.jpg)
AI is reconfiguring how events are marketed, sold, programmed and monetised, much like e-commerce reconfigured retail. Call it the industry's Amazon moment: the organizers who understand the shift early will lead it.
That was the premise of Change agents: How AI will transform the business of events, a Grip webinar featuring Ben Fletcher, Chairman of Oxford Global and founder of Fast Growth Icons; Marco Giberti, Founder and CEO of Vesuvio Ventures and co-author of Reinventing Live; and Tim Groot, CEO and founder of Grip. The conversation drew on an upcoming Grip report based on interviews with more than 12 industry leaders about how AI will change the business of events.
Here is what they said and what it means for anyone responsible for the strategy and revenue of B2B events.
The most consistent finding from Grip's research was also the most reassuring. Nobody serious believes AI will replace live events.
"It doesn't matter where in the world or how people are involved in events, the people that we've interviewed all believe that in-person is here to stay, and if anything is becoming more important in the age of AI, as a trust accelerator." — Tim Groot, CEO and Founder, Grip
The market agrees: as Marco Giberti pointed out, private equity has validated the industry with billions of dollars invested, including major acquisitions in recent weeks. And as AI floods digital channels with synthetic content, authentic human connection becomes scarcer and more valuable. Events are one of the few places where trust is built at scale.
What changes is everything around that moment of connection. Whatever their background, every leader interviewed believed they would soon do more with the same number of people, or the same with fewer. Budgets will shift away from people and towards systems that automate huge amounts of workflow.
Marco Giberti dedicated chapters of his book Reinventing Live to moving events "from three days to 365". At the time, colleagues told him the idea was decades away.
"Now with AI acceleration, it's probably not going to take decades. It's going to take months or years. And that idea of moving the three day event to the 365 platform is becoming more and more real." — Marco Giberti, Founder and CEO, Vesuvio Ventures
His advice to owners and C-level leaders was unambiguous: use AI for efficiencies, and use it to augment the event experience from a revenue perspective. "If you are an organizer right now, AI is no longer an option," he said. "Every single event category should be experimenting with AI as aggressively as possible."
AI also unlocks audience expansion that was never operationally possible before. Tim shared an example from the report interviews: one CEO described having 5 million people in a database of which only half a million attend his events. The opportunity is to use AI to create experiences, including highly targeted regional micro-events, for the 4.5 million who know the brand but never engage physically. That reframes AI from a cost-saving tool into a growth engine.
Ben Fletcher made the strategic case for treating events businesses as media businesses sitting on uniquely valuable data.
Large language models are trained on historical data. Ask one to write your event programme and it will produce something plausible built from last year's information. An organizer whose programme is grounded in current interviews with buyers and practitioners will beat it every time, and that current knowledge is exactly what events generate constantly. Ben took it to the logical extreme: the data inside an events business may become valuable enough that an AI company acquires one purely to access it.
So what should organizers capture? Tim prioritised two assets:
Session transcripts. Record and transcribe every session at every event. Transcripts become fuel for content and a validation engine for future programming. Grip now validates new webinar topics against the registration performance of every previous topic.
Connection data. Badge scanning is the minimum. Grip's click-to-connect captures an average of 20 connections per person, producing far richer data on who talks to whom. Feed that into AI and you can run a gap analysis on your event's network: who is missing, and what would make the networking mix stronger.
Ben added a third: intelligent surveys that branch based on each answer, extracting far more personalised information from the same send. That data improves marketing, powers personalised agendas and makes matchmaking dramatically more effective. His prediction: attendees will eventually arrive with their own context files, ready to hand their objectives directly to an organizer's AI.
When an audience question asked about prompting, Ben pushed back on the premise.
"Prompts is a very last year way of thinking about things, because agents are so great. The whole point about agents is they're constantly learning and building up this memory of you, so you don't have to prompt them." — Ben Fletcher, Chairman, Oxford Global
Agents hold persistent context about your business: machine-readable files describing what you do, how you work and what you are trying to achieve. You maintain that context once and the agent applies it to every task.
The panel was clear-eyed about where we are on the curve. This is the 1996 of the internet: messy, fast-moving, tools changing every week. But the direction is set. Agents will increasingly create software, build experiences and, crucially for this industry, book and recommend events on their users' behalf.
Ben has tested all of this in the field. Over the past nine months he built a working AI system for Oxford Global, then released a public version called Assyrio after a LinkedIn post about it went viral.
It works like a virtual SDR for sponsorship sales. Load in your events: programmes, delegate lists, speakers. It crawls competitor event websites to find sponsors, flags which are already your clients, and runs weekly so you spot a new sponsor the moment they appear on a competitor's site. A lookalike function finds companies resembling your best-fit sponsors, including those that just raised funding. It enriches contacts, scores every prospect against your event context and drafts tailored outreach, with humans approving matches before anything is sent.
Two lessons stood out. Proof of concepts are now astonishingly fast: Ben produced eight prototypes in a couple of weeks. Getting one production-ready took vastly longer, and adoption was harder still. An AI system delivers nothing until people are trained to use it and trust it.
His end state is instructive: a Monday morning where the agent has already run the system, found new sponsors, assessed the top five prospects and drafted the outreach.
Marco Giberti, who spent over a decade as President and Regional Partner at Reed Exhibitions, framed the strategy question around scale.
Large organizers with hundreds of events will build AI layers that generate serious efficiencies across SG&A, operations and marketing, but their tech stacks, complexity and culture make them slow. Small organizers with five to ten employees and one or two shows can be far bolder. That asymmetry is creating something new: the AI-first event organizer. In December, Marco asked to meet the first event founder building with an AI-first mentality. He is now talking with two.
His caution, drawn from daily conversations with investors: do not over-commit to any single tool. There is a new winner every week. Whatever you build or buy, architect it so you can swap components without starting over.
The panel converged on the same practical starting point: begin with low-value administrative work.
"It's so easy with AI to do something that's 80%, 90% effective, but to get it to 99% effective will take you at least twice as long, and possibly 10 times as long." — Ben Fletcher, Chairman, Oxford Global
Ben references the march of nines. The implication: target processes where 90% accuracy is acceptable and a human can handle edge cases. Basic research, contact enrichment and data entry go first. Programme research and sponsorship selling stay human for much longer. A 10% headcount efficiency is plausible, but not overnight: Oxford Global still runs human checks on everything its virtual SDR produces, handing over more as confidence grows.
Tim shared the rule Grip uses internally for every automation: it must not require people to change their behaviour, and it must not require new data to be entered. His example is post-sales-call discovery notes. Salespeople used to finish a call and manually fill in Salesforce fields. Now AI reads the call transcript, applies context about Grip's business and completes the fields automatically at 90 to 92% accuracy. That is good enough: the notes need to be directionally right and present for every meeting. The result is five hours a week returned to the sales team.
His sharper question: once the human leaves the loop, ask whether the process needs to exist at all.
Tim closed with the webinar's boldest claims. Companies now measure themselves on revenue per employee, heading from roughly $500,000 five years ago towards $5 million. When your attendees' time is worth ten times more, premium events can charge accordingly.
"All of those events might charge something like $2,000 or $3,000. I think that's going to go to $15,000 to $20,000 per ticket in three years' time." — Tim Groot, CEO and Founder, Grip
Crucially, he argued this only works for new events. There is no incremental path from a $2,000 ticket to a $20,000 one. The only route is to launch something ten times better and ten times more expensive from day one. Combine that with AI agents becoming the primary discovery channel, where a field marketer simply asks their agent which events to attend, and Tim's prediction follows: within 12 months, an event will launch and generate over $10 million in first-year revenue.
Marco pushed back. Ten million dollar launches are risky: he would rather build to $10 or $15 million by year three with the right buyer-seller ratio and proven ROI than maximise sponsorship too fast and hollow out the community. But he fully agreed on the threat to incumbents, with new organizers monetising through subscriptions and online education that legacy "three days a year" brands cannot credibly copy.
"Some of the number one events for decades are going to be losing their position soon, because they are not going to be able to change fast enough." — Marco Giberti, Founder and CEO, Vesuvio Ventures
It is the innovator's dilemma, playing out on the show floor.
If participants, exhibitors and organizers all deploy agents, those agents need one place that agrees on the facts: who is registered, which sessions exist, where a meeting is happening and at what table. That is the system of record, and it is where Tim announced Grip's next step: the Grip MCP, which lets organizers plug their AI agents directly into their event data.
Organizers can interrogate how an event is configured, compare it against previous events, monitor it continuously and surface the actions they should be taking. Early projects go further, generating fully personalised outreach web pages on the fly from Grip registration and insights data. Registration for the MCP beta is open on the Grip website now.
Three moves emerged from the conversation. Start capturing proprietary data now, because it compounds and no model can replicate it. Automate the invisible administrative work first, using the march of nines to pick your battles. And stay flexible: experiment aggressively, but architect everything so you can swap tools as the landscape shifts.
The business of events is being rebuilt around the connections only live experiences can create, with AI handling everything else. The organizers who treat that as an opportunity will own the next decade.
This post covers the highlights. The full discussion, including Ben's live demo of his AI sponsor discovery system and the audience Q&A, is worth your time.
No. Every one of the 12+ industry leaders interviewed for Grip's upcoming report believes in-person events are here to stay and becoming more important. As AI-generated content floods digital channels, live events act as a trust accelerator.
Start with session transcripts and connection data. Transcripts fuel content and validate future programming. Connection data, from badge scans to click-to-connect tools capturing an average of 20 connections per person, enables AI-driven gap analysis of your event's network.
It depends on scale. Small organizers can build boldly and experiment aggressively. Large organizers gain more from efficiency layers across their portfolio. Either way, avoid locking into any single tool: the AI landscape produces a new winner every week.
Low-value administrative work: manual data entry, contact enrichment, basic research and post-call notes. Choose processes where 90% accuracy is acceptable, requiring no behaviour change and no new data entry from your team.
Expect premium ticket prices of $15,000 to $20,000 for new high-value events, year-round monetisation through subscriptions and online education, AI-powered audience expansion into dormant databases, and AI agents becoming the primary channel through which participants discover events.
