Introduction

Frequency

Frequency Distribution

Give Again rates

Giving Patterns

Value

What about converting to Regular Giving?

Ranking

Key takeaways

The data

Introduction

What can we learn from a million single givers?

While it’s widely recognized that focusing your recruitment strategy on regular givers typically yields greater long-term rewards, many charities still prominently feature single giver recruitment in their acquisition strategies.

In conversations with those clients who rely significantly on recruiting single cash givers, we’ve heard various thoughtful rationales: the often prohibitive cost of recruiting regular givers, the importance of offering options for supporters who prefer one-time donations, the interesting theory that single givers might actually develop stronger emotional connections with your cause (since they actively choose to donate each time), and perhaps most commonly, that it’s part of a strategic two-step approach where single givers are eventually encouraged to become regular givers or at least continue making occasional one-off donations.

During these discussions, we’ve noticed a subtle but meaningful difference in how organisations perceive regular versus single givers. While most charities seem to believe regular giver attrition follows fairly predictable patterns across organizations, they view single giver performance as much more diverse and organization-specific. There’s a prevailing sense that single giving patterns are uniquely shaped by each charity’s approach rather than following some universal “law of giving”!

Double Jeopardy

This got me thinking about the fascinating work of the late Andrew Ehrenberg, who first observed ‘Double Jeopardy’ – the principle that larger brands typically attract more customers AND those customers purchase more frequently, while smaller brands struggle with fewer customers who purchase less often.

Applied to charities, this would suggest that larger organizations (in terms of income) enjoy higher giving frequency among supporters. If true, this supports the view that brand growth primarily comes from expanding your supporter base, rather than the belief that smaller, niche charities can achieve higher giving rates.

It sounds a bit far fetched – surely the charity sector is immune from such a ‘natural law’, how can you apply behavioural patterns from other sectors to our own!

While Double Jeopardy is extensively discussed in advertising circles due to its implications for customer differentiation, I’ve found only one paper addressing this topic for charities: ‘An Investigation of the Structure of Charity Markets’ by Margaret Faulkner:

“findings on how (charity) brands vary are likely to apply giving us important information about loyalty patterns that could be useful to fundraisers (i.e. loyalty levels across charities are unlikely to differ greatly but the size of donor bases will, meaning there are big and small charities)… Bigger brands tend to have more customers and also have more loyal customers than smaller brands (Ehrenberg et al., 1990). For charities this means larger charities have more donors than small charities and their donors donate slightly more often on average. To get bigger, charities should focus on increasing the absolute numbers of donors”.

So there appears to be some evidence supporting the idea that giving behavior follows common patterns. But what do a million single givers tell us?

We examined just under a million single givers recruited by various charities in 2018 and tracked their Single Giving behaviour over five years. A more detailed explanation of the data can be found at the bottom of this article

Here are some of the key insights we’d love to share!

Frequency

‘A mile wide and an inch deep’

‘A mile wide and an inch deep’ is an often used quote to describe the public attitude towards development aid and global poverty – but it could also be used to illustrate the mean average number of times our cohort of recruits made a donation to the same charity during the five years after they were recruited.

The mean average frequency was 1.53, meaning on average a new Single Gift recruit gave an additional 1.53 single ‘one-off’ donations after their initial recruitment donation during a five year period. Yep, we were pretty surprised by that figure as well!

Of course the 1.53 mean average doesn’t mean anything – someone either gives or doesn’t, but the mean average is a useful way to quickly compare the fascinating differences across markets, organizations, different profiles and behaviours, so below we explore frequency and frequency distribution in more detail to help uncover some useful insights.

Frequency by market

The chart above shows the mean average number of single ‘one-off’ donations per recruit made within 60months after their recruitment for each market represented in the data.

Clearly there is a wide range of performances across different markets, however before we pass judgement on each market it’s well worth considering the spread of frequency performance of different organisations within each market.

In the chart below each circle represents a different participating organisation within that market and by hovering over the chart you can see the Median, 1st quartile and 3rd quartile values for each market, [the first quartile (Q1) is the value that separates the lowest 25% of the data, while the third quartile (Q3) separates the lowest 75%, meaning 25% of the data lies above it. They mark the edges of the middle 50% of the data in a chart].

By hovering over each circle you can see the mean average of that organisation and a count of single giving recruits the mean is based on.

We can also see from the chart below that those organisations with a mean average frequency between 0.5 to 1 donation post recruitment represents the most populated band with just over 1/3 of orgs, while just over 60% of all the organisations fall between 0.5 and 1.5 donations.

The role of recruitment channel

We’ve seen that there are significant differences in frequency by organisation and across markets, so can we see clearer patterns by other factors? First we look at whether the frequency of post recruitment donations is different for those supporters recruited by a digital channel (Email, Digital ads, SMS), and non-digital channels (DRTV, direct door drop, face to face, press etc).

If we look first without considering market, we see that the average mean frequency of those recruited by non-digital channels is double that of those recruited by digital channels.

And this picture is consistent across the 19 markets (shown below) with only three markets (Denmark, New Zealand and Taiwan) in which the frequency value for those recruited through a digital channel is either the same as, or higher than non-digital recruits. Suggesting that, other things being equal, whether a new recruit is recruited through digital or non digital type channels will impact their future behaviour.

By channel

If we look at specific channels we see that door drop then direct mail shows the highest average frequency of post recruitment donations at 3.09 and 2.53 respectively, while perhaps unsurprisingly retail network delivers the lowest frequency of 0.48 – indicating a huge proportion gave a single gift at the recruitment stage but never gave again.

It’s worth noting that the volume of recruits by each channel differs considerable, and this can be seen by hovering your cursor over the channel bar.

The chart above breaks out each channel by individual markets so you can see the spread of mean average frequency. For robustness we’ve only included markets where the volume of recruits for the channel was 1000 supporters or above, so many market channel combinations were excluded.

Spontaneous recruits [those donations received that were unprompted by a communication or cannot be attributed to a communication] have been included because the volume was notable with over 80,000, and we thought their ongoing frequency was surprisingly high at an average of 1.5

So overall we see that recruitment channel is a significant driver of future donations with direct mail and other non-digital channels giving a better performance than digital channels.

We’ll look at the comparison again when we look later at overall value – nevertheless the information does suggest careful consideration is needed when defining the optimum nursery stage and onward communication journeys to deliver additional donations for recruits based on their recruitment channel.

The role of age

You’ll know from our analysis of regular giving recruit performance that age is a highly significant driver of Lifetime value (LTV), with older supporters typically giving more, and for longer, often making them a charities most valuable supporters.

Below we look at the mean frequency by age band for each market.

The median frequency of gifts increases with supporter age, with 18-30 year-olds giving the least frequently post recruitment, and those aged 76-100 the highest average. With a strong correlation (0.66) of age to frequency of single ‘one-off’ donations, this suggests, similar to regular giving patterns, that other things being equal focusing on recruiting older supporters will likely deliver greater returns than targeting younger acquisition profiles.

There are some differences to this pattern of ‘older is better’ and we can see this below. The majority of markets do fit the pattern we’ve seen above, however there are some exceptions, notably New Zealand, Taiwan and in Australia while in the UK recruits aged 18-30 over perform.

Age or channel?

Talk of targeting acquisition activity always raises the relationship of channel and profile – and which drives the other!

While the result may differ within certain markets, overall we see a consistent pattern that regardless of the recruitment channel, recruits within the older age bands have a higher frequency of post recruitment donations than younger recruits.

So, other things being equal, we evidence that recruiting older Single Giving supporters will offer a greater opportunity for future gifts than younger SG recruits – regardless of recruitment channel.

Does proposition matter?

We do see some quite sizable differences in average frequency by proposition – whether the recruit was recruited via a communication that focused on an emergency or not an emergency – a key distinction in development / humanitarian fundraising given we known each type of proposition can attract different types of audiences.

The chart below shows that the average frequency of single gifts made post recruitment is significantly higher among non-emergency recruits at 1.6 donations, while emergency recruits recorded a mean average of 1.02 post recruitment donations within a 5 year period.

Given the influence of recruitment channel type, the chart also shows the proposition values split out by channel type.

For both digital recruits and non digital recruits non emergency recruits gave a higher number of post recruitment donations, however the difference is far greater among non digital recruits.

The next chart shows that in all but three markets we see that the future giving frequency is higher for those recruited on a non-emergency proposition than those recruited on an emergency proposition. However in many markets the difference is small, indicating that recruits from emergency appeals do offer good potential for future Single Giving behavior.

The chart above does not factor in the importance of recruitment channel type (digital or not digital), but the chart below does enable you to select a specific market and see the performance of proposition type by recruitment channel type.

Frequency by recruitment donation value

We’ve shown that the frequency of additional donations for digital recruits are significantly different from non-digital recruits, so for this chart we’ve shown them separately.

The chart above shows two useful insights – firstly that the frequency pattern for each recruitment donation band depends upon whether the recruit is from a digital channel or not. For digital recruits we see a clear pattern that as the recruitment donation band rises, so does the frequency of future donations. In contrast, for non-digital recruits it’s the lower donation values that deliver the higher frequency of future donations.

Secondly, we see the volume of recruits within each recruitment value band – illustrated by the size of the circle. The £20-£50 band is the most populated, and the volumes for recruitment donations over £250 falls considerably.

Frequency distribution

So far we’ve looked at the frequency of post recruitment donations by using the mean or in some cases the median average – which provides us with a simple metric to make comparisons by.

However, as fundraisers we know that the potentially high percentage of recruits who fail to make even a single extra donation after recruitment can distort the picture of how well our supporter development programme is actually working. Because of this, looking at the distribution of frequency can be more insightful.

The chart above shows the percentage of recruits within each market who make a specific number of post recruitment ‘single ‘one-off’ donations. The frequencies of giving above ten are banded for simplicity. We should make an important note that the communication frequency, the volume of donation requests, the campaign selection criteria and the channel through any request was delivered will differ for each organisation, and not something we’re reporting on here.

The markets are ordered by the percentage who failed to make any donation post recruitment – from least (Switzerland 41%) to most (Hong Kong at 82.9%). As mentioned the data is for recruits acquired in 2018 only, and each has had five years to generate additional gifts.

In addition to the proportion not giving any more gifts we see the percentage giving 1, 2,3 and so on, additional donations within the 5 year period – these may either be within the same year or across the years.

Two features stand out – firstly as a rough rule of thumb, the percentage giving 2 more gifts is roughly half that of the percentage giving 1 more gift, and this halving for every additional gift continues until you get past five donations. And secondly, that the level of difference between markets diminishes once you discount the recruits who failed to make any gift and concentrate just on those who give one or more additional donations post recruitment.

Does size matter?

When we look at the donation frequency distribution by the size of the organisation we see an interesting pattern, with the percentage giving one or more donations after their recruitment increasing as the organisational size gets larger. Here we’re using the value of the organisations single giving income in 2018 as a proxy for organisational size.

If we look at the same data but remove those who did not make an additional donation post recruitment we see that that the distribution pattern is very similar across the different size of organisations – suggesting that if the size of the organisation makes any difference it’s to the give again rate – not the frequency among those who do give again.

Or put another way, larger orgs may be more successful at gaining a second gift, but no more successful at generating high frequencies than smaller size organisations – questioning the double jeopardy theory.

Recruitment channel

We previously showed that we can expect differences in ongoing donation frequency depending upon the channel used to recruit the new supporter – so it’s no surprise that we see significant differences when looking at distribution by channel. But perhaps this chart really brings home how differently the recruits of different channels perform post recruitment with a spread from 37% not giving again up to 89% for those recruited through a retail network.

And remember this is not the percentage who did not give again within 12 months, but 60 months!

Below we’ve split out the distribution to allow us to compare the channels in greater detail.

Distribution by channel type

If we group channels into whether they are classed as digital or non-digital we get an even clearer picture by showing the cumulative percentage of each who have made X donations or less – so the lower curve is performing better. If we cap the chart at ’10 or less’ donations, we see at each point non digital recruits are outperforming digital recruits in terms of frequency.

Recruitment donation level

For the distribution by recruitment donation value we can see a trend that the lower the recruitment donation value the higher the percentage who failed to donate again – although we previously saw how the direction of this trend differs depending upon whether the recruit was acquired through non-digital or digital channels.

What we can see is that any trend wears off by the proportion who gave three or less donations post recruitment.

Looking at the importance of age

Previously we saw a strong correlation between age and the average frequency of donations – as age increases the average quantity of post recruitment donations also goes up.

The chart below looks at the relationship between frequency and age in a different way, by the proportion of recruits within each age band who make X donations or less – up to 10 donations or less. Meaning the lower the curve the better, as it means a higher percentage made over 10 donations.

The chart highlights how differently recruits over 60 perform compared to younger recruits, with 67% of under 45’s not making any additional donation, compared to 45% of the over 75’s. At each frequency level up Bands E and F show better frequency performance, before evening out at the ’10 or less’ donations point

Give again rate

As a fundraiser you’ll know that a vital KPI for most charities is the give again rate – highlighting that for many fundraisers the question is not the distribution shape of their supporters donations, but more simply, how many actually make a second gift at all! No doubt because getting that first donation after recruitment makes such a difference to the economics of that recruit and because many fundraisers use an additional gift as evidence of future potential.

We’ve already shown a little of the give-again values in the distribution charts, so we make an importance distinction between ‘Give again in Yr1’ and ‘Give again by Yr5’.

The first chart below compares the percentage of recruits for each market who gave in Yr1, verses the percentage who gave again at any point during the five year period.

What captured my attention was the relatively high values for the 3rd column which shows the increase in give again rate between the end of Yr 1 and end of Yr 5. I didn’t expect these to be so high, and it’s worth highlighting that for Sweden, the UK, Taiwan and the Netherlands the value for the 3rd column was equal to or exceeded the give again rate in Yr1.

This challenges the often held view that Give Again rate for Yr1 will determine the 5yr give again rate – i.e. if you’re not getting supporters to give again by end of year 1 it’s unlikely they will over a longer period of time.

The chart below enables you to see the 5Yr give again rate by each organisation within a chosen market (via the drop down), and the 5Yr rate is split between digital and non-digital recruits.

I wanted to include the two charts below to acknowledge that the averages used above do not show the spread of performance at an organisational level. In the following charts each circle represents a single organisation allowing us to see that even within markets there can be a wide spread of give again performance.

Give again rate for the first 12 months post recruitment

Give again rate for the 5 year period

The following four charts display the Give Again rates by some of the variables we’ve previously looked at, and as you would expect repeat the same pattern of behaviour, with age, channel and recruitment donation value suggesting trends in performance, while looking at give again rates by size of the organisation seems not to show any noticeable pattern.

Give again by age band

Give again by channel type

Give again by recruitment gift value

Give again by proposition

As we expected we did find a significant difference in give again rates depending upon whether the recruited was acquired on an emergency proposition or not.

Overall a higher proportion of Non-emergency recruits made one or more additional gifts. However if we then factor in channel type of recruitment we can see that the difference is largely dependent upon whether the recruit was acquired by digital or non digital channels.

Give again by size or organisation

Giving patterns

First year of giving

Let’s start with a stat that really surprised us – that a quarter of all the single donations that were made after recruitment were made within 12 months of the supporters recruitment date. To be honest we thought this proportion would have been far higher!

In contrast if we look by supporters, rather than donations, we can see below that of those who made a post recruitment donation, on average 54% gave their first post recruitment Single Gift donation within 12months. Looking at the pie chart you’d be tempted to adopt a rule of thumb that the proportion roughly (very roughly!) halves for each year post recruitment. So by year 5 only a very small proportion of recruits will make their first post recruitment Single Gift donation.

However we can see huge differences if we look at the market level below, with Australia and New Zealand standing out. We will see later that 90% of Australian recruits gave in only one year, and nearly 90% gave in Year 1, which may suggest a highly unusual communications journey!

For many other markets where a far higher proportion first gave in years 3 or more, it perhaps reminds us that applying a 24month recency rule to new recruits may be causing some supporters to be excluded from campaigns too early, notwithstanding the additional cost and risk of including them.

Are ‘early givers’ worth more?

In the chart below we compare the average financial contribution (£ gross) of new recruits and the year of their first donation post recruitment – these values exclude their initial recruitment donation.

While the mean average donation contribution for those whose first gift was in Year 1 was £322.59 it fell in a consistent way for each year so that for those whose first donation was in Year 5 after recruitment their mean value was £100.35

We wanted to understand to what extent this decline in value was due to frequency of gifts or the value of their donations – i.e. is the average donation of those who give in year 1 higher than those whose first donation was in later years.

One way to examine this is to look at those supporters who made only one donation post recruitment, and compare the mean value based on in which year they made that first and only donation.

We can see that for these recruits there is a slight increase in their gross value the later they give again, thus supporting our assumption that frequency of giving is driving value rather than the average donation gift value.

Giving across multiple years

Lets look at those recruits who did make one or more donations after their recruitment and we can see that a high proportion gave within either just one of the years or two of the years, while the percentage who gave in more than two of the years fell considerably, as expected given the frequency of giving. The value for four years is extremely low and not a single person gave across all five years!

We can see that there are large differences across different countries in the proportion who gave in multiple years, with 90% of Single Giving recruits to Australian charities giving only in the first 12months post their recruitment date. In comparison less than 60% of recruits to charities in Canada, Germany, Switzerland, Austria and Sweden only gave in a single year.

As you would imagine, the gross value of recruits increases as they give in more than one of the years. To see if this is simply due to frequency of donations rather than when they were made, we then look at the value by frequency as well.

About the data

Source

We’re incredibly grateful to the IFL Forum for supporting this analysis and giving us permission to publish this ‘deep dive’ into the behaviour of Single Giving recruits.

The data has been collated from the INDIGO benchmarking study – globally the largest study of its kind. The INDIGO study is run by Adroit on behalf of the International Fundraising Leadership Forum (IFL Forum), a peer group of international fundraising directors in some of the world’s biggest NGOs and UN agencies raising funds from individuals, companies and foundations in multiple markets or markets.

INDIGO is an IFL Forum study of Individual Giving behaviour across different organisations and markets.   The study is built from anonymized donor and gift records stretching back for the previous 6 calendar years. For the INDIGO study Adroit audit, collate, model, analyse and present the data to draw out key insights. Typically, over 1 billion gift records are processed on 80-90 million individual donors.  This is shared via Power BI reporting as well as webinar insight sessions with members from each market enabling organisations to understand market level trends and make peer comparisons, providing insights into donor acquisition, retention, upgrades, reactivation, life-time value and many more metrics.

For more information about the INDIGO study please click here

Composition

The dataset used for this analysis focussed upon the transactional behavior of Individual Giving supporters recruited in 2018 to the organisations included and were acquired from a single ‘one-off’ donation. Supporters whose recruitment donation or subsequent donations were over $2000 were excluded. Checks were included to identify any obvious miscoding of single gifts and outliers, and these contacts were removed bringing the number of supporters included from the initial 1,000,000 to 997,152.

The data from 75 organisations from 19 countries (markets) were included in this analysis. An important similarity between the organisations is that their main focus is upon humanitarian aid, disaster relief, development issues, human rights and/or the environment.

Analysis and commentary were written in June 2025 by James Long with support from Ella Briggs and Alexis Jones

Notes

Pearson’s correlation scores

Pearson’s correlation score measures the strength and direction of a linear relationship between two variables. It ranges from -1 to +1:

• +1 means a perfect positive linear relationship
• -1 means a perfect negative linear relationship
• 0 means no linear relationship

R squared values

R-squared (R²) is a number between 0 and 1 that shows how well data fits a statistical model. For fundraisers, it tells us how much of the change in say donations, can be explained by a factor you’re analyzing—like marketing spend or age

For example, an R² of 0.8 means 80% of the variation in donations is explained by your model, suggesting a strong relationship. A low R² means other factors are likely influencing results.

---

Ways we can help