Data is one of those words that makes plenty of fundraisers tense up. You picture spreadsheets, pivot tables and a dashboard you do not fully understand, and you quietly hope someone else will deal with it. The good news is that using your data well is far less technical, and far more useful, than that mental image suggests. It is simply about knowing who your supporters are, what they care about, and what is actually working, so that you spend your limited time and money in the right places.

Heather Egleton, a consultant at Hubbub, puts the worry plainly: "Some people, and you may be one of those if you're on the call today, hear the word data and kind of get a bit panicky. You think of spreadsheets and pivot tables and all sorts of other things that can blow some of our minds." Her argument is that you can move past that and start to enjoy it. "There are sometimes spreadsheets and pivot tables, but that's not all there is. There is the bigger picture as well."

This guide draws on two Hubbub webinars: "Why you shouldn't be daunted by data" and "A primer to web analytics using Google". It covers five things in turn: getting your data in order, segmenting it so your communications land better, the handful of metrics that genuinely matter, the basics of web analytics on your donation journey, and how to turn all of it into better decisions rather than prettier dashboards. It pairs naturally with our guide to donor acquisition and retention, because almost everything here is in service of acquiring, keeping and upgrading donors.

Working templates. Sign in to download our segmentation template, data health checklist and analytics dashboard template.

Start simple, and just start

The single most important thing is not to wait until you have a perfect system. Heather's advice is deliberately small: "If you do one thing after listening to this today, it is really just to make a start. It's very easy to kind of put data and planning into your virtual in-tray or your actual in-tray. But pull it out and crack on with it."

That first thing can be modest. "If that one thing is just that you take a step back and review what you have, how you're recording it, where you're getting it from, what use it could be to you, and looking at your end goal and can it help you get there, that's brilliant." You do not need a "fancy pants database", as she called it, to begin. A clean spreadsheet and a clear question are enough to get going.

There is also a useful reframe for the daunted: most of the work is interpretation, not arithmetic, and that is the part you are already good at. Being reactive has its place, and sometimes it is exactly right, as it was for the appeals many teams ran during Covid and around the war in Ukraine. But for your day-to-day programme, using historical information to make predictions is what makes you "a lot more cost efficient" and lets you "be really confident in the decisions you're making and the actions that you're taking".

Getting your data in order

Before you can segment or analyse anything, the data underneath it has to be trustworthy. The principle Heather kept returning to is an old one: "Good data in, good data out. Bad data in, bad data out. That has been said to me a long time ago and I will never forget it." Trustworthy data comes from three things: consistent entry, clean records, and proper consent.

Consistent entry. The biggest cause of useless data is inconsistent recording, especially when several people touch the database. Agree protocols for how things are entered, because "unless that information is in there accurately and consistently, it's no use really, because you can't report and analyse on it: you'll be comparing apples and oranges." Wherever you can, design the inputs to remove guesswork. Use "a tick box or a drop down choice" and restrict free text, so you avoid the classic mess where one record says a supporter studied "French and history" and another says "history and French", or an organisation appears once by its initials and once spelled out in full. Those small inconsistencies are exactly what break a report later.

Clean, deliverable records. Cleansing does not have to mean shipping your file off to a third party and hoping. The kind Heather recommends is the sort you can do yourself, with your donors' help. Before a big mailing or an impact report, build in a step to ask supporters to check and update their own details: "If you have an engagement platform, ask them to double check their data and update it. And then they're actually then also giving you consent because they're providing you with the most recent information." That keeps your data deliverable and earns you fresh permission at the same time. It is also worth periodically streamlining duplicate fields, because "if you've got the same thing said on this tab and that tab and then in this box, it's a lot harder to keep that all up to date".

Consented and proportionate. You do not need to be a GDPR expert, but you do need a simple test. Heather's version is memorable: "If someone asked to see it, would they be absolutely horrified about what you're collecting? Don't put anything in there that you wouldn't happily let someone look over your shoulder and see." Collect what is accurate, factual and necessary, keep it up to date, honour the communication preferences people give you, meet your GDPR obligations, and lean on whoever in your organisation understands the rules. Our data health checklist turns these into a short, repeatable review.

One habit underpins all three: start from the end. Think first about what you want to know, then work backwards from the result you want. If your aim is to grow regular giving, ask what data that needs. If it is to find legacy prospects, ask what signals someone is likely to leave a gift in their will. You will usually find three or four aims with a lot of overlap, which means you are streamlining what you collect rather than drowning in surveys and attributes.

Segmentation, and why it improves results

Once your data is clean, segmentation is how you make it pay. Segmentation means splitting your supporters into groups so you can talk to each group in a way that fits them, and the reason to bother is straightforward. As Heather explained, it "really helps us all to create more personalised communications with our audiences, which will in turn increase the warmth that they have for you and your mission and their affinity, which then hopefully will in turn increase the support they're able to provide you."

The blunt instrument segmentation replaces is the scattergun appeal. "Gone are the days, if they were ever actually there, where you just throw your whole database in a telephone campaign and hope for the best." Instead, you channel limited resources where they will work hardest, and the savings are real: print and postage are expensive, so "you really want to make sure you're segmenting and you're focusing in the right areas with any resources that you have."

You can segment on many dimensions, alone or in combination, depending on your goal:

Donation history. For an upgrade campaign, target those who have given regularly for a while, or made several one-off gifts, and steward them towards a higher or regular gift.
Relationship. Not only the relationship with you (alumnus, parent, former staff) but between supporters: a matriculation year, a subject cohort, people who were there at the same time.
Engagement. Those who always come to events, those who never do, and everyone in between, plus the activities and interests they have shown.
Communication preferences. Honour them once collected. Do not phone someone who asked never to be called.
Location. Domestic or worldwide, and finer breakdowns within that.

Two practical points make segmentation stick. First, record which segment each donor was placed in, and keep it in your appeal history, because what works for this summer's campaign may differ next year as you learn. Second, segmentation is also how you test content. You can "literally just split your data randomly in half. One half has a video in it, one half has a photograph", or one version is written from the head and another from a student. It is "not 100 per cent scientific, because there are lots of other variables", but it surfaces trends you can act on. Our segmentation template gives you a structure to define and tag your segments.

A connected technique is profiling: looking at the donors you already have to find more like them. Pull your recent givers into a spreadsheet, "not hideous ones, I promise", with columns for what they gave, when they last gave, their relationship and their age bracket, then look for the commonalities. A light RFM analysis, weighting by how recently, how frequently and how much people give, lets you "weight the ones that are giving the most, the most often, the most recently, and then start to look at what traits they share". You can then tag the unasked people in your database who match that profile.

Profiling also punctures lazy assumptions. As Heather warned, "you can't ask somebody over 70 because they don't know how to use computers, and you can't ask someone under 30 because these poor people can't afford to pay the rent". Those are assumptions, not data. "Throw all those assumptions out the window and look at what your data is telling you about who should be included and how they should be communicated with."

The metrics that actually matter

It is easy to collect dozens of numbers and report none of them usefully. A small, consistent set tells you most of what you need: participation, retention, average gift and lifetime value.

Participation. How many people gave, and how many were new. New-donor numbers in particular tell you whether your acquisition is working and where your future regular givers will come from.

Retention. Whether donors stay with you, lapse, or come back. Heather frames this as part of profiling: "Are they staying with you? Are they upgrading?" Tracking lapsed donors, and whether your activity re-engages them, is how you protect income you already have rather than constantly buying it back.

Average gift. What people give, and how that changes over time, both across the file and within a segment you are trying to upgrade.

Lifetime value. The longer view: what a donor is worth across the whole relationship, not just this appeal. This is the metric that justifies investing in stewardship and in acquiring new supporters who can be cultivated for years.

The discipline that makes metrics worth having is consistency. As Kat Carter put it, you want metrics "that are consistent" and "reliable enough to be collected on a regular basis to show change over time, whether that's positive or negative, and to see trends that help you to make informed decisions". A metric you measure once is a curiosity. A metric you measure the same way every month is a tool.

The basics of web analytics

Online giving has one big advantage over direct mail: you can see far more of what happens. With a mailing, Kat noted, "you're kind of only really gaining two data points": who gave and what came back undelivered. Online, "you can learn a little bit more about at what point did somebody decide that they want to make a gift, how long did it take them to make that decision, and what were the things that influenced that that were on the website."

The first step is almost embarrassingly simple, and it is Kat's top takeaway: connect Google Analytics to your website, especially if that website takes donations. "If you don't have Google Analytics connected and the Google Analytics code connected to your website, especially if that website is taking donations, that's the first thing that you should do." Even if you are not ready to analyse anything, you are at least collecting the data, so that "when you are ready to analyse it, you have something to get started with". If a website provider runs your site, they can add the tracking code for you.

Once it is connected, a handful of measures cover the donation journey:

Users and sessions. New and returning visitors, and how often they come back. People often need several visits "before they overcome their skepticism and make a gift", so returning users are not a failure.
Devices. Make sure the journey works on mobile, tablet and desktop. People browse and often give on mobile, so "if you have a website that's great on desktop but it isn't great on mobile but you get lots of users on mobile, then you could be missing out on conversions."
Channels, sources and mediums. Where traffic comes from: email, social, organic search, referral, direct. This tells you where to focus effort.
Average time on a page. A long time can mean engagement, or a friction point. "If somebody is spending too much time there, then maybe there's a friction point on that page."
Exit pages and bounce rate. Where journeys end. The question is whether people leave from the thank-you page, having given, "or are they exiting before they even get to the donation page". A higher bounce rate on a donation page is normal, because there is only one action to take, but you need to know your own benchmark.

To see which messages actually drive gifts, use UTM links, the tagged URLs Google provides free. They let you label a link by campaign, medium and source, so Analytics can tell email apart from social, and one email apart from another. For a giving day Kat recommends being granular: tag each email by day and send time, because then "I'm able to see more information about which emails are driving traffic, but which emails are converting." The pairing is the point: "It doesn't matter if it gets lots of engagement if it doesn't convert." When you find content that produces both high traffic and high conversion, "let's do more of that."

When you are ready for more, Google Tag Manager adds event tracking: how far people scroll, which of several donate buttons gets used, whether a video or a photo converts better. Hubbub tested two donate buttons on a giving day page and found the main button under the totaliser drew 609 clicks and 438 conversions, while the button in the navigation bar drew only around 100 uses. The lesson was not to remove the nav button, but to confirm both have a place: data over opinion. Our analytics dashboard template gives you a simple, consistent place to record these numbers over time.

Turning data into better decisions

The aim is decisions, not dashboards. Kat's framing is the one to keep: "Without data, you're just another person with an opinion. So let's be data-led, not opinion-led." Without it, "you might end up being totally unaware of what's working, what's not working, and just making guesses as to how you move forward."

Benchmarks turn raw numbers into judgements. Hubbub uses a few it has built up over time: on a giving day, around 85 per cent of donations typically come from email in the first year or two, and their donation forms convert at around 70 per cent. These are not targets to chase blindly. A team sitting above 85 per cent on email may simply lack channel diversification, while a team below it has often grown social and ambassador giving, which "is sometimes a good thing". The value is in the conversation the number starts: if your conversion rate is low, is it the audience, the messaging, or the user experience? Test each in turn, "do a test of your audience, do some messaging changes, test your giving on one platform versus another", and you can isolate what actually moved the needle. If you have no benchmarks of your own yet, Kat pointed to public ones from researchers like NextAfter and M+R.

Crucially, learning often comes from what did not work. "You almost want to find the things that aren't working to learn from them, because if you're looking at everything that's working, you're probably not going to learn too much from that other than it worked, but not maybe as to why."

A note on AI, since it comes up constantly. Both Kat and Heather see it as a helper for the manual parts, not a replacement for judgement. Where a task is "very manual and can almost be a blocker to doing the next step, which is the actual analysis", letting AI compile, cleanse or totalise data frees you for the work that matters. But, as Kat cautioned, "just letting AI do everything isn't the answer, because it doesn't have the savvy". You remain the professional interpreting what the data is telling you.

Finally, bring your team with you. People can be wary of data because "they think that they're just being tracked and watched". Be transparent about why you are doing it and what difference it makes, and find the person who lights up at this: "there generally is more than likely someone in your team that would get as excited about this as I do." Hand them the project, and the whole thing becomes sustainable rather than a one-off. As Heather summed it up, "data can really have an impact and a great success for you to acquire, retain and upgrade more donors."

Frequently asked questions

I find data intimidating. Where do I actually start? With one small thing. Review what you already hold, how it is recorded, and whether it helps answer a question you care about. You do not need an expensive database; a clean spreadsheet and a clear goal are enough to begin.

How do I keep my data clean without it taking over? Build consistency in at the point of entry with drop-downs and tick boxes rather than free text, agree protocols if more than one person records data, and use natural moments, like an impact report or a mailing, to ask supporters to update their own details. That keeps records deliverable and refreshes consent at the same time.

What is segmentation and why does it improve results? It is splitting your supporters into groups so you can communicate in a way that fits each one. It increases warmth and affinity, lets you focus limited budget where it works, and replaces the scattergun appeal where you "throw your whole database in and hope for the best".

Which metrics should I track if I only track a few? Participation and new donors, retention and lapse, average gift, and lifetime value. Measure them the same way every time, because a metric is only useful when it shows change consistently over months.

What is the first thing to do with web analytics? Connect Google Analytics to your website, especially if it takes donations. Even before you analyse anything, you will be collecting the data so it is there when you are ready. Then use UTM links so you can see which emails and channels actually convert.

Will AI do my data analysis for me? It can handle the manual parts, compiling, cleansing and totalising, which frees you for interpretation and relationships. But it lacks the savvy to make the call for you, so you stay the one reading what the data means.

Members' resource pack

Tools

Hubbub's Giving Day Simulator: model likely results before you commit to a target
Fundraising Copy Checker: sharpen your appeal and email copy
QR Code Generator: link in-person moments straight to a donation page

How to use your data and segmentation