How AI is Rewriting the Rules of Supporter Engagement

"Do things that don't scale" - Paul Graham's 2013 essay has become gospel in the startup world, encouraging founders to embrace high touch, labour intensive processes to win their first customers before worrying about automation.

For years, mass participation fundraisers across the nonprofit world have dismissed this advice as an unaffordable luxury. The standard directive for most fundraising teams has been crystal clear:

"Deliver your fundraising proposition to the widest possible audience at the lowest possible cost, with success measured primarily by this years ROI."

But what happens when technology fundamentally changes this equation?

Twelve years after Graham's essay, AI has created a world where personalised, human relationships can potentially scale to thousands or millions. The question is no longer whether nonprofits can afford to "do things that don't scale" - it's whether they can afford not to embrace technologies that make deep connection possible at unprecedented scale.

The accessibility of AI creates an entirely new playing field that transforms supporter expectations. The challenge for nonprofits isn't just adopting new tools - it's maintaining a competitive edge in building authentic relationships when emotional connection at scale becomes the norm rather than the exception.

Damned if you do, damned if you don't

Since wrapping up at GivePanel (a company who works on bridging the gap between getting a potential supporter's attention and their first fundraising activation), I've intentionally taken time off to meet with folks working on their proposed futures of nonprofit tech.

Without the lens of a specific product, I've had the privilege of seeing how people, across the sector, are thinking about the future of mass pax fundraising.

What's become abundantly clear to me is that nonprofits have lived with a frustrating dichotomy for years:

Option A: Provide deep, personalised engagement that creates meaningful supporter relationships but requires tremendous human resources.

Option B: Automate as much as possible to reach more people, inevitably sacrificing the quality of those interactions.

This trade-off is particularly painful in fundraising, with most teams settling for an uncomfortable middle ground between these extremes. However, as supporter expectations evolve, I'm concerned that this compromise approach will soon prove not to be good enough.

From "Don't Scale" to "Exponential Thank You Economy"

I should start by saying that I'm not pushing any particular solution in this post (hence there's a lack of name dropping and case studies) - I'm just sharing my perspective on this both for feedback & challenges and to explore new opportunities that might emerge from this conversation.

I'm also aware that AI in itself isn't new.

Even in the nonprofit sector, I'm very late to fully grasp how it's being deployed by fundraisers and technologists around the world today.

However, what we might call "Democratised AI" - the mainstream acceptance and understanding that erupted when ChatGPT went viral in late 2022 - has fundamentally changed what experimentation is now seen as both acceptable and necessary.

This shift hasn’t just made room for another efficiency play - it's transforming what previously "didn't scale" into precisely what CAN scale, putting enterprise-level capabilities into the hands of organisations regardless of their resource constraints.

Gary Vaynerchuk's decade old "Thank You Economy" concept advocated for personalised attention that made customers feel valued, but required significant human hours. Today's AI enables what we might call an "Exponential Thank You Economy" - where personalised/authentic notes to thousands of donors and meaningful conversations with event participants happen without armies of staff, volunteers or telefundraisers.

The high touch, labour intensive work that Graham suggested startups embrace is now available to every fundraiser, almost regardless of team size. This is already changing the game.

So, What's Holding Us Back?

As I noted in a previous article, a lot of the challenges in fundraising innovation are primarily human, not technological (borrowing from human centric tech enthusiast Tim Lockie's concept of the "human stack" versus "tech stack").

Despite having powerful off-the-shelf AI tools available, organisations resist implementation.

Graham observed that founders avoid recruiting users individually due to "a combination of shyness and laziness."

This same hesitation appears in fundraising innovation - as an example, we often default to mass communications partly because we fear the work that might come from deeper engagement. Investing in solutions that potentially create more work seems counterintuitive.

The irony is that this is precisely where the greatest value lies. A one-time blast to 10,000 supporters might generate only a couple of replies, while creating opportunities for genuine conversations—aided by AI tools that handle the prioritisation, setup and scaling—might initially create more work but can deliver dramatically better ROI. The reluctance to embrace this productive discomfort is understandable, given the limited resources available to us & the sometimes prohibitive costs of trying new technology solutions. However, this mindset shift often separates transformational results from incremental improvements and may soon become the baseline expectation from your supporters.

This friction is described in what Clayton Christensen calls the "Innovator's Dilemma" - noting that established organisations are often structurally unable to pursue disruptive innovations because they're optimised to serve existing needs with existing solutions.

Nonprofits are facing this classic dilemma now. We are incentivised to maintain current fundraising approaches that reliably deliver short-term ROI often at the expense of investing in potentially transformative but initially unproven technologies. I fear that sadly, the risk is no longer just missing an opportunity but it's finding ourselves scrambling to catch up when these approaches become industry standard.

A Short Note

Outside of this, I'm aware that there are very valid concerns in embracing AI at scale that are holding us back at an organisation level. Of which, most reasons could be grouped under either:

1. Institutional inertia: Established systems and processes make change psychologically and financially costly.
2. Technology transition costs: The immediate expenses of migration, retraining, and potential disruption outweigh uncertain future benefits.
3. Fear of the inauthentic: Leaders worry AI-enabled personalisation will feel fake and organisations worry about the wider ethical concerns of using AI.

These are all fair concerns and AI adoption undoubtedly comes with dangers. However, 2025 isn't 2022 and there does now seem to be an increasing numbers of ways to lean into AI-enabled human connection without needing to bet the house.

A Starting Point

Whilst, I have nowhere near the expertise to suggest how to become a team that embraces rather than fears the power of AI. I would suggest that we need to be doing more as individual's and teams to make more use of AI than in just token ways, such as drafting communications and tweaking your content for different channels.

To take your approach beyond the basics, you could start in far worse places than exploring the below:

• Investigating proactive contextual outreach to your supporters based on behavioural patterns that indicate both risk and opportunity.
• Looking at behavioural journey mapping for events that responds to supporter signals in real-time, not just segment-based automation.
• Producing content that evolves based on what resonates with individuals, not just broad personas.

All this and more is possible today and all while keeping humans squarely in the loop.

Fundraisers still approve, refine, and personalise AI-generated communications, applying their empathy, intuition, and relationship knowledge where it matters most. The technology allows fundraisers to be more human, not less - freeing them from repetitive tasks to focus on meaningful interactions and strategic thinking that truly require the human touch.

In my conversations over the past few months, I've noticed a clear divide: those asking "How do we better automate what we're doing?" and those saying "What would we do if we could have a meaningful conversation with every supporter?"

My prediction for 2025, is that the latter group is quietly creating extraordinary results that the rest of the sector will be racing to catch up with in the coming year whereas those orgs tipping their hat to AI only by using GPT wrappers to help create content and "improve internal efficiencies" may be falling behind.