Download as pdf or txt
Download as pdf or txt
You are on page 1of 7

International Journal of Nonprofit and Voluntary Sector Marketing Volume 5 Number 3

Deciding to recruit only donors with high


lifetime values

Tony Masters
Zalpha, 1 Riverside, London W6 9WA, UK;
Tel: ⫹44 (0)20 8735 8883; Fax: ⫹44 (0)20 7221 0520; e-mail: tony@zalpha.com
Received: 6th June, 2000

Tony Masters is Senior Strategic Partner at meet fundraising cost-to-income and net
Zalpha, the strategic consultancy of WWAV income targets. However, recent changes
Rapp Collins Group. He has spent 25 years in in data manipulation techniques suggest
direct marketing, starting on the client side in that a different strategy — based on lower
1973, but has worked with WWAV Rapp Collins recruitment volumes but increased donor
for 15 years. He contributes to a wide range of lifetime value — can give a better result
client areas, but has a special interest in for some charities.
charities. He has written articles about a variety It is common practice to increase donor
of fundraising issues, and spoken at national and lifetime values by careful consideration of
international fundraising conferences. communications. Welcome programmes,
donor-care programmes, and fundraising
programmes are all intended to increase
ABSTRACT donor lifetime value. But looking at donor
This paper argues that it is important to choose segmentation suggests that lifetime value is
the right donors to recruit, in order to boost partly nature as well as nurture, and that
average lifetime donation values. This choice choosing the right prospective donors can
needs to be made by making decisions, before have a big, positive effect.
donor recruitment, based on data that are
available in the market. A case history is
shown, based on a group of charities asking for CURRENT SITUATION
cash donations, to demonstrate how such deci- Typically, direct marketing-sourced
sions may be reached. Finally, there is a donors give to a charity for only a few
discussion about implications for direct mail, years. Exactly how long they stay with
other recruitment media, and for charities’ any particular charity will depend on a
fundraising organisations. number of factors, including the donor’s
affinity with the charity and their
donation method (monthly giving, direct
INTRODUCTION debit or standing order, or cash). However
Many charities seem to have a ‘high a charity defines lapse, ultimately very
recruitment, low donor value’ strategy for nearly all donors will lapse.
direct marketing fundraising. This strategy Cash donors have higher lapse rates International Journal of Nonprofit
and Voluntary Sector Marketing,
has been driven by economic reality: it than monthly giving donors. After four Vol. 5 No. 3, 2000, pp. 241–247.
䉷Henry Stewart Publications,
generates sufficient long-term income to years, about half of all new cash donors 1465–4520

Page 241
Deciding to recruit only donors with high lifetime values

have already failed to give in the previous but that is an average, and looking at
year. For donors recruited to a monthly particular cases suggests a very wide spread
gift, it perhaps takes eight years (assuming of results.
99.3 per cent renewal each month) before
half of new donors have ceased giving
regularly. To keep up with this rate of loss, ECONOMIC MODELLING FOR
a charity recruiting mainly cash donors SEGMENTATION APPROACH
needs to recruit new donors equal to
25–30 per cent of its active donor base Regular giving
each year, just to stand still. For charities Take the situation of a charity promot-
using mainly regular giving as their pay- ing a £2 per month ask. Many people
ment method, the number is lower, but respond to this type of offer with more
still substantial: 13 per cent. And most than £2 per month, but there remains a
fundraising charities seek to grow their very substantial group that does respond at
donor bases, not just to stand still. exactly £2. Some of those donors will be
In real numbers a charity with 500,000 persuaded to trade up to a higher regular
active monthly giving donors needs to monthly donation, or to give an oc-
replace about 65,000 donors every year, casional cash gift, or to add value in some
more if it wants to grow. If it had a donor other way. But maybe 25–35 per cent of
base of mainly cash givers, the numbers new donors recruited will never give
would be higher: 125–150,000 every year. more than £2 per month.
These are very large numbers of donors to The economics of this do not look
replace each year. If a third of these targets attractive. The following costs are in con-
are to come from mailings to prospective venient round numbers, but are not too
donors, then at typical response rates a far from reality:
charity would need to mail 2.5 million
people to reach the cash target, or 3.6 — Marketing recruitment costs: promo-
million to reach the regular giving tar- tion at £0.50 per pack, with a 0.6 per
gets. cent response rate gives a cost per new
These days, recruiting new donors in donor of £83.
these volumes is very expensive. Long — Set-up costs: data capture, database
gone are the days when it was possible to amortisation, welcome pack, standing
recruit substantial numbers of new donors order or direct debit set-up costs. Say
at an immediate profit. Now, instead of a total of £10.
immediate profit, many charities using — Monthly administration costs: bank
direct marketing find that only 60–90 per transaction charges, database ad-
cent of recruitment cost is covered by ministration charges, costs of three
first-year income. The remaining 10–40 communications per year. Say a total of
per cent is regarded as an investment £0.25 per month.
against the donor’s lifetime value with the
charity. Today, a charity’s investment in The total costs of winning a new donor,
recruiting new donors is very considerable and the set-up costs — the first two items
indeed. above, totalling £93 — need to be paid
Generally, this investment policy is a from a net £1.75 per month (£2 per
good one. Lifetime values are, on average, month less the monthly charges in the last
sufficiently high to generate income many item above). This means that the £93
multiples of the original recruitment costs, costs are finally recovered with the donor’s

Page 242
Masters

54th monthly payment — four and a half charity raises from them more than offsets
years after recruitment. the money lost on those people who only
give once.
Cash donation
Taking only the donors who pay just £2
per month, their discounted lifetime value NEED TO REDIRECT MARKETING
over the average eight-year life of a SPEND
monthly giving donor is only £113. This For regular monthly giving there is a
figure assumes a discount rate of 8 per group of donors — probably 25–35 per
cent per annum. It seems a very poor cent of donors — who never give more
return on marketing and set-up costs of than they were originally asked for, and
£93. who contribute only a tiny amount of
Looking instead at cash donors, the value to the charity. Similarly, there is a
situation is somewhat worse, because group of cash donors — probably 35–40
there is a group of donors who only ever per cent of donors — who never give a
give once, and no amount of subsequent second gift, and who cost the charity
mailing can encourage them to give again. dearly.
For some charities, this group represents Charities may consider that they want
upwards of 40 per cent of their new cash to divert their fundraising efforts away
donors. from these people, so that more may be
Cash donors have a different, and spent on those donors who will provide a
simpler, cost structure than monthly good lifetime value to the charity. Un-
donors, but costs are still substantial: fortunately, these poor-value donors are
brought in from the same recruitment
— Marketing recruitment costs: promo- activity that produces the large numbers of
tion at £0.45 per pack, with a 1.5 per good, high-value donors: mainly door-
cent response rate gives a cost per new drops, DRTV and direct mail. This ac-
donor of £30. tivity remains viable as long as on average
— Set-up costs: data capture, database donors go on to give more.
amortisation, and welcome pack. Say a If progress is to be made in redirecting
total of £5. marketing spend towards the best donors
— Donor mailings: say eight mailings and away from the poor-value ones, deci-
over two years — after which the sions have to be made at the level of the
donor’s lapse is recognised — at an individual first-time donor, not at the
average cost of £0.62 per pack. So a level of the postal sector, the TV spot or
total of £5. the list. The direct mail process is prob-
ably the one that lends itself most easily to
These add up to a total cost of £40 to consideration of individuals, but findings
recruit and service a donor who gives, from direct mail will have implications for
perhaps, just £15. The charity loses £25 the other major recruitment media.
for each donor that does this. Working with a number of charities, it
Of course, the programme of new has been possible to rank donors from
donor recruitment remains viable, because high lifetime value to low lifetime value.
those donors who do give again, and go This ranking can be forecast with con-
on to give one or two donations a year for siderable accuracy at six months after the
a large number of years, will be very recruitment. The forecasts depend in part
valuable indeed. The net income that the on the initial media of recruitment, but

Page 243
Deciding to recruit only donors with high lifetime values

Table 1: Household structure donors

Household structure Best donors (%) Poor-value donors (%) Index

Single male 11.2 7.7 1.44


Single female 23.7 19.9 1.19
Unmarried couples 4.0 4.5 0.90
Married couples 36.5 41.5 0.88
Older families and similar 14.2 17.5 0.81
Other combinations 10.4 9.0 1.16
Total 100.0 100.0 1.00

mainly on donation amounts and patterns. gift. These charities have general market
These forecasts enable charities to get a appeal and are not restricted to small
focus on the best donors at an early stage, market niches. In every case, less than 60
but the major costs — those associated per cent of first time donors go on to give
with recruitment — are already incurred. a second gift.
So forecasts have limited impact on the Looking at the profiles of best- and
charities’ ratios of donation income to poor-value donors, against standard
fundraising costs. geodemographic and lifestyle profiles,
If ratios of income to cost are to be little difference is found. However, some
changed, then charities need to make progress has recently been made using
decisions about individuals, in terms data from the electoral roll. This work
of their likely long-term value, before indicates that a worthwhile segmentation
recruiting them. This means that the data of prospective donors can be made before
available to use is restricted to what can be mailing them as part of the recruitment
ascertained in advance of promotion: process.
data from lists, from lifestyle and elec- Table 1 takes a sample of the charities’
toral roll overlays, and from standard best donors: those who give repeatedly;
geodemographics. If charities also want and a sample of poor-value donors: those
this decision process to be applicable to a who have given only once, despite a first
wide range of donors, then they also need gift many months ago. These were both
the data to be available as universally as profiled by looking at their household
possible, and also cheaply. Thus the data structure, as derived from the electoral
on which decisions can be based have to roll. Percentages are given against each
be standard geodemographics — for household structure classification, for each
preference, since they are cheapest — or of the two donor groups. Finally, there is
the electoral roll. an index, derived by dividing the percent-
age for best donors by the percentage
for poorest donors. If this index exceeds
CASE HISTORY: A CHARITY ASKING 1.00, then that household structure is
FOR CASH DONATIONS better represented among good donors
The data below come from work that has than among the poorest donors. If it is less
been carried out with a small group of than 1.00, the household structure is less
charities asking for low-level cash dona- well represented among best donors than
tions, some with an incentive for the first among the poorest donors.

Page 244
Masters

Table 2: Length of residence of donors

Length of residence summary Best donors (%) Poor-value donors (%) Index

0–2 years 10.0 9.2 1.08


3–5 years 9.8 11.2 0.87
6–10 years 18.7 21.4 0.87
11⫹ years 61.6 58.2 1.06
Total 100.0 100.0 1.00

Table 3: Combined household composition and length of residence

Household structure Response/Mail index Best/poor-value index

Single male 0.94 1.44


Single female 1.24 1.19
Same sex sharers 1.42 1.16
Unmarried couples 1.03 0.90
Married couples only 0.92 0.88
Older families and similar 0.96 0.81
Other combinations 0.95 1.16

Length of residence summary


0–2 years 1.19 1.08
3–5 years 0.96 0.87
6–10 years 0.99 0.87
11 ⫹ years 0.96 1.06

Table 1 shows, as was expected, that single of recent movers (0–2 years residency) in
people make up a higher proportion of the best donor group was unexpected.
the best donors than they do of the While these segmentations are interest-
poor-value donors. But the degree of ing, they are not sufficient to begin
over-representation of single men was a to make decisions about recruitment
surprise, as was the under-representation strategy; in recruitment, response as well
of older families (recent empty nesters, or as long-term value will be important.
people with an adult child still at home). There was a risk of finding that those
A similar pattern, although less extreme, potential donors with the best long-term
is generated when looking at how long value also had low response rates. Ob-
someone has lived at the same address (see viously, what a charity really needs is for
Table 2). It was no surprise that around 60 high response to be positively correlated
per cent of donors had lived in the same with donor quality.
house for more than ten years. This is an In order to understand the response–
indicator of age, and it correlates with a quality correlation, the study also looked
high proportion of donors being single. at the propensity to respond among the
However, the slight over-representation household composition and length of

Page 245
Deciding to recruit only donors with high lifetime values

Figure 1 Initial Response

Low Medium High

Single men
Single women and
(all residences)
Best same sex sharers
‘Other’ households
donors (<2 or >10yrs resident)
(<2 or >10yrs resident)

Unmarried couples
Single women and
(>10yrs resident)
same sex sharers
Average-value Married couples
(2–10yrs resident)
donors (<2yrs resident)
Unmarried couples
‘Other’ households
(<2yrs resident)
(2–10yrs resident)

Unmarried couples
(2–10yrs resident)
Poor-value Married couples Married couples
Married couples
donors (2–10yrs resident) (<2yrs resident)
(>10yrs resident)
Older families

residence segments. Combining both ele- look attractive. If the lifetime values and
ments into the same segmentation gives recruitment costs for each of the nine cells
the chart shown in Table 3. The table in this matrix are considered, then
indicates two indices. The first, the these charities only make an acceptable
‘Response/Mail index’ is greater than economic return in the four cells labelled
1.00 if response is higher than average in ‘Best donors’, ‘Average-value donors’, or
this household structure or length of ‘High response’. It follows that one of
residence segment, or less than 1.00 if these charities would do better not
response is lower than average. The mailing married couples more than 2 years
‘Best/Poor-value index’ is the one seen resident, unmarried couples 2–10 years
previously, and is greater than 1.00 when resident, and older families no matter how
the segment has a higher representation long they have been resident. These
among the best donors than among the groups are about 50 per cent of the
poor-value donors, and vice versa for charities’ usual mailing file. But mailing
values less than 1.00. only the other half generates real long-
Combining these two indices allows us term value from the remaining half of the
to plot, for this particular charity group- mailing.
ing, those segments that give high,
medium and low response, overlaid with Implications for other charities and
the likely quality of the donors that are other recruitment media
gained. (See Figure 1.) Low response This analysis works for charities asking for
combined with poor-value donors — cash donations, because a donor who
those least likely to give again — does not gives only once actually loses money

Page 246
Masters

for the charity. With charities asking for electoral roll data to be overlaid on
regular monthly gifts, someone who gives mailing files to allow individual deci-
only £2 every month does eventually sions to mail (or not) to be made
recover their costs of promotion: they do — a change in the list-buying process, to
not make a loss for the charity. This might ensure that names can still be bought at
suggest that this type of analysis does not reasonable cost.
work as well for regular-giving charities.
But it can also be argued that it is better There is a bigger change that will be
to replace the monthly donor who only required, and that is a change to charities’
pays their acquisition costs with a donor fundraising targets. Many charities set a
from another list who will have a good ‘cost-per-donor’ recruitment target, based
lifetime value. Charity recruitment results on the actual recruitment income plus an
will be improved by replacing poor-value allowance for additional future giving; and
donors by high-value donors regardless of a volume target. If a charity is to
whether poor-value donors are paying for change to choosing which individuals to
their initial recruitment. mail, judging them by forecast lifetime
The implications for a charity’s recruit- value, then it is essentially changing to
ment mailings are clear, but can the same a ‘higher-value, lower-volume’ strategy.
analysis be used to aid the targeting of the Fundraising targets, in both cost-per-
other major recruitment media: door- donor and volume forecasts, need to
drops and DRTV? DRTV in particular reflect this changed strategy.
generates a different audience and would
need to be analysed separately, but it
seems likely that consideration of in- SUMMARY
dividual lifetime values from these media Many charities that fundraise by di-
will aid targeting either in the choice of rect marketing use a strategy of ‘high
TV spots or in the choice of door-drop recruitment, low donor value’. Economic
areas. modelling, however, suggests that many of
The argument has been made that con- these donors lose money for the charity or
sideration of lifetime values at time of have only a tiny net value. Charities may
recruitment will change the way a charity wish to change from this strategy to one of
uses lists, and could affect the way other lower recruitment volumes, but with high
recruitment media are targeted. Such a donor lifetime value. Changing to this
change will have implications for mailing strategy requires having a decision process
programmes, specifically: that allows a charity to forecast a donor’s
likely lifetime value in advance of recruit-
— a need to carry out fundamental donor ment. The data and technology exist to
analysis prior to a change of recruit- allow a charity to create such a decision
ment policy process, but there are implications for the
— a need to add an additional stage (and charity’s fundraising organisation, as well as
time) into mailing schedules, to allow for the donor recruitment process.

Page 247

You might also like