Exhibition and event organizers are far too familiar with the pre-event deluge of repetitive questions from attendees and exhibitors that can overwhelm their team. One association discovered a solution that reduced its staff’s workload by 80% while simultaneously improving the user experience: implementing a chatbot independently.
Let’s take a closer look at the real-use case study presented by Jessica Lawshe, Director of Digital Transformation at PMMI, The Association for Packaging and Processing Technologies at a recent IAEE Show Organizer Insights Hour. We will explore the practical steps, considerations and best practices that can help your organization successfully implement a chatbot from the ground up.
Understanding the Value Proposition
Before diving into implementation, it’s worth understanding why a chatbot makes sense for your event. For attendees, the frustration of long wait times for basic questions about schedules, directions or amenities disappears. Instead, they receive instant answers and can even get personalized recommendations for relevant sessions and exhibitors based on their interests.
Exhibitors benefit significantly too, especially first-timers who often struggle with logistics. A well-designed chatbot can answer questions about move-in times, shipping deadlines and booth setup while intelligently suggesting relevant upgrades. For instance, when the chatbot identifies it is interacting with an exhibitor, it can highlight premium offerings that align with their needs, driving additional revenue for your event.
Perhaps most importantly, your staff can focus on strategic work rather than answering the same questions hundreds of times. A chatbot addresses this pain point directly.
Laying the Foundation
Success starts well before you build anything. Begin by conducting a thorough website audit to ensure your information is accurate and well-structured. A chatbot is only as good as the data it can access, and a poorly organized website will lead to frustrating user experiences.
Next, think beyond your public-facing website. Gather additional information sources that would be valuable but aren’t necessarily published online. Work with frontline staff to compile frequently asked questions, import detailed exhibitor and session lists, and include practical information like parking locations, nearby restaurants and badge pickup details. If your industry uses unique terminology, add those definitions to help the chatbot serve as a comprehensive resource.
When preparing your data, format matters significantly. Use CSV files with clear, consistent headings such as “Exhibitor Name” versus just “Name.” When possible, convert documents to PDF format rather than Word. AI systems can extract data from PDFs much more cleanly because Word documents contain embedded styling and metadata that complicates the process.
Building with Intention
The heart of your chatbot lies in its system prompt, which provides instructions on how to act and respond. Be specific about what the chatbot should and shouldn’t answer. Configure it to only address questions related to your show and to politely decline inappropriate or off-topic inquiries. Implement detailed security guardrails from the start.
You have flexibility in how you add data sources. The chatbot can scan your entire website or just specific URLs, which is useful if you want to incorporate information from your venue’s parking page, for instance. Consider connecting a Google Drive folder that automatically feeds new documents to the chatbot, though note that edited documents may only update during scheduled recrawls.
Configure lead collection thoughtfully, whether by requiring contact information upfront, using keyword triggers or creating workflows that know when to ask. Consider enabling a hybrid search option that allows the chatbot to pull information from the internet for questions about airports or directions (but add extra guardrails to maintain control).
Testing and Refinement
Do not launch without thorough testing! Ask the chatbot numerous questions and correct its responses to train it to understand your brand voice, tone and preferred response style. Create a shared Excel document where internal staff can document questions asked, responses given, corrected answers and notes. This collaborative approach is especially valuable when the person building the chatbot isn’t familiar with every detail of your event.
Start with a targeted beta launch on select website pages, making it easier to monitor accuracy and fine-tune responses before going live.
The Payoff
PMMI saw remarkable results: 80% fewer calls and emails to staff, with the average attendee question resolved in under 30 seconds. Both attendee and exhibitor satisfaction increased measurably.
When getting started, many platforms offer free trials of seven to 14 days. Use this time to test features and find the right fit. And if you get stuck on system prompts, don’t hesitate to use tools like ChatGPT to help you craft initial instructions since you can always refine them as you go.
The key to success isn’t perfection from day one. It’s starting with solid preparation, testing thoroughly, and continuously monitoring and adjusting based on real user interactions. Your future self – and your team – will thank you!
IAEE’s Community Insights Hours bring like-minded members together for a think tank tailored specifically to their interests and needs. Learn more here. Not an IAEE member? This is just one of many benefits you don’t want to miss out on! Get in on the action here.
