Measure Up!
This type of analytics answers the question: "What happened?"
descriptive analytics
KPI stands for these three words:
Key Performance Indicator
Our presentation says each KPI should have one of these, not a committee.
Owner
The first step in the optimization framework is to measure the current state, called this.
Baseline
This gives the team a real-time view of all KPIs.
Dashboard
This type of analytics looks for root causes and patterns behind results.
diagnostic analytics
This five-letter framework is used to create strong KPIs.
SMART
Weekly or monthly reviews of KPI progress are called this.
Check-in cadence
The second step asks what could break your strategy.
Stress test
The lecture says metrics should be broken down by product, channel, and customer ______.
Segment
This type of analytics tries to forecast what will happen next.
predictive analytics
In SMART, this letter asks whether the team can actually influence the metric.
A for Actionable
Apple is used as an example of a company organized by this instead of by product.
function/functional teams
The lecture says to make this many changes at a time when optimizing.
ONE.
Besides a dashboard, each KPI also needs one responsible person called this.
Owner
Before you can improve or optimize something, the lecture says you first have to do this to it.
measure (you can't optimize what you can't measure)
“Increase website traffic” is used in the lecture as an example of this kind of weak metric.
a BAD KPI/vanity metric
Clear roles create clear ______ for KPIs.
accountability
After implementing a change, you do this to track its impact on KPIs.
Monitor
This is the pre-decided rule for what happens if a KPI drops by a certain percent.
Action threshold
If a team studies past traffic, figures out why signups changed, and then forecasts next month’s results, they are using these three levels in order.
descriptive, diagnostic, then predictive analytics
A good KPI should answer this question: “If this number goes up, are we ______?”
WINNING
In the presentation, when everyone is partially responsible for a KPI and it drops, this usually happens.
EVERYBODY BLAMES EVERYBODY ELSE
In the CAC example from the lecture, the cost dropped from $50 to this number after the test worked.
$35
In the lecture example, if CAC rises more than 15% in a month, the team should automatically do this.
Shift budget to the highest-performing channels