Co-authored by Cantarus (Mark Eichler, Chief Product Officer & James Jemmett-Page, Senior Consultant) and Betty AI (Johanna Byrne, Chief Operating Officer & Emily Stamm, Director of Customer Growth)
Please note our content disclaimer in relation to blog posts.
Most membership organizations have now experimented with AI in some form. Teams are using ChatGPT, Copilot and AI-powered search tools. Vendors are rapidly adding AI features to existing platforms. AI is no longer a future consideration – it is already part of the digital landscape.
Yet many organizations are still struggling to answer a simple question:
What value is AI actually delivering?
The challenge facing membership organizations is no longer adoption. It is proving impact.
Despite a 21% increase in AI adoption across the sector, only 6% of organizations currently have a formal AI strategy and just 3% have deployed member-facing AI tools. The result is a growing gap between experimentation and measurable organizational value.
In a recent webinar, Cantarus and Betty AI explored how associations can move beyond isolated AI experiments and focus on outcomes that matter: operational efficiency, member value and organizational growth.
The latest MemberWise Digital Excellence Report, published in mid-2026, provides a useful baseline. AI adoption among UK membership organizations has grown by 21% in a single year, with 26% of organizations now using AI in some form. That’s a meaningful shift. But the figures beneath the headline tell a different story.
Mark Eichler: "Only 6% have a strategy on AI, which is really an eye-opener for me when I saw that figure. And only 3% have member-facing AI tools."
The predominant use case remains LLM-powered chatbots and content search – a sensible starting point, but one that leaves significant opportunities on the table. Personalization, content discoverability, data-informed member engagement: these are capabilities AI can support well, and most organizations have yet to pursue them in any structured way.
Mark framed the broader picture bluntly: the sector is in a "dabbling phase." Organizations are experimenting, but without formal strategy or governance in place, many are doing so in ways that carry unexamined risk – particularly around data safety and their responsibilities as data controllers.
For many organizations, AI activity is outpacing organizational readiness. Teams are experimenting with tools, pilots and use cases, but few have established how success will be measured or who owns the outcome.
Without that foundation, AI risks becoming another technology initiative that generates interest but struggles to demonstrate long-term value.
One of the clearest messages from the session was a call to reverse the typical sequence. Rather than selecting an AI tool and then searching for a use case, associations should start with the problem they need to solve – and work backwards to the technology.
Johanna Byrne: "We don’t want you to spend hours or days or months putting together a 100-page strategy document. Nobody looks at those strategy documents."
The alternative proposed was a four-question framework, designed to be completed in a short team session rather than a drawn-out strategy process:
1. What problem are we solving? Stated in plain language: "members can’t find what they’re looking for on our website," rather than "we want to use AI."
2. What does success look like? A clear outcome, equally plainly stated.
3. How will we measure it? What metric, where does it live, and what’s the baseline today?
4. Who owns it? Without a named owner, AI initiatives tend to lose momentum and fail to embed.
The emphasis was on intentionality over comprehensiveness. The point is not to produce a document but to establish enough clarity that every subsequent implementation decision can be tested against a defined objective.
Too often, organizations begin with a product demonstration rather than a business challenge. The result is a pilot that generates initial enthusiasm but ultimately fails to scale because nobody has defined what success looks like.
When AI initiatives struggle, it is rarely because the technology failed. More often it is because the organization never established what problem was being solved, how success would be measured, who owned delivery, or how outcomes would be reported.
This is where many AI projects stall.
The webinar identified three areas where AI can create demonstrable impact for associations: operational efficiency, member experience and revenue. The speakers were clear that these aren’t theoretical – they’re grounded in what both organizations are seeing with clients.
For many membership organizations, the most immediate opportunity is not reducing costs. It is increasing capacity.
Teams are being asked to deliver more services, more content and more personalized experiences without a corresponding increase in resources.
AI can help organizations reclaim time spent on repetitive activity and redirect that capacity toward higher-value work such as member engagement, service development and strategic planning.
The most immediate gains tend to come from reducing the volume of routine queries that consume staff time. One Betty AI client reported a 50% reduction in the time staff previously spent fielding member questions, with an AI assistant handling the initial discovery instead.
Emily Stamm: "One of our associations reported this year that since they implemented their Betty, they’ve seen 50% more time back for their staff that they previously had used to deflect questions."
On the Cantarus side, AI is accelerating project delivery – particularly during discovery and content migration phases, where synthesizing large volumes of stakeholder feedback or auditing years of accumulated website content has traditionally been labor-intensive work.
James Jemmett-Page: "The process of assembling and synthesizing and pulling insights out of that data traditionally has been incredibly manually onerous. But utilizing AI allows us to pull the nuances out of people’s feedback, get much richer results to help inform our website projects."
Mark added that AI-assisted content audits can help teams assess quality, identify duplication and evaluate relevance at scale – shifting the conversation from evaluating individual pages to thinking about buckets of content. For organizations sitting on 10 or 15 years of accumulated web content, that reframing saves substantial time.
A recurring theme was that this is about expanding capacity, not reducing headcount. Johanna was direct on the point.
Johanna Byrne: "Rather than ‘we have to use AI,’ it becomes ‘we get to use AI’ – which I think is a really important shift."
Membership organizations don’t have a content problem. They have a discoverability problem.
One organization highlighted during the webinar found that in some months around 95% of its content was being consumed by only 5-10% of its membership base.
The value already existed. Members simply couldn’t find it.
This is where AI has the potential to transform the member experience – not by creating more content, but by helping members discover and use the expertise organizations have already invested in.
Associations invest heavily in producing content for their members, and members consistently cite access to content as a primary reason for joining. Yet the two frequently fail to connect. James shared an anonymized client example that illustrated the scale of the problem.
James Jemmett-Page: "In some months, around 95% of their content by volume was only being consumed by around 5 to 10% of their membership base."
The only content gaining meaningful traction was whatever sat on the homepage or was pushed directly via email. Deeper resources – policy documents, white papers, guidance notes – were going largely unread, despite survey data confirming that members valued exactly that type of material.
The fix came in two phases. First, improved content categorization and tagging, followed by AI-powered implicit personalization to surface relevant content based on browsing behavior. Second, the deployment of a natural language knowledge assistant that allowed logged-in members to search the full content library conversationally. The client is now reporting improved content performance across the board and stronger member engagement.
Emily extended the point beyond findability to what members do once they’ve found the right content – using AI to translate answers for different audiences, generate reusable outputs, and put membership benefits into practice rather than simply reading about them.
The insight layer matters too. Several organizations are using data from AI-powered interactions to identify content gaps, understand what members are searching for in real time, and create new material in response. One Betty AI client is now producing podcasts directly informed by the questions their members are asking.
Emily Stamm: "They’re taking that real-time insight and turning it into content that is helping their industry with what they’re facing today."
AI’s revenue potential for associations goes beyond cost savings. The webinar surfaced several models already in use: AI-powered buyer’s guides with paid vendor listings (one association charges $250 per listing), public-facing AI assistants that drive new membership sign-ups by helping prospects experience the value of the knowledge base, and specialized AI instances – study buddies, certification prep tools – offered as premium member benefits or separate subscription products.
Mark noted that study buddy functionality is proving particularly effective for engaging younger, career-focused members who are app-first in their habits.
One theme surfaced repeatedly throughout the discussion: confidence.
Many organizations are not questioning whether AI works. They are questioning whether they can trust it enough to scale it.
That confidence comes from four areas: clear governance, appropriate data controls, defined ownership, and measurable outcomes.
Without those foundations, AI remains an experiment. With them, it becomes an organizational capability.
Data quality is one of the most common reasons organizations delay AI projects.
Ironically, it is often the same organizations that are struggling with member engagement, content discoverability or operational efficiency that could benefit most from getting started.
The reality is that very few organizations have perfect data. The objective should not be perfection. It should be readiness.
Mark Eichler: "You can think of your data that you have as the bricks and the AI tool to help you build a new extension to your digital estate – the AI tools and the AI capacity might be the mortar to put those bricks together."
Most membership organizations have decent levels of data availability from their CMS, CRM, or AMS platforms. The data doesn’t need to be perfect – it needs to be sufficient, with a clear metric and a named owner. In many cases, one of the early outputs of an AI deployment is better data collection, as the system surfaces gaps that weren’t previously visible.
James Jemmett-Page: "We don’t feel that you have to have perfect data before kicking off an AI project. You need to have sufficient data. You need to have a clear metric. You need to have an owner."
The governance dimension matters here too. James observed that UK organizations in particular tend to get bogged down in pursuit of the ideal AI policy, and that while caution around member data is appropriate, at a certain point it becomes a drag on progress.
James Jemmett-Page: "You shouldn’t let trying to attain perfect, with regards to your strategy and governance, stop you making progress."
The session closed with a forward look at where AI in the membership sector is heading. Mark outlined a near-term future where semi-autonomous AI agents act on behalf of members – assembling personalized briefings, flagging relevant new publications, supporting career progression – rather than waiting passively for a search query.
Central to this shift is the Model Context Protocol (MCP), an open standard developed by Anthropic and now broadly adopted, which enables AI tools to interact directly with the systems where an organization’s content and data already live. James highlighted that Umbraco has integrated MCP functionality into version 17 of its CMS, enabling AI agents to read content directly from the CMS content tree – cross-referencing articles, checking tone of voice guidelines, flagging outdated terminology, and identifying metadata gaps.
Mark Eichler: "As people use a semi-autonomous agent to go build that dossier of ‘here’s what you need to know, here’s what we recommend the most important priorities are for you to accomplish today’ – that kind of stuff is going to be much more possible very soon."
For associations, the implication is that the content and data infrastructure they build now will determine how well they can serve members through these emerging channels. Websites built on platforms with strong MCP support will be better positioned to deliver authenticated, personalized experiences via AI agents without requiring entirely new systems.
The organizations that will see the greatest impact from AI over the next three years will not necessarily be those using the most advanced tools.
They will be the organizations that start with a clear business problem, define measurable outcomes, establish ownership and use AI to enhance member value rather than simply automate tasks.
The message from the webinar was clear.
Don’t wait for perfect data. Don’t wait for the perfect governance framework. Don’t wait for a 100-page strategy document.
Start with a problem worth solving, measure the outcome and build from there.
Because AI maturity is no longer about experimentation.
It is about creating measurable organizational value.
This blog captures the key themes from our recent webinar, but the full conversation goes further. Watch the replay of Measuring the Impact of AI in Associations to hear the discussion in full, including practical examples and audience questions.
If your organization is exploring how AI can deliver measurable impact – whether through improved content discoverability, operational efficiency or new revenue models – the Cantarus and Betty AI teams would be happy to discuss your priorities, challenges and opportunities.
Whether you are considering your first AI initiative, looking to scale existing projects, or exploring how AI can enhance member experience and organizational performance, we’d welcome the opportunity to share what we’re seeing across the sector.
Get in touch with Cantarus or visit Betty AI.
Together, Cantarus and Betty AI help membership organizations identify high-impact AI opportunities, establish meaningful success metrics and deliver AI initiatives that create measurable organizational value.