Can artificial intelligence (AI) tell you which conferences to sponsor? Calculating an event’s return on investment (ROI) has never been easy, according to a Demand Metric survey, only 17 to 28 percent of companies currently track it. Yet the same report indicates trade shows take up to one-fifth of the average respondee’s annual marketing budget. From booth design to sponsorship to travel, trade shows are not cheap affairs.

AI can’t eliminate trade show costs, but it may help optimize them. New York-based SummitSync uses machine learning to predict whether the event you’re considering is financially worth it. SummitSync CEO and Co-Founder John Corrigan said,  “the $350,000 they were going to invest in CES [the International Consumer Electronics Show] in a sponsorship would have been a complete waste because CES had less than 10 percent of their prospects attending.”

How Machine Learning Predicts Conference Potential

According to Corrigan, the system uses machine learning to predict a conference’s potential to jump start deals by analyzing prior attendee numbers, sponsor lists and session content. SummitSync then compares this data against information from your customer relationship management (CRM) platform, “map[ping] what conferences, trade shows and events their clients and prospects are attending,” he said. “From there, the client can use that information to weigh the concentration of clients, revenue and costs of attending or sponsoring a trade show or conference.” In other words, the company is using machine learning to predict ROI.

It’s intent-based marketing, but specific to conferences. Previously limited to sales pipeline platforms like TechTarget’s Priority Engine, intent-based tools predict which customers actually want more information about your product. For example, Seattle-based startup Outreach forecasts meeting success while Cincinnati company NaviStone identifies consumer receptiveness to direct mail.

Learning Opportunities

Related Article: How Artificial Intelligence Will Impact the Future of Work

As to whether it works, Corrigan can’t offer client names for verification, claiming his clients’ investors would be upset to learn how much money they may have wasted at prior events. However, in addition to the CES customer, he says SummitSync recommended a programmatic advertising client not attend SaaStr, a software conference that would have cost the company $50,000. “[The event] just didn't match their target customer or partner despite the sales pitch,” Corrigan explains. Whereas event marketers stereotypically tell prospective exhibitors what they want to hear, machine learning only looks at data.

Additionally, SummitSync offers a post-event sales tracking dashboard, as does competitor Jifflenow. Corrigan says predictions are most effective for tech, media and telecom events with more than 1,000 attendees. Expect the software’s analytical abilities to extend to more trade shows soon, though. In a July A round led by Loeb Enterprises, venture capitalists invested $3 million in the company.