Online Giving and the Donor Pipeline in Muin

Most giving problems are not giving problems. They are setup problems that nobody saw until a donor hit them.

A giving page goes live with no program behind it, and the receipt comes out with no purpose on it. A kiosk points at a bare fund, so the gift reports as undesignated. A Zakat appeal points at a program that is not Zakat-eligible, and the money lands somewhere it should not be spent from. Nothing errors. The gift goes through. You find out in January.

Muin now has one screen that answers “is our giving actually set up?” before a donor finds out for you.

Fund and program are two different questions

This is the distinction everything else rests on, and it is the one most often collapsed.

A fund answers which spending rules apply to this money? It is an internal management bucket. Most organizations need exactly one to start, and at most two: a general fund and a Zakat fund, because only Zakat has genuinely different spending rules.

A giving program answers what is the money for? This is the name a donor recognizes on their receipt.

They are independent. One program can be funded from more than one fund. One fund can pay for many programs. So a fund per cause is the wrong shape, and it is the shape people reach for first.

A fund never appears on a donor-facing surface. Receipts, giving history and year-end statements name the program. If a receipt came out with no purpose on it, the page is missing a program, and adding another fund will not fix it.

Three steps, then a list of everything a donor can reach

Update: Giving Setup is now the Online Giving page (Nonprofit → Giving → Online Giving), and Run setup fills in your programs and a draft giving page for you. See Set up Online Giving.

Open Nonprofit → Giving → Online Giving, or press Run setup to have your programs and a draft giving page filled in for you. Before you have anything, it shows Set up giving in three steps, and the steps tick themselves off as you finish them:

Create a fund. The pot the money lands in.
Create a giving program. What the money is for.
Add a way to give. The page, kiosk or embed a donor actually uses.

If you would rather not assemble it piece by piece, the How does your organization give? picker proposes a whole structure that fits your organization. Name the rows it suggests, press Review what will be created, read the list, then Create this structure. Nothing is created until you confirm.

Below that sits Ways to give: every cause a donor can reach, with its program, its fund, which surfaces it is live on, its tiers and its status. There are deliberately no totals, trends or dollar figures on this screen. It answers one question, and money has its own reports.

Needs attention

The useful part is the strip at the top. Warnings cannot be dismissed, and each one carries the link that fixes it:

A giving page, kiosk or embed resolves no program: Set a program
A membership or giving tier has no program: Set a program on the tier
A Zakat-eligible surface points at a program that is not Zakat-eligible: Make the program zakat-eligible
No fallback program for gifts nobody designated: Choose a fallback program
A kiosk points at a bare fund: Re-point the kiosk, or Point it at a page
A fund name is reaching a donor-facing surface: Name a program

When it is clean it says so: “Nothing needs attention. Every way to give names a program and a fund.”

The failure state is worth reading too. If the check cannot run, the strip says “We could not run the setup check, so nothing here means ‘all clear’.” An empty warning list after that message is not a pass, and it does not pretend to be one.

Where the donor goes next

Setup is the front half. The back half is what you do with the donors it brings in.

Cultivation Pipeline (Nonprofit → Giving → Cultivation Pipeline) shows five columns: Prospect, Qualification, Cultivation, Solicitation, Stewardship. Each donor row shows how long they have been in their current stage, their next planned action and its date, and a Stalled badge once they have sat in one stage for 90 days or more. Overdue cultivation actions appears below the board when a next step is past its date.

The board is read-only on purpose. A stage is set on the donor’s own record, so there is one place a stage changes and one history of those changes, and the board reads it rather than becoming a second source of truth. A full-width Not yet staged card sits above the columns, because donors with no stage at all are the ones a pipeline tends to hide.

On a donor’s record, two numbers are calculated from their own giving history rather than generated: an Engagement Score out of 100, built in four equal parts from recency, frequency, monetary value and growth, and a Lapse Risk percentage. The AI parts of that screen are labelled as such and are generated when you ask for them: a stewardship briefing, a handover dossier, and a chat scoped to that one donor.

Giving History on the donor record lists charitable, tax-deductible gifts. Payments that were not gifts are listed separately, under their own heading. That separation matters more than it looks: a Sunday-school tuition payment is revenue the organization earned, not a gift somebody gave, and counting it as giving makes a donor out of a customer.

Forward-looking, not fortune-telling

Giving Intelligence adds three tools that run on your own numbers:

Revenue Forecast projects five streams from your trailing and prior windows, with a diversification score. A stream with no data is flagged “no data” and left out of the projection rather than counted as zero.
What-If Simulator gives you three sliders, from 0.5x to 2.0x: donor retention, average gift, new-donor acquisition. It returns a year-end total with a confidence band and the change against your baseline.
Risk Radar scores composite risk from six weighted factors: funding concentration, top-donor dependency, cash runway, restricted-funds reliance, donor lapse and seasonality. A factor with no data scores as unknown rather than as safe.

All three are deterministic calculations, not model output. The same inputs give the same answer, and you can explain the number to a board.

The pattern

The common thread is refusing to report something as fine when nobody checked. A setup check that cannot run says so. A forecast stream with no history is flagged rather than zeroed. A risk factor with no data is not scored as low risk. A pipeline board shows the donors nobody staged instead of quietly leaving them out of the counts.

A dashboard that is always green tells you nothing. These tell you where to look.