How I built a 91-asset year-end campaign in eight hours

5:05

How fast can AI make a whole campaign's worth of assets?

In my test, 91 assets for a complete nonprofit year-end campaign took eight hours of my time. That covered the strategy, the appeal, GivingTuesday, social, email, video and an internal hub. The speed came from a foundation that was already built, and the result still needed the customer's team to refine it.

I wanted to see for myself how far one person could get in a single working day. When Scott and I sat down to talk about it afterward, I opened with the headline.

I'm living in this world where we're going to talk about this today, where I built the whole year end campaign for a nonprofit in eight hours using AI.

Colleen Cook · Recorded conversation, Sep 30, 2026

The organization is an If Possible customer with a small team and a big new vision. Their campaign needed to do more than raise the December number. It needed to start a new kind of conversation with their donors about where the organization was headed.

What was in the 91-asset campaign?

Everything a development team would normally scramble to produce between October and January. The campaign started with a plan the team could take to leadership, then moved through the appeal, a GivingTuesday push, a post-GivingTuesday push, stewardship pieces and an internal hub. Most pieces were made in several versions for different donor segments.

The social content alone was meant to carry them through November, December and into January, organic and paid. The hub is the piece I'm proudest of. Your staff, and any board members who want to help, can work from a single place.

Where did the eight hours go?

About half went to thinking. I split the work into four two-hour sessions, three on a Wednesday and one on Thursday morning. I started from a recorded strategy conversation and spent roughly four hours making decisions in a campaign command center before most assets existed.

over the course of those eight hours, I would guess I probably spent four of the hours in the command center thinking through decisions, building strategy, and working with the plan.

Colleen Cook · Recorded conversation, Oct 14, 2026

Before the clock started, I recorded myself talking through what I wanted the campaign to be, including the focus of the messaging, the ask, the goals, and where I saw opportunity in the donor base. Then I put myself in the director of development's shoes and asked what everything is that actually gets done in a year-end campaign. The command center surfaced the gaps, like whether we had consent to use a story, and each answer became a creative decision.

The making happened in Claude Design, and the method mattered more than the tool.

My first request was a proof sheet, a deck showing every asset I could make from the prompt that came out of the command center. My first prompt left out some of the copy, so I went back, asked for everything, and got a new proof sheet. I gave feedback on that and moved forward. The last thing I built was a review board, because 91 pieces is too much for me to track all at one time.

What would a campaign like this normally cost?

Based on ten years of running an agency, Scott and I put a campaign like this at about 250 hours of agency work. Scott does the math at $150 an hour, which comes to $37,500. For a small nonprofit, that price usually means the campaign never happens at this scale.

it would have typically taken 250 hours for our agency to do it. Now at $150 an hour, that $150 an hour is $37,500.

Scott Williams · Recorded conversation, Oct 13, 2026

Those are our estimates, and they are the only numbers I'll claim here. The 250 hours is what our team has seen a campaign of this size take. The eight hours is my own clock.

What had to be in place before the eight hours started?

The foundation. We had spent about a year with this customer on their brand, their offering and their case for support, and none of that was AI generated. They also had an AI brain, a content engine and their brand rules recorded in one place. That groundwork is why eight hours was enough.

You can't just go to AI, ask for a year-end campaign and expect something good. You'll get something that looks like AI made it and sounds like AI wrote it, and that can quietly undercut your brand, because people long for things that feel human and real. From a fundraising standpoint, you want it to be yours.

When your AI knows your organization, has teammates trained in different skills, and has a content engine that understands how your brand expresses itself in different forms, then work like this becomes possible. If you're starting from zero, start with the foundation. Our post on making your brand voice your first AI system is the place to begin, and keeping a person in the director's chair is how you keep the output from turning into slop.

Can someone who isn't a designer make campaign assets with AI?

Yes. My background is in fundraising, and I built every asset in this campaign, including the video and the web pages. The skills that matter most are knowing what the campaign needs to say and being able to judge whether a piece is good.

Earlier this fall I built a giving day campaign for another If Possible customer, an international nonprofit, on a Saturday afternoon with Claude Design. I asked their team whether everything was correct and whether it felt like them. Their director of communications said she "would happily deploy this right now." Another time, after a workshop, I spun up an eight-page workbook on the themes we'd discussed in about 25 minutes, edits included.

The customer from this campaign has seen the same thing on their own. After getting their intelligence layer set up, their team produced a month of content in an hour.

What should you expect if you try this yourself?

Expect to get far, and expect to keep refining. My honest read is that eight hours got this campaign to third base. The customer's team will sharpen the messaging, make it fully theirs and decide which pieces to actually run. That's the right division of labor.

Scott framed it for the customer as a toolkit. They "might only use 10, 20, 30% of it, and that's okay." The value is seeing every option before you decide, which almost never happens with a traditional campaign.

If I'd had more time, I would have gone much deeper on the review process and gotten hypercritical about every detail. I kept myself to eight hours on purpose. With your own campaign, give the review the time it deserves.

To aim your campaign before you build anything, read build it, aim it, then launch your year-end campaign. For the appeal itself, see writing a year-end appeal with AI, and for social, 10 nonprofit social media prompts that earn attention and gifts. There's more on the fundraising hub.

If you want ideas like this in your inbox each week, join the newsletter. And if you'd like to build your own year-end campaign this way, alongside other nonprofit teams, get on the cohort waitlist.