dumb
- Google analytics for all!
the only news that got out from the Marketo conference to CNN was that Hillary was “thinking about running in 2016″ from her stage Q&A with CEO Phil Fernandez.
I think “google analytics for all!” is bigger news than Hillary running, but that is just me.
While I haven’t participated in the conference, one of my team members was there and he referenced me to your blog.
These 4 are very good, I agree
– Measure what you can, not what you want
– Use sampling, but low sample sizes drive wrong conclusions
– Dashboards make you dumb
– Google analytics for all!
I think many people don’t understand the value and meaning behind fundamental things (like above) and are looking for some “magic” techniques, while completely missing the point on the basic stuff.
I will click around to check for more good articles 🙂 Thanks!
LikeLike
Question: What sample size you (in Marketo) consider to be reliable ?
– for emails
– for website pages
thanks!
LikeLike
That is a really really good question. For site pages, the A/B testin algorithms will tell you when you reach statistical significance. For emails……hmmmmmm….seems like a good blog posting on how to figure this out (if you can). I generally ignore email results where there are less than 1000+ emails sent. But let me see if I can find a way to calculate this or see if someone else has. Thanks for the comments.
LikeLike