What HMRC Charities Online taught us about church giving

The £560M problem nobody talks about

UK churches claim roughly £300M in Gift Aid annually. The potential is closer to £860M. The gap isn't laziness or ignorance. It's friction.

Church treasurers juggle rostas, maintenance budgets, and volunteer schedules. Gift Aid compliance is one more admin task, and it's dense. HMRC Charities Online works well if you're a large charity with a dedicated finance team. For a congregation of two hundred, it's Byzantine. Forms have conditional fields that change based on donor postcode. Declarations expire. The rules around small donations were rewritten in 2024, and hardly anyone noticed.

When we started interviewing churches, we heard the same story: treasurers manually cross-referencing donor names, phone numbers, and addresses against HMRC guidance documents. Some delegated the work to volunteers. Others filed incomplete claims because the alternative was spending six hours on paperwork.

What struck us was this: the problem wasn't technical complexity on the church side. The problem was that no platform existed to translate what congregants do (give money) into what HMRC demands (auditable, compliant Gift Aid data).

Learning to speak HMRC's language

We spent weeks inside HMRC Charities Online documentation. Not once. Many times. We read the submission schema. We tested the validation rules. We submitted test claims and traced what happened when fields were wrong.

Here's what we learned: HMRC doesn't want your best effort. It wants precision. Donor name format matters. Postcode validation is strict. If you submit a claim with a donor who has moved house, the claim gets rejected. If you submit incomplete data, you wait weeks for rejection letters.

But we also learned that HMRC Charities Online is actually elegant once you understand it. The schema is logical. The validation is there for a reason. There's no ambiguity. If you submit clean data in the right format, the system works without friction.

That became our north star: capture data so cleanly during giving that submission to HMRC becomes automatic, not a chore. When a congregant gives via Givr, they scan a QR code and give in fifteen seconds. In those fifteen seconds, they answer a few questions about Gift Aid eligibility. Givr captures name, postcode, and consent. Fifteen seconds of good data at the point of giving eliminates hours of manual work later.

The result is that treasurers can submit Gift Aid claims at the click of a button. No manual cross-referencing. No spreadsheet errors. No rejected submissions that arrive weeks later.

Why QR codes beat app downloads

Early on, we considered building a mobile app for congregants. Download our app, create an account, store your payment method, give with one tap. It seemed obvious. It's what most platforms do.

Then we talked to a church treasurer who said: 'We tried an app. Twelve people downloaded it. In six months, three of them uninstalled it. We went back to the collection plate.'

The friction of app downloads kills giving. Not because people dislike apps, but because asking someone to download an app to donate money feels like asking them to sign a contract to buy a coffee.

We scrapped the app idea. Instead, we built QR codes that open a browser page. No account creation. No stored payment methods. Congregant scans a code at the end of service, browser opens, they give in fifteen seconds, the browser closes. They never think about it again.

This sounds simple. But it forced us to rethink everything. How do you maintain security without an account? How do you capture Gift Aid consent without annoying the congregant? How do you make the experience smooth enough that someone who's never given digitally before feels confident?

The answer was obsession with the fifteen-second flow. Every extra field, every confirmation screen, every animation cost us in completion rates. We removed everything that wasn't essential. What remained was Gift Aid declaration, a little more breathing room, and the donation field.

The Stripe Connect decision

Building a UK giving platform means touching regulated payments. That typically means obtaining an FCA licence, hiring compliance officers, and burning cash for eighteen months before you launch.

We chose a different path. We use Stripe Connect Express, which lets churches connect their own Stripe account without us needing an FCA licence. This solved three things at once.

First, it simplified our compliance burden. We're not a payments processor. Stripe is. That's their job.

Second, it gave churches transparency. Your money goes into your Stripe account, then into your bank account. You control the flow. You see the fees. No mystery about where your giving goes.

Third, it meant we could launch to the market in months, not years. We could focus on the real problem: automating Gift Aid.

This architecture decision shaped everything else we built. It's why Givr stays narrow and focused. We're not trying to be a full financial platform. We're the bridge between congregant and HMRC.

What we're still learning

Six months in, we're still discovering edge cases. A church treasurer submitted a Gift Aid claim and discovered that the GASDS small-donation scheme rules had changed again. We've now built it into Givr. Another church asked if we could handle standing orders. That's why Gather tier adds GoCardless recurring giving.

But the core lesson remains: if you solve the thing that actually costs people time and money, everything else becomes easier to justify. Churches don't care about the elegance of our architecture. They care that they reclaim Gift Aid without spreadsheets and that their congregants can give without apps.

That's what HMRC Charities Online taught us. The platform works. The process is clear. What was missing was a tool that spoke both languages: the language of the congregant (simplicity, speed) and the language of HMRC (precision, compliance). Everything we've built since stems from that realisation.