Upload
fiona-mcphee
View
186
Download
0
Embed Size (px)
Citation preview
Practical ways to use your data to drive up
results
Exploring SegmentationFiona McPheePareto Fundraising
Segmentation as your starting point
Targeting, testing & analysis as the goal
I’ve had to make a few about what you already know and do so we have time for the good bits
Assumptions
Bothering has a discernable benefit You have buy in, why is understood and it’s used consistently so it’s not:
We Agree That You Have To Bother With Segmentation
Underutilised Tossed Aside Disliked by users
Communicating with everyone, every time is not effective
fundraising• Not everyone is the same• Not everyone wants the same thing
We agree that:
Understand how your supporters
behave
Make informed decisions
We agree that implementing some form of segmentation will benefit your fundraising program by
primarily helping you to:
We acknowledge segments of One is the goal?Tailored to
the individual
Right ask, right time,
right channel
Best ROI possible
(short-term / long-term)
Acknowledge segments of One groups is the reality?
Major Donor Program
Middle Donor ProgramCor
e Program
Gran
ular
ity o
f seg
men
tatio
n
Use Segmentation To Consistently Target & Measure
Who What When Outcome Inform next move
Use Segmentation To Consistently Measure
•Increase gross / net income•Increase response rates•Increase average gift
Program / Strategy
•Measure / Improve effectiveness of every dollar you spend – ROI•Measure / Improve revenue generating efficiency of every dollar you spend - CPRCampaign
•Increase Life-Time-Value•Increase retention / Decrease attrition•Increase gifts per donor•Increase multi-interaction (cross-sell)
Donor Base / Donor Type
Use to create fairly homogenous
groups which are likely to respond in similar ways to our
strategies
Use to prioritise these groups in
order to target and understand their
behaviour
I used segmentation primarily by
Recency, Frequency, Value (RFV). RFV is a tactic – there are
other options
My starting point for segmentation
Psychographic Segmentation• Powerful potential – the goal of many Marketing &
Fundraising teams• Often revealed by qualitative research
• e.g. personal experience of services offered by charity• Data very hard (expensive) to capture and maintain at an individual
donor level (and nearly always an additional cost to standard database stored information such as transactions)
• Psychology is more complex than a data point or two (e.g. people change their minds) and so even psychologists find it hard to determine practical marketing psychographic segments that can be effectively used for targeting.
• May be easier to use psychographic segments to inform creative treatment only
Demographic Segmentation• Social and economic information is fairly
commonly available data (internally or externally) and therefore somewhat practical
• Age can be a useful additional overlay• Widowers with absence of grandchildren
correlates strongly with bequest realisation• Not a reliable predictor of behaviour - no
obvious practical value for a targeted communications programme
• Helps us to know who is in our segments
Geographic Segmentation• Basic and obvious insights
• Affluent areas/ higher donor penetration areas• Proximity to Charity service areas if relevant
(e.g. a Hospital)
• No obvious practical value for a targeted warm communications programme
• Can be helpful for acquisition targeting• Might be useful for local level prospect
targeting/ broadcast media/ door-drops
Behavioural Segmentation• The product of psychographics and
geodemographics
• A charity’s (your) supporters are already a highly niche, homogenous market segment (why swaps work)
• RFV is the most common behavioural segmentation • Adapting for fundraising & your program / donors
and your activity is key
Common Fundraising RFV
Recency Frequency Value
0-12 months >1 a$1000+
13-24 1 b$500-$999.99
25-36 c$250-$499.99
37-48 d$100-$249.99
49-60 e$50-$99.99
61-72 f$25-$49.99
73-84 g$10-$24.99
85+ h$0.01-$9.99
Likelihood of transacting again
Potential value of the
transaction
Donor Type
RFV
Tailoring
Cash Donor Lottery In Memoriam Other Event 0%
2%
4%
6%
8%
10%
12%
14%
16%
Response Rate by Supporter Type Re
spon
se R
ate
Group lower
volume for
ease
Regular Givers Cash Donors Campaigners Lottery
Players
Event Attendees
In Memoriam
In Celebration
Peer-to-Peer Fundraisers
Peer-to-Peer Sponsors Members Virtual Gift
Purchasers
Regular Givers
Regular Givers Only
Regular Givers & Cash
Donors
Regular Givers &
Members
Member & Cash
Donor
Cash Donors
Members Only
Virtual Gift Purchaser Other
Lottery Players
Event Attendees
In Memoriam
In Celebration
Cash Appeal
Cash Donors
Members Only
Member & Cash Donor
Virtual Gift Purchaser Other
Lottery Players
Regular Giving Conversion
Practical Applications To Improve Results
Use segmentation to manage tactics and testing to increase response rates and net income
The question “How much of our warm donor base is optimum to mail (or phone, offer lottery)?” is often asked of a data set. The best way to answer this is via a segmentation model which will help identify not only those donors that will respond to appeals, but the value of the responses.
0-12 13-24 25-36 37-48 49-60 61-72 73-84 85+0%
2%
4%
6%
8%
10%
12%
14%
16%
18%
20%
Typical effect of Recency on Response
Recency (months since last gift)
Resp
onse
More recent supporters respond better than those donors who haven’t transacted for some time.
1 4 7 10 13 16 19 22 25 28 31 34 37 40 43 46 49 52 55 580%
5%
10%
15%
20%
25%
30%
Typical effect of Recency on Response
Recency (months since last gift)
Resp
onse
Those who did it recently are the most likely to do it again
The purest way of considering frequency in a RFV model is to look at the giving interval.
<3 3-5 6-8 9-11 12-14 15-17 18-20 21-23 24-26 27-29 30-32 33-36 >360%
5%
10%
15%
20%
25%
30%
35%
Typical effect of giving interval on response - Multi Donors
Cash giving interval
Resp
onse
Rat
e
Split by 12 month intervals – highly predictive, but more so for 0-12 donors than the 13+ that may be unprofitable
<12 12-23 24-36 >360%
2%
4%
6%
8%
10%
12%
14%
16%
18%
20%
Typical effect of giving interval on response - Multi Donors
Cash giving interval
Resp
onse
Rat
e
Resting (not giving a chance to do it again)
will affect response in short-term
An alternative, and simpler to implement, way of looking at frequency is simply to look at the number of gifts a donor has made.
1 2 3 4 5 6 7 8 9 10+0%
5%
10%
15%
20%
25%
Typical Effect of frequency on response 0-12 recency
Cash Gifts
Resp
onse
Rat
e
Jump in response once
a donor has made 3 or more
gifts
When we look at the 13+ recency donors, the jump in responsiveness comes at the 2nd gift
1 2 3 4 5 6 7 8 9 10+0%
1%
2%
3%
4%
5%
6%
7%
Typical effect of frequency on response13+ Recency
Cash Gifts
Resp
onse
Rat
e
The more gifts given in the
past the more likely they are to give, useful
for reactivation
In response to the original question, who are profitable to mail, then this becomes a question of focusing on the lapsed donors to eliminate those that are not profitable. For practical reasons, splitting by single and multi donors provides good distinction in for these lapsed donors.
1 >10%
2%
4%
6%
8%
10%
12%
14%
16%
Typical Effect of frequency on response
Cash Gifts
Resp
onse
Rat
e
“How much of our warm donor base is optimum to mail?” • Those who gave more recently• Giving interval /number of gifts predictive of
response– More gifts = better (for both recent and lapsing/lapsed)
Summary
When we look at the value, we have several options to consider. Total giving, Value of last gift, Average gift or Highest Gift. The total giving is usually excluded because it effectively combines the frequency and value.
$0$30$60
Typical effect of Value on IncomeHighest Gift
Value (Highest Gift)
Inco
me
per d
onor
mai
led
$0$40$80
Typical effect of Value on IncomeLast Gift
Value (Last Gift)
Inco
me
per d
onor
mai
led
$0$30$60
Typical effect of Value on IncomeAverage Gift
Value (Average Gift)
Inco
me
per d
onor
mai
led
PotentialActual
© Pareto Fundraising 2013
TargetingExample Charity
© Pareto Fundraising 2013
Targeting
Donor type
Example Charity
© Pareto Fundraising 2013
Targeting
RFV
Example Charity
© Pareto Fundraising 2013
Targeting
Strategy
Example Charity
© Pareto Fundraising 2013
Targeting
Counts
Example Charity
© Pareto Fundraising 2013
Targeting
Strategy
Example Charity
Active Regular
Giver
F2F
Recency of recruitment
Upgrader (Y/N)
Phone
Recency of recruitment
Upgrader (Y/N)
Regular giving example
• System (Database)• Expertise (staff expertise)• Capacity (staff resource)• Size of program (what level of granularity is
worth it)
Behavioural / RFV Challenges
SCORING
Another application of segmentation analysis is to take the significant variables, and create a scoring model that combines response and value predictors.
• Good for the phone (lottery, RG, cash, BQ) as the fixed costs are limited but the marginal costs are high / cost per contact is high therefore you need a more precise view of the return per donor– With the mail for example the marginal costs are low and hard to calculate, lot more
fixed costs– Requires significant variables (impacting response/desired outcome) to be identified
(and validated and refined over time) which are determined by initial analysis which should be refreshed and assessed for accuracy/changes over time
– Example of use: higher scored donors would be called more frequently• Mail programs tend not have the frequency fluctuations but say 12 x per year your mail saings may be
good
What about scoring?
$0.00
$10.00
$20.00
$30.00
$40.00
$50.00
$60.00
0%
10%
20%
30%
40%
50%
60%
70%
80%
Response & Income by score
Actual Resp Income per donor
RFV Bequest Prospect
Bequest Scoring
Demographic indicators
Bequest Prospect
Donor Attribute Score
Loyal Donor (RFV driven)
Number of gifts (e.g. 10+, 5+)
Length of giving (e.g. 10+ years)
Donor Type (e.g. RG & Cash, Cash Only)
Gift value band (RFV driven)
Payment type(s)
Age (</>)
Has children / grandchildren
Proximity to cause
Responded to survey
Changed details
RFV Scoring Overalys
Bequest Score
Bequest Scoring
SEGMENTATION WITHIN SEGMENTATION
Segmentation within Segmentation
• Analysis to identify your value band / profile
• E.g. $1k - $5k active multi, new $1k - $5k
Middle Donor Progra
m
Some form of direct response Age Longevity
Upgrade Other channel usage e.g. SMS
Response to a declined payment
communication
Mailing F2F Recruits – What overlays may indicate potential?
OTHER IDEAS
Topic of Acquisition
Premium Type / Value
Premium vs. non-Premium
recruits
Recruited via a swap
Have been swapped
Subjective list then analyse
Other overlays to consider
Number of touch points
2nd Gift rate by contact type segments & recruitment type
0
200
400
600
800
1,000
1,200
1,400
0%
10%
20%
30%
40%
50%
60%
70%
80%
Addr
1
Addr
1 Em
ail
Addr
1 Em
ail L
andl
ine
Addr
1 Em
ail L
andl
ine
Mob
ile
Addr
1 Em
ail M
obile
Addr
1 La
ndlin
e
Emai
l
Addr
1
Addr
1 Em
ail
Addr
1 Em
ail L
andl
ine
Addr
1 Em
ail M
obile
Addr
1 La
ndlin
e
Addr
1
Addr
1 Em
ail
Addr
1 Em
ail L
andl
ine
Addr
1 Em
ail L
andl
ine
Mob
ile
Addr
1 Em
ail M
obile
Addr
1 La
ndlin
e
Addr
1 M
obile
Emai
l
Emai
l Mob
ile
Addr
1
Addr
1 Em
ail
Addr
1 Em
ail L
andl
ine
Addr
1 Em
ail L
andl
ine
Mob
ile
Addr
1 Em
ail M
obile
Addr
1 La
ndlin
e
Addr
1 M
obile
Emai
l
Cash Recruit Virtual Gift Recruit Emergency Recruit Disaster Recruit
2nd Gift Rate (Year 1) RecruitsBase: 2009 - 2014 Recruits
The more info provided = better second gift rate
2nd Gift rate by simplified contact type
0
500
1,000
1,500
2,000
2,500
3,000
3,500
4,000
0%
5%
10%
15%
20%
25%
30%
35%
40%
45%
50%
Addr
1
Addr
1+
Emai
l
Emai
l Lan
dlin
e
Emai
l Mob
ile
Land
line
Mob
ile
Addr
1
Addr
1+
Emai
l
Emai
l Lan
dlin
e
Emai
l Lan
dlin
e M
obile
Emai
l Mob
ile
Land
line
Mob
ile
Addr
1
Addr
1+
Emai
l
Emai
l Lan
dlin
e
Emai
l Mob
ile
Land
line
Land
line
Mob
ile
Mob
ile
Cash Recruit Emergency Recruit Disaster Recruit
2nd Gift Rate (Year 1) RecruitsBase: 2009 - 2014 Recruits
Simplified view of the contact segments shows us that Address plus something
else or just address are best 2nd gift opportunity.
Donor Base / Donor Type Tracking
1 2 3 4 5 $-
$200,000
$400,000
$600,000
$800,000
$1,000,000
$1,200,000
$1,400,000
-
0.50
1.00
1.50
2.00
2.50
3.00
3.50
4.00
Net Income All & ROI All
Series1 Series2
1 2 3 4 5 $-
$50,000 $100,000 $150,000 $200,000 $250,000
- 0.50 1.00 1.50 2.00 2.50 3.00
Net Income Newsletters & ROI Newslet -ters
Series1 Series2
1 2 3 4 5 $-
$200,000
$400,000
$600,000
$800,000
$1,000,000
$1,200,000
- 0.50 1.00 1.50 2.00 2.50 3.00 3.50 4.00 4.50 5.00
Net Income Appeals & ROI Appeals
Series1 Series2
Donor Type Tracking
Donor satisfaction with the quality of service provided by the fundraising team is the single biggest driver of loyalty towards the organisation
Satisfaction Levels
Commitment Level
Motivational Segmentation
Right thing to do
Environmental issue
Rights issue Its all three
Advocacy
Education
Basic Needs
FIONA [email protected] TWITTER: FIMCPHEEPHONE: +6421 336 905WEB: WWW.PARETOFUNDRAISING.COM