Last week we were joined by Georgia from Refuge, alongside Jess Hodge, Partner Strategy Lead at Forward Action, to unpack how Refuge built a winter fundraising appeal that raised £530,000 – 127% of their goal, with half of that income coming from new supporters.
With so much ground to cover, we couldn’t get to every question live, so we’ve pulled together this write-up with answers to what was asked. If you want the full context behind any of these answers, watch the webinar recording here or read our key takeaways blog post on planning a winter appeal.
Strategy & channel shifts
Was there a direct mail element to the winter fundraising appeal?
No – this was an entirely digital campaign. Refuge finds that digital delivers stronger results for them, and it also gives more flexibility during the campaign itself: if something isn’t working on Meta, you can turn it off immediately rather than waiting weeks to see how a mailing performs. Of course, this will vary depending on your own audience, so it’s worth testing what works for you rather than assuming digital-first is right for everyone.
Was the shift in income by channel based on re-strategising and allocating budget from insights, or was it live performance-led?
A bit of both. Refuge looks at trends across the whole year, not just the winter appeal – if a channel is consistently performing well, more budget goes behind it going into the next campaign.
But there’s also room to pivot mid-appeal: if PMax is delivering strong returns during the campaign window, budget gets shifted there from Meta in real time, and vice versa.
Having tracking and testing in place throughout the appeal is what makes that in-flight flexibility possible.
Tracking and targeting
Could you clarify what PPC and PMax are?
PPC (Pay-Per-Click) is paid search advertising – essentially paying to appear at the top of search results.
PMax (Performance Max) is Google’s automated ad programme that runs campaigns across its full network (Search, Display, YouTube, Gmail and more) from a single campaign setup.
If we’re not currently tracking anything, where can we learn more about setting that up? Literally the basics!
Start by agreeing what you actually want to track, as a team, before worrying about tools. At minimum, that’s:
- Who is coming to your donate page
- Where they’re coming from
- How much you’re paying to get them there
Once everyone agrees on those questions, you can build your tracking set-up around answering them, rather than trying to track everything at once.
What do you use to track donations on different channels?
Refuge uses a mix of tracking. Generally they refer to Engaging Networks’ (their donation platform) built-in tracking to see how donations are coming in, but they also make use of whatever in-platform tracking is available.
So for Meta, they might look at the ROAS or CPA using EN tracking data, but they’ll also consider things like the frequency of ads or number of clicks, as shown in Meta’s ad platform.
Then they also have Google Analytics to get a bit of a picture of how supporters move around their site / to identify any major drop off points.
Did you test any adverts on Microsoft Bing?
Not this year – Refuge is still seeing strong performance from Meta and PMax within individual giving, so that’s where budget is focused for now. Other teams at Refuge may test Bing in future, and as with most things in this space, that could easily change next year.
Will you still build your own audiences, even with Meta pushing us towards Advantage+ so strongly?
Yes – there isn’t a one-size-fits-all answer here.
Refuge tests Advantage+ (or broad match) audiences against more specific ones on most appeals, and which one wins varies by time of year and can feel fairly unpredictable. Specific audiences like mums or grandparents have performed well for Refuge, particularly at Christmas, even though a broader audience might win at other times of year.
The advice is to keep testing both rather than committing to one approach.
Gifts, messaging & donation framing
Are the gifts – colouring books etc. – restricted donations? / With the messaging and tangibility, how do you approach this when funds are unrestricted?
It’s a mix. For campaigns like Christmas, Refuge fundraises a portion that is restricted – for example, parcels for women and children, often fulfilled as vouchers so recipients can choose what they need. Donors see illustrative costs and a clear description of what their gift supports (including product-style items like colouring books and teddy bears, often offered as an on-page top-up). Once that restricted target is met, further donations go towards whatever is most needed at Refuge at the time – and that caveat is included on the donate page itself.
If you’re worried that “this could fund…” framing undersells the ask: we usually find that a strong, tangible story about impact does most of the persuading. The “could fund” line is a smaller part of the picture than we in the sector sometimes assume.
Do the top-ups come after the initial donation is completed?
Yes – the prompt appears just before the donation is finalised, as a pop-up along the lines of “would you consider adding X as well?”. Refuge sees around 50–60% of supporters add a top-up when it’s framed this way, styled similarly to e-commerce checkout flows (e.g. “you bought a teddy bear, would you like to add a colouring book too?”).
As with everything else, it’s worth testing different top-up wording to see what resonates with your own audience.
What would you suggest putting on a shopping list for an LGBT+ young people’s charity where we mostly do schools programmes? It’s hard to convey the emotional pull with our “products”.
When your product set is less tangible, or the ask amount doesn’t map neatly onto a physical item, it can help to shift from “£X buys Y object” to “£X buys Y outcome” – for example, what does £5.50 make possible for one young person in terms of support or access, even if there’s nothing you could physically put in a box? The tangibility comes from the specificity of the outcome, not from it being an object.
A few other angles worth testing:
- Lean into urgency, especially around key moments like the start of a school term – giving people a reason to act now.
- Talk about funding gaps directly, rather than trying to force a literal shopping list.
- Consider a collective framing (“if this many people club together, we can make X happen”) rather than shoehorning in individual item costs – forcing a shopping list that doesn’t naturally fit can weaken the ask rather than strengthen it.
Brand & channel balance
Do you focus on brand awareness as well as fundraising over winter? If so, how do you balance the comms and keep it unified for the supporter?
Refuge works from one unifying proposition so all channels are working from the same core message and tone – but that doesn’t mean every piece of activity ties directly back to the Christmas appeal. If a strong press opportunity comes up, the press team will run with it, and it’s linked to the appeal where it makes sense, but not forced when it doesn’t.
In reality, supporters aren’t seeing your Instagram post, your email and your PMax ad all on the same day – they’re building a general impression of your organisation over time. As long as your tone of voice is consistent, small mismatches in messaging across channels matter far less than we often assume. Given fundraising messaging is often what makes up the bulk of what supporters actually see, keeping that tone aligned with your broader brand is what really counts.
Email strategy
What numbers are we talking about when it comes to testing for email? Do you take a sample of the audience and see what works?
It depends on your list size.
For smaller lists:
A straightforward 50/50 split test works well.
For larger lists:
The gold standard is testing a subject line or content variant on around 10% of your list, then sending the winning version to the remaining 90% – so the majority of your audience always gets the best-performing version.
As a rough rule of thumb, you probably want at least 10,000 recipients before running this kind of test, to get a meaningful result.
Refuge segments their list based on past giving behaviour, testing within relevant segments before rolling out to the wider list.
We’ve noticed you’ve been testing “Refuge” as the email sender instead of “[Person’s name], Refuge” – could you share any results?
We don’t have exact figures to share, but the shift was informed by M&R and Rally benchmark data suggesting fundraising appeals from an organisation name tend to outperform those sent from an individual.
For stewardship-focused emails, though, putting supporters in touch with a named person can still build valuable connection, even if the email isn’t sent directly from their personal inbox.
It’s a good reminder to keep testing sender name as behaviours shift, and to lean on resources like the M&R and Rally benchmarks as a starting point if you don’t have your own historic data to test against.
Suggestions or links to resources for email list growth?
Here’s a few blog posts we’ve written over the years (these contain links to watch previous webinar recordings too):
- Lessons in lead gen: integrating channels to make meaningful connections
- How to mobilise your next generation of supporters online
- The general election series: Goal 1 – Using digital to build your base
- Lessons in paid social: how to cut through, mobilise supporters, and drive digital growth
Check out our YouTube channel for all our webinar recordings on demand.
Paid social creative
How many creatives per theme or angle did you run at once on paid social?
Refuge separated out campaigns per audience and per channel (IG or FB), and tend to run about 5-15 variants for each one. That’s not to say that they’re unique – sometimes the same ad works well across a range of audiences and channels. Often they’ll start with a smaller number of variants and then introduce more throughout the campaign as they gain a better understanding of what is working.
Getting started on a smaller budget
What are some easy wins for small charities with limited budget, staff or skills to run a full marketing and testing strategy?
A few starting points that don’t require a big budget:
- Start with email. The supporters on your list are already warm, and sending an email is free – so it’s the highest-leverage place to begin testing messaging and offers.
- Look into Google Ad Grants, which give eligible charities free ad spend to test paid search.
- Focus on activity that gets you a result, even if it’s not the end result. Lead generation and email list growth (rather than asking straight for a regular gift) mean that even supporters who don’t donate immediately are now on your system, ready for you to build a relationship with over time – which is far more cost-effective than paying for cold donor acquisition outright.
- Try daisy chains (multi-step supporter journeys, e.g. from a petition or hand-raiser into a donation ask) – Refuge has seen these work so well that they’ve effectively become a donor acquisition tool in their own right, not just a list growth one.
Thank you to everyone who joined us live, and to Georgia for being so generous with Refuge’s insights and results. If you’re starting to plan your winter appeal and want a hand thinking it through, get in touch – we’d love to help.