The strongest argument for a website investment usually loses to a calendar.
Here is how it happens. For most January fiscal years, the board votes on the 2027 budget in November or December, but the document they vote on takes shape in September and October, built from whatever requests were on the table then. If your digital line item wasn't in that draft, the vote is mostly a formality, and it happens without you. You'll be told to bring it back next year, and next year the same thing happens.
After nearly two decades of working only with nonprofits, we've seen far more digital requests die this way than get voted down on the merits. So this post covers the part of the process most guides skip: not what to ask for, but how and when to ask so that the request survives budget season.
Fall is the window, not December
For January fiscal years, the real decision window is September and October. That's when department heads submit requests, when finance builds the draft, and when the trade-offs get made. By the time the board meets to vote, the shape of the budget is largely settled. The meeting adjusts numbers at the margins, and it rarely adds a whole new category.
The timing is cruel in a second way. Late in the calendar year is exactly when a weak website hurts most, because sector studies consistently put December around a third of annual giving, and it's also exactly when nobody has time to fix anything. So the pattern repeats itself: the site struggles through year-end, everyone agrees in January that something must change, and by the time the organization catches its breath, the next budget cycle has already closed.
Breaking that loop doesn't take anything sophisticated. It takes a written request, in front of the right people, before the draft hardens, and the rest of this post is about making that request easy to approve.
Speak board, not marketing
A board rarely thinks of itself as funding a website. What it funds are mission outcomes, retired risks, and staff capacity, and a request written in marketing language forces board members to do the translation into those terms themselves, which makes the request easy to defer. In practice, most of how to present a marketing budget to a nonprofit board comes down to doing that translation yourself, before the document ever reaches them:
- "Website redesign" becomes "donors and program participants who search for us can find us, understand what we do, and give without friction." (If you're not sure the case is there, our rundown of the signs your nonprofit website needs a redesign is a useful gut-check.)
- "Accessibility remediation" becomes "we retire a legal risk and make sure the people we serve can actually use our site, which makes it a mission item rather than a purely technical one."
- "Maintenance and support contract" becomes "staff hours currently spent wrestling with the website go back to programs."
Each translation has the same two ingredients: a person the board cares about, and a change in what happens to that person. If a line item can't be stated as an outcome for donors, participants, or staff, it isn't ready to go into the document yet.
Bring evidence, not aspirations
A request that says "we want a better website" reads as a wish, and boards are slow to fund wishes because a wish has no boundary. A request that says "here is what's broken, and here is what it's costing us" reads as a repair instead, and repairs are easier to approve because a repair ends when the broken thing works.
So before you write a single budget line, spend an afternoon documenting the current state. Complete your own donation on your phone and note every point of friction. Check when the site was last updated, and by whom. Ask your newest staff member to find three basic things (your programs, your annual report, your donation page) and watch where they get stuck. Screenshot what you find.
This honest audit does two jobs. It converts the request from a want into a repair, and it becomes your baseline, meaning the thing progress gets measured against when you report back to the board next year.
It also lets you name a credible number. Boards distrust round guesses, so anchor the ask in real ranges; our breakdown of how much a nonprofit website costs exists precisely so you can cite something defensible instead of a hope.
Lead with the free money
There is an ordering decision that changes how the whole request reads, which is to put the items that cost nothing first.
The biggest one is the Google Ad Grant, a Google program that gives eligible nonprofits up to $10,000 per month in free Search advertising. The money is real, and so are the rules: ads are capped at a $2 bid, and the account has to maintain a 5% click-through rate (the share of people who see your ad and click it) or the grant can be suspended. Those rules are why the grant takes management to keep, and we've written a full Google Ad Grant strategy guide on running it well. But eligibility itself costs nothing, and applying costs nothing.
When your budget document opens with "step one costs $0 and brings up to $10,000/month in advertising," everything after it reads differently, because you arrive as the department that found money before asking for any, and the paid lines become the modest investment that makes the free part work. Boards tend to find that kind of plan much easier to approve. You can check your eligibility with our Ad Grant eligibility quiz before you put it in writing.
Add a line for AI visibility
There is one item that was on almost no 2026 budget and belongs on the 2027 one. A growing share of the people who used to search for an organization now ask an AI assistant instead: which food banks serve this county, which after-school programs take referrals, where to give for a cause they care about. Whether your nonprofit shows up in those answers depends on whether the assistant can read your site and find clear, specific statements about what you do and where you do it. The practice of making that happen is called generative engine optimization, or GEO, and our guide to GEO for nonprofits covers it in full.
Two things make it a good budget line. First, the baseline is free: run your site through our AI Visibility Scorecard and you get a graded report you can paste into the request as a documented finding, the same way you would a screenshot of a broken donation form. Second, the fix usually overlaps with work you are asking for anyway. Google's own documentation says there are no additional requirements to appear in AI Overviews or AI Mode beyond the fundamentals of a well-built, crawlable site. In the scorecards we run, the pattern we see constantly is a visually excellent site whose raw code is nearly empty because everything on the page is rendered by JavaScript, so the assistant has nothing to read. That is the same defect that trips up screen readers, and it gets fixed in the same redesign or accessibility line.
Some of the organizations we work with have already put this line in. A family-building nonprofit that started GEO work with us in August 2026 is now cited inside Google's AI Overviews for 22 different searches, up from 19 in July, and the searches it is cited for grew from about 2,500 to about 4,000 a month, including the highest-volume question in its field, measured with Semrush between July and October 2026. A Washington, DC women's shelter went from being cited only on unrelated queries to being the source the AI Overview cites for "homeless shelters for women and children," a search made about 590 times a month. Neither needed a new website to get there; both needed their pages to say plainly what they do, for whom, and where. Our AI visibility service is that work, packaged.
In board language: "when someone asks an AI assistant which organizations do what we do in our area, we are in the answer, and we can show the board how often." That last clause is what makes it fundable, because the scorecard gives you a number to report against at the next meeting.
Pre-answer the three objections
Nearly every digital request meets the same three objections, and answering them inside the document, before anyone asks, changes how the meeting goes.
"Why now?" Because of the calendar in section one. Waiting for a more convenient year means going into another December giving season with the current site, and then waiting through another full budget cycle after that. Name that cost plainly.
"Why can't current staff handle it?" Answer with your evidence rather than with a defense of your team. If staff could have fixed these problems in the margins of their week, the problems in your audit wouldn't exist. Frame the request as protecting their time: it moves specialist work to specialists so that staff hours go back to programs.
"What if it doesn't work?" This is really a question about accountability, so answer it with a measurement plan: the baseline from your audit, two or three numbers you'll track, and a commitment to report against them at set board meetings. A measurement plan doesn't guarantee a result, but it does guarantee the board will always know whether the investment is working, and that visibility is what governance actually requires.
The one-page structure
Everything above fits on a single page, and one page is the right length, because a board can absorb it during the meeting and a champion can forward it without explanation. Copy this structure:
Digital investment request — FY2027
1. Current state. Three to five documented findings from your audit, each one sentence, each observable. No adjectives.
2. Cost of inaction. What continues if nothing changes — donation friction through another year-end season, unmanaged risk, staff hours lost. Two or three sentences.
3. Proposed investment, by outcome. Free items first (Ad Grant application and management, the AI visibility baseline), then each paid line stated as the outcome it buys, with an honest cost range. Never a tool list.
4. How we'll report back. The baseline, the two or three numbers you'll track, and which board meetings get the update.
5. The ask. One sentence: the total, and the decision you need by which date to hit the fiscal year.
That last line matters more than it looks. In our experience, a request that names a decision date usually gets an actual yes or no, while a request without one tends to be received warmly and then pushed to a future meeting.
If you're drafting one now
If you're putting a 2027 request together and want a second set of eyes on it, bring it to a strategy call. We'll pressure-test the framing, flag what a board is likely to push on, and tell you honestly which pieces you don't need an agency for. We've been nonprofit-only since 2007, so we've read a lot of these documents over the years, and we have a fairly good sense by now of which arguments boards accept and which ones they send back for another year.
