How’s this different from other analytics tools (GA, Coremetrics, etc.)
How does the SOASTA approach leverage analytic tool data vs data from mpulse (performance)
It’s all about the mPulse DATA – not products/product names
It’s all about the output/results: real-time campaign analytics
All of the data, all of the time, forever:
Data science driven by real math
Precision
Predictive analytics
Session analytics
Setup the story – this is what you’re about to see
This is the customer’s problem: XYZ (no visibility to promotions that weren’t working until the next day)
Tell the campaign story – what happens when something goes wrong? What does the marketer do as a result of this visualization?
Top left : Activity -- Anomaly detection (seasonality within day) – (“Intelligent alerting”??????)
Standard deviation
Simple alerting thresholds don’t work – need deep data science to know how YOUR users use the site
Give an example of when you might go out of the bands: i.e. mis-priced products at $9 instead of $90 and it causes page views to go way up
Events mean different things at different times (midnight-8am, going below 1,000 page views is OK, but not ok during rest of day)
Bottom left: activity by campaign
Top right: cumulative revenue
Bottom right: cumulative revenue by campaign
Tell the campaign story – what happens when something goes wrong? What does the marketer do as a result of this visualization?
This dashboard is geared toward the campaign manager.
Same dashboard as previous, but Isolated to a single campaign
Session path: key points
What is the customer doing on their site?
Abandonments & performance relationship? IN REAL-TIME
OD story: Landing page is slow, but users are still going around it and converting via home page