Analytics

The key to maximizing value and sustained growth is getting the right
information to the right people at the right time

ANALYTICS
 What to know
Data Sources
Types of Analytics
Campaign Analytics
Forecasting Analytics
Monitoring Outcomes
Old Model
New Model need a plan to integrate

DATA SOURCES
 Sources: The Old Model

TM Web Sites

Donor/
Customer

Direct Events/Major
Mail Gifts
You Need a Plan that Integrates the Data
Integrating Data to Develop Strategy
Balancing Analytical Needs
Description of Types of Analytics

TYPES OF ANALYTICS
 Analytics Balancing LCLS
There are basically 4 types of data that need
to constantly be balanced in decision making
each has different value
L Lagging Benchmarking
C Current Campaign Reports/Testing
L Leading Forecasting/Budgeting
S Strategic Scenario Building/Multi-Year
Growth
 Analytics a Few Types
Campaign Analytics
Performance Reports
Package Testing Results
Forecasting Analytics
Budget Forecasting
Cash Flow Forecasting
Analytics a Few Types
Digital Media Analytics
Web Site Traffic Analysis
Paid Advertising ROI
Communication
Analytics
Segmentation Analytics
Predictive Analytics
 Analytics a Few Types
Performance Indicator Reports
Trend Analysis (historic donor value)
Multi-Year Forecasting (future donor value)
Performance Reports
Metrics

CAMPAIGN ANALYTICS
 Performance Reports
Key metrics help determine campaign success
What are they?
How are they calculated?
Why is it important? Why may it be limiting?
Which is most important when?
It depends on your goals!
It depends on the program!
 Response Rate
What?
The percentage of the quantity mailed that
responded
How?
(# of Gifts/Quantity)*1000
Why important?
Indicator of package success. More gifts the better!
Why limiting?
Does not factor in revenue, size makes a difference.
 Average Gift
What?
The average size of the contribution
How?
Gross Income/# of Gifts
Why important?
The higher the gift, the more valuable the donor
Why limiting?
Does not factor in response, large gifts can skew
performance
 Gross/M
What?
Gross income per thousand pieces mailed
How?
(Gross Income/Quantity Mailed) * 1000
Why important?
The more revenue per piece, the better
Why limiting?
Does not factor in cost
 Cost/M
What?
Cost per thousand pieces mailed
How?
(Total Cost/Quantity Mailed) * 1000
Why important?
Tracks costs
Why limiting?
Test vs. rollout pricing need to be considered
 Investment/Donor (CTA)
What?
Net Investment to acquire a new donor
How?
Net income/# of Gifts
Why important?
Factors in response AND revenue AND cost
Why limiting?
Only used in prospecting, but its a key indicator!
Acquisition Campaign Report
 Net/M
What?
Net income per thousand pieces mailed
How?
Gross/M Cost/M
(Net Income/Quantity Mailed) * 1000
Why important?
Factors in gross and cost, measures net productivity
Why limiting?
Not really applicable in prospecting when there is
traditionally no net revenue
Appeal Campaign Report
Which variable is most important when? It
depends, but general rule of thumb is
Acquisition Investment/Donor (CTA)
Appeals Net/M
Renewals Response Rate
Upgrades Average Gift
 Package Testing Why?
Build on whats working, change what isnt
Tweak Tests, New Package
Move the needle on a specific variable based
on program goals
Response rate, Average gift, Net (etc.)
Stay one step ahead of your peers!
 Package Testing When?
As often as possible!
Try and test at least one variable in every mailing
Tests arent limited to creative/design
RFM segment testing in appeals
List testing in prospecting (80/20 rule)
 Package Testing How?
Ensure there is enough quantity (and gifts) to
be statistically significant
Test only one variable at a time
Make sure segments are equally distributed
Ensure you wait long enough to make a
conclusion!
 Package Testing For Statisticians
Confidence Intervals
Bell Curves
Z Score
P Value
 Bell Curve

CONFIDENCE INTERVAL

Lower Actual Upper

 Bell Curves Intersecting
Package A Package B

Overlap?

Lower (A) Upper (A) Lower (B) Upper (A)

Acquisition Testing Report
Appeal Testing Report
 Package Testing For Everyone Else
9 times out of 10 the result will be the same
Lower and Upper Bounds
Best Case/Worst Case Scenarios
Overlap
Which metrics to consider depend on goals of
the test
Budget Forecasting
Cash flow Forecasting
Multi-year Forecasting

FORECASTING ANALYTICS
 Budget Analysis
Question: Based on where I am today in my
fiscal year cash flow and how campaigns have
been performing, where am I going to end the
fiscal year.

Approach: Combining current cash flow to

date along with a forecast of active
campaigns gives an excellent real-time
approximation of total fiscal year revenue.
34
 Budget Analysis: Tools
Campaign Budgets: All TM, DM, Web, etc.
budgets with expected outcomes.
Predictive Flow: using different tools for each
channel flow out budget (i.e., DM would use a
standard days out).
Estimation of Carry In/Carry Out: Based on
timing of prior FY campaigns how much
revenue will be carried into this FY. Same for
current FY Campaigns.
35
Budget Analysis: Results

36
 Forecast Analysis
Question: Recent performance metrics are soft
compared to last year, how is recent
performance going to affect my fiscal year
budget.

Approach: Look at recent cash flow trends

based on which campaigns are active and use
tools to forecast the remaining months.

37
 Forecast Analysis: Tools
Historical performance by day.
Time-series analysis to utilize recent trends
along with historical performance to predict
outcome.
Create several forecasts based on assumptions
of environment changes.

38
Forecast Analysis - Results

39
 Multi-Year Forecasting
Question: Determine the baseline scenario of
business models based on declining performance.
Analyze what investment levels are realistic given current
trends in each business model

Approach: Establish benchmarks for growth in each

program areas including investment levels, return on
investment targets, donor return rates, retention
targets, donor conversion, and upgrading targets, etc.
Identify the donor segments of greatest value to guide
investment decisions.

40
 Multi-Year Forecasting - Tools
Segment the file into the investment channels
(i.e., DM acquisition, Web Ads, F2F, etc.)
Estimate performance of the new acquisition
channels based on the investment.
Use historical activation, retention and
upgrading measures to predict future donor
value.
Create scenarios based on modifications of
investments or changes in the base file.

41
Multi-Year Forecasting - Results

42