Good UX for a nonprofit is not about delight. It is about removing the specific friction between a person who needs something and the thing they need, whether that person is looking for help, offering time, or trying to give money at eleven at night on a phone with one bar. Every other definition eventually turns into a conversation about hero images.
This piece is deliberately structured as work rather than as reading. There is a test you can run in the next ten minutes, a scored audit, and a prioritization step that most audits skip, which is why most audits end up as documents nobody actions.
01 · The gap nobody is treating as urgentHalf the visitors. A quarter of the money.
The 2026 M+R Benchmarks put numbers to something anyone who has tried to give on a phone already suspects. Mobile drives 52% of nonprofit website traffic but produces only 28% of revenue. Mobile donation pages convert at 8% against desktop's 11%, and for small organizations mobile drops to 4%. The average mobile gift is $88 against $168 on desktop.
View as table
| Measure | Mobile | Desktop |
|---|---|---|
| Share of website traffic | 52% | 48% |
| Share of transactions | 43% | 57% |
| Share of revenue | 28% | 72% |
| Donation page conversion | 8% | 11% |
| Average gift | $88 | $168 |
Some of this gap is real behavior. People do make larger considered gifts on a laptop. But a 24-point spread between traffic share and revenue share is hard to explain by donor psychology alone, and the conversion difference on the donation page itself points at something in the experience rather than in the donor. The benchmarks do not isolate a cause, so treat that as the likely explanation rather than a finding. Digital wallets are worth checking first: PayPal is accepted by 79% of the nonprofits M+R surveyed, Google Pay by 58%, Apple Pay by 57%, and those are the payment methods that remove typing from a phone.
Before any redesign conversation, complete a donation on your own phone, on cellular, and count every tap and every keystroke. In the audits we run, the same five things come up again and again: no digital wallet, too many required fields, a small tap target, a slow form, and a recurring-gift option that is hard to see.
02 · The cheapest research you will ever runFive seconds is all you get.
The five-second test is the highest value-per-minute research method available to a small team. Show someone your homepage for five seconds, take it away, and ask three questions. What does this organization do? Who is it for? What could you do here? If the answers are vague, your homepage is not underperforming because of the design. It is underperforming because it is not making a claim.
Run it below to feel the timing, then run it properly with three people who do not work with you. A neighbor, a board member's partner, someone at a coffee shop. Three people is enough to find the obvious problem, and the obvious problem is usually the one costing you the most.
Run it on your own homepage
Open your homepage in another window. Press start, look at it until the timer ends, then come back here and answer the three questions honestly. Doing this on your own site is instructive; doing it with a stranger is the real test.
Ready when you are
Open your homepage in a second window or on your phone. When you press start, look at it and nothing else until the timer reaches zero.
If you hesitated on any of the three, so will a first-time visitor, and they will not have your context. The fix is almost never a bigger image. It is a plain sentence, above the fold, that says what you do and for whom.
In our sessions, three things go unrecalled more than anything else: the geography you serve, whether services cost money, and what a visitor is supposed to do next. Those facts are usually present somewhere on the page, just not in the first five seconds' worth of attention.
03 · The floor, not the ceilingAccessibility is getting worse, not better.
The 2026 WebAIM Million found detectable WCAG failures on 95.9% of home pages, up from 94.8% the year before, with an average of 56.1 errors per page, up 10.1%. Both measures moved the wrong way this year. In the nonprofit sector specifically, only 26% of organizations report designing their site for people with visual and hearing impairments.
The encouraging part is how concentrated the failures are. Six error types account for 96% of everything detected, and every one of them is a decision made once in a template rather than a thousand times in content.
View as table
| Failure type | 2025 | 2026 |
|---|---|---|
| Low contrast text | 79.1% | 83.9% |
| Missing image alt text | – | 53.1% |
| Missing form labels | 48.2% | 51.0% |
| Empty links | – | 46.3% |
| Empty buttons | 29.6% | 30.6% |
| Missing document language | – | 13.5% |
Two things worth saying plainly. First, automated scanners such as WAVE and axe DevTools catch only a portion of real accessibility problems; they will not tell you that your link text says "click here" seventeen times or that your focus order jumps around. Run the scan, then navigate your primary journey with the keyboard alone and watch where the focus indicator goes. Second, for a mission-driven organization this is not a compliance exercise. The people most likely to be excluded by an inaccessible site frequently overlap with the people you exist to serve.
Fix contrast, form labels, and empty links and buttons in your templates and you have addressed the majority of what an automated scan will ever find, permanently. Then do the keyboard walkthrough, because that is where the problems a scanner cannot see are hiding.
04 · The auditTwenty checks. About ninety minutes.
Five areas, four checks each. Do it with your actual site open, and with a phone in your hand for the mobile section. Score honestly; "sort of" exists as an option because most real answers live there. Below the audit, your answers assemble themselves into a priority matrix, which is the part that turns an audit into a plan.
Score your site, get your plan
Each check carries an impact weight and an effort estimate, which is how the matrix below sorts your fixes. Everything runs in your browser and nothing is sent anywhere.
Your score, area breakdown, and a prioritized matrix will build as you answer. Twenty checks is enough to find the problems that matter and few enough that you will actually finish.
Every failed check is placed by how much it affects your mission outcomes and how much work it takes. Start top left and move clockwise.
One rule about the matrix that matters more than the matrix itself: assign an owner and a date to every item in "Do now" before you close this tab. An audit without owners becomes a document, and a document becomes evidence at the next planning retreat that you already knew about all of this and did nothing.
How to measure whether it worked
Pick metrics tied to the mission rather than to the site. Donation page conversion split by device is the sharpest single measure; sector averages give you a target of 11% desktop and 8% mobile to beat. Then add volunteer or service inquiry completions, form abandonment rate, and the share of visitors reaching your eligibility or "get help" page. Record the numbers before you change anything, because the most common failure in nonprofit UX work is improving something real and being unable to prove it.
06 · Turning findings into changesThe playbook, by role.
Your job is the five-second answer and the words that carry it.
Put one plain sentence above the fold
What you do, who for, where. Not a tagline. "We provide free legal help to renters facing eviction in Santa Clara County" outperforms every abstract mission statement ever written, and it is the single change most likely to move your five-second test results.
Run the test with three real strangers this month
Not colleagues, not board members who know the programs. Three people with no context will find the problem that a year of internal debate has not. It costs an afternoon and it settles arguments that budget cannot.
Own contrast during brand work
Low contrast text now appears on 83.9% of home pages and rose almost five points in a year, and it is most often introduced by a well-intentioned palette refresh. Check every brand color pairing against WCAG before it reaches a template.
Your job is the giving path, and it is where the largest recoverable value on the site sits.
Complete a mobile gift yourself, this week
On your own phone, on cellular, standing up. Count taps and keystrokes. Then remove half of them. The gap between mobile's traffic share and revenue share is the biggest single number in nonprofit digital right now and most of it is form friction.
Add the digital wallets
Among the nonprofits M+R surveyed, PayPal is at 79%, Google Pay 58%, Apple Pay 57%, Venmo 44%. They remove typing from a phone, which is most of what makes mobile giving hard. If you are missing them, this is a configuration task, not a project.
Make the recurring option visible, not buried
Monthly giving is the most durable revenue a nonprofit can build, and on mobile it is frequently a small toggle below the fold. Surface it, label it in plain terms, and show what a monthly amount actually funds.
Your job is the path for people seeking help, which is usually less tested than the path for people giving money.
Answer "do I qualify?" before anything else
State eligibility as conditions in plain language, on the page, in text. Not in a PDF, not in a paragraph about your service philosophy. This is the single most consequential piece of content on a service page and it is routinely the least accessible.
Watch three people try to get help
Give them a realistic task and do not help, however painful that is. Where they hesitate is your navigation problem, and it will not be where you expected. This is also the fastest way to discover that your programs are named for internal reasons.
Keep hours, locations, and thresholds current
A stale eligibility number is worse than a missing one, because someone acts on it. Put a visible reviewed date on the page so both visitors and answer engines can see it is maintained, then actually review it.
Your job is to make sure the findings become owned work rather than a document.
Assign owners and dates before the meeting ends
Every "Do now" item needs a name and a date attached in the room. An audit without owners is evidence at the next retreat that you already knew. The prioritization is the deliverable, not the findings.
Track device-split conversion as a board metric
Donation page conversion by device, measured before and after. Sector benchmarks give you 11% desktop and 8% mobile to beat. Two numbers on a board slide will fund more UX work than any amount of qualitative description.
Treat accessibility as mission, and say so
Only 26% of nonprofits report designing for visual and hearing impairments while sector-wide failure rates are rising. For an organization that exists to serve people, an inaccessible site excludes the population you are most accountable to. That framing moves budgets that compliance language does not.
We will run this audit with you.
A working session on your real site: the five-second test with fresh eyes, the mobile giving path timed end to end, a keyboard and contrast pass, and a prioritized matrix your team can start on the same week. No deck, no discovery phase.
Sources and method: Mobile and desktop traffic, transaction, revenue, conversion, average gift, and payment method figures from M+R Benchmarks 2026. Accessibility failure rates, error counts, and year-over-year changes from the 2026 WebAIM Million, an automated analysis of the top one million home pages; note that automated testing detects a subset of real accessibility barriers. The 26% figure for nonprofits designing for visual and hearing impairments is from the 2026 Nonprofit Tech for Good report, as compiled in its 2026 website statistics roundup. The 20-point audit, the weighting, and the priority matrix are Socient's own. This is an independent article and is not affiliated with or endorsed by M+R or WebAIM.