The study explores trends in attendee preferences, factors in destination selection, and how incentive travel...
Research / AI: Uses and Possibilities for Incentives Professionals
by Incentive Research Foundation
The IRF recently conducted a series of focus groups with suppliers, third parties and program owners who shared their insights, practical use cases, and hesitation around the AI technology. While these discussions revealed that AI use among incentive industry professionals is still in its early stages, use of AI and interest in the technology are steadily increasing. This report explores how incentives professionals are using AI both in the travel and non-travel incentive arenas, information about the tools and programs they find most valuable relative to incentive programs, and the gains and impacts they’re experiencing as a result.
The topic of Generative Artificial Intelligence (AI) is all over the Incentives, Rewards, and Recognition industry. You’ll find that AI is a mainstage discussion or a breakout at nearly every industry event. The possibilities for application of AI within the industry are vast, from creative and rewards selection, to hyper-personalization and data analysis. To understand the use of AI today among industry professionals, we gathered a group of leaders including program owners, third parties, DMCs, hotels, and others to discuss their experiences with AI.
While AI solutions have been around in various forms for years and integrated into some of the tools we use regularly, our discussions focused on Generative AI solutions such as ChatGPT or Spark.AI that take multiple inputs and consistently learn, thereby improve output. These are the technological advances that have the industry talking about possibilities, engaging in experimentation, and discussing implications for our industry’s workforce.
source: ttps://www.eyerys.com/sites/default/files/ai-fields.png
This paper explores how AI is being used in programs today, some of the current challenges, and the state of mind within the industry from the standpoint of suppliers and program owners. The general sentiment of the group was that AI use among incentive industry professionals is still in its early stages. Most incentive professionals we spoke to who have explored AI are using ChatGPT or a similar large language model to increase the efficiency of their marketing, writing, and data analysis. Hesitation around the technology centers around privacy issues, accuracy, and training time.
For many, AI means the opportunity to tap into efficiency allowing their teams to focus on more strategic tasks. For others, it is seen as a shortcut for tasks that are not their core competency such as drafting communications or helping with data analysis.
Some view AI as a journey. It’s a change that many aren’t yet comfortable with in large part due to privacy concerns, including leaders of organizations in many cases. The journey is about understanding how this technology can aid and advance the workforce – augmenting and enhancing what they’re capable of delivering day-to-day.
“It’s about automating tasks that then relieve people from doing busy work. For example, we’re looking at implementing an AI solution to answer question after question that incentive program attendees ask. We had one coordinator answer 4000 emails that were all questions answered in the registration materials. Pick up, drop off, expenses. AI can quickly answer these questions pointing back to registration materials. It means that that person can do other things that the organization needs their brain power for; more strategic things.” – Corporate End User
“I look at it as a way to help if I’m stuck writing. Only used it a few times, but it helps me. I have a lot of caution because I don’t know where the information I put in is going. It’s important to keep inputs generic and not include any protected information.” – Industry Supplier
“I can ask it to help me communicate information in an improved way. They’re still my ideas, but the bot gets it there faster and helps elevate my intention.” – Third Party
Others view AI with hesitation. They see the potential benefits and are keeping an eye on it but have not yet really explored the possibilities. Others are worried about exposure and data privacy, so they are keeping their distance. One likened it to their adoption of search engines or email. It’s simply something to make a habit. All non-users noted that they recognize it is important to get past the hesitation, make time in their schedules, and start familiarizing themselves with AI.
Our discussions overall speak to an industry in transition when it comes to AI. Exploring the level of experience using Generative AI today, most focus group members consider themselves to be beginners with a few characterizing themselves as more intermediate users, saying they use it every day.
Those using tools like ChatGPT are using it for a wide variety of reasons, but most consistently point to content generation, some data and analysis, and working on itineraries. Most use is basic now, but the desire for more sophistication is there.
Some respondents had not yet tried using a generative AI tool. The make-up of the group of non-users included individuals across multiple age and demographic groups. This is consistent with our experience with a group of 70 incentives professionals at the IRF’s Leadership Insights Forum Event in 2023 where we had hands-on AI-based experimentation rooms. The assumption that younger generations would be quicker to adopt this technology seems to be incorrect relative to the incentives industry.
Hesitation in using AI is primarily privacy related. One way to manage some of the privacy concerns is to employ a private version of AI. Private Generative AI delivers accuracy, helps protect copyrighted materials, and blocks data sharing from public models, creating a more secure and controlled environment. Where a private AI instance is not available, users should absolutely take care to avoid uploading or sharing any proprietary information or Personally Identifiable Information (PII).
Individuals also cited the simple challenge of finding time to experiment with a tool like ChatGPT as a deterrent. Still others worry about what it means to human connections and critical thinking.
“I sound like a curmudgeon, but what are we losing in critical thinking in our quest for efficiency? It’s important to sit down and think about how to solve problems. I also wonder where the human-ness goes when we’re using technology like this. For us and our property, the human connection is so much of what makes us who we are.” – Hotel Supplier
Regardless, all agree that using and understanding AI is increasingly important to anyone in the industry. One respondent noted, “New technologies always create fear. But this is really less about adopting new technology – it’s change management.” The sentiment resonated with this group as those using it agree the same level of technological skill required to learn a program like Photoshop simply isn’t required for using most Generative AI tools. The skill required is more akin to having a conversation where you’re requesting specific outcomes.
A few respondents, primarily those working for larger organizations or corporations, indicated that their organizations had policies in place around the use of AI. Smaller organizations tended to have no policy in place, other than issuing cautions around uploading any documents with proprietary information or documents containing PII.
“So far the internal feedback / guidance is much more caution than use. PII is key. The few bits of guidance have been about protecting customers and proprietary information.” – Industry Supplier
According to the 2024 Deloitte Technology Trust Ethics study, to meet the ethical needs of emerging technologies, the most common roles organizations are planning to hire include AI ethics researchers (53%), AI compliance specialists (53%) and technology policy analysts (51%). This underscores the importance of company-wide alignment around use, disclosure, and application of this technology to our day-to-day roles.
This policy gap also applied to disclosure around use of AI for proposals, content, and more. And, for reference, Generative AI was not used to write this paper. There is some difference of opinion between whether it is necessary to disclose when AI is being used. Some are proudly disclosing to demonstrate adoption of the technology to their customers. Others believe clients expect them to be using it and it’s not important to call it out every time it’s being used.
“Getting a lot of questions [about how we are using AI] but no specific directives because they want to know how they can use it. They recognize the opportunity for efficiency but it feels like we need to use it to become a better provider for our clients. I think most people assume it’s a part of the process (but that may be a big assumption). We don’t use it for the sensitive data.” – Third Party
Program owners also vary in their expectations around disclosure for partners and suppliers using AI for ideation, program design ideas, and more.
“Our budgets are tight and when we’re paying a third party, I expect them to be as efficient as possible with the time I’m paying for… I expect them to know more than I could.” – Corporate Program Owner
According to the Deloitte 2024 study, a higher percentage of organizations with revenues of $1 billion+ are taking steps such as hiring AI talent, delivering training, and providing internal tools for their employees (64%) vs. smaller organizations (36%).
But interestingly, smaller organizations are outpacing larger ones when it comes to developing guidelines or policies regarding the ethical use of AI.
“We have a risk and responsibility training module to make sure there is a human in the mix. Making sure that we don’t lose sight of ethical and regulatory pieces.” – Corporate Program Owner
Within the incentives industry, it is important for suppliers and their clients to come to agreements and have up-front discussions to ensure expectations around the use and disclosure of use of Generative AI related to their programs. Without these explicit conversations, suppliers set themselves up for potentially difficult and awkward situations with clients. In fact, suppliers have an opportunity to establish clear policy for their organizations and set standards for their employees to avoid any conflicts with clients regarding AI use and disclosure.
As program owners begin to incorporate AI into program design, there are many possibilities being explored. Corporate program owners understand the challenges they face as the workforce shifts, the desire for personalization and more custom rewards continues to increase, and leadership increasingly looks to them to provide data demonstrating the effectiveness of the investment in incentives, recognition, and rewards.
Utilizing AI to identify rewards structures that can be motivating and effective for a group using data from past programs is something AI can do. Some are using it today.
One10, a performance improvement provider, has leveraged “explainable AI” for its customers to improve the design of incentives. A global information technology distributor that wanted to analyze its incentive promotions targeted at channel partners engaged One10’s Predictor Model. By scouring sales and product data, AI predictive models were provided to the client to guide them in which incentives most impacted the revenue, and which were lower performing. This explainable AI technology not only improved the accuracy of the client in house analysis but also allowed the customer to reallocate its investments in the most effective incentive promotions resulting in millions of dollars of improvement.
Where that data is not available, for example when pursuing new business, savvy third parties are asking AI questions about the organizations and industries they are pursuing. By asking about the performance of the company and industry overall, key competitors (and any information about incentives they may offer), pending or recent regulation, average profit / price points, the rules structures and incentives recommended are much more likely to be effective for the client.
“From a new customer perspective, agencies need to know our business better. They’ll present rules engines that don’t reflect our demographic or our business very well. For example, they’ll present rules engines that are more for a pure sales org vs. a channel structure.” – End User
Third parties are also experimenting and embracing the technology to help address diversity, equity, and inclusion objectives, and recommend program adjustments based on data from past participants.
“Diversity and sense of belonging is the key use for my team. By using specific prompts you can get 80% there. I’m searching for what I’m not thinking about relative to belonging and inclusivity. So I can run my program through asking if there is anything in the program that could be difficult for someone who is differently-abled. I might not get everything, but the AI often returns results that I may not have thought about.” – Third Party
Communications seem to be a focus for all involved. The group collectively appreciates the power of Generative AI to deliver draft communications for program rules structures, updates, and launches. But everyone is clear that communications must have human oversight, review, and intervention before they are utilized.
“Our company has an internal version of Chat GPT. We have built in our corporate guidelines and voice / tone. So we can run all website content through that to get the copywriter much farther down the road. It is saving us time.” – End User
Consider exploring the following areas using Generative AI:
The uses for Generative AI within incentive travel programs are exciting and exploration is underway by many of our respondents. Program communications, brainstorming for themes, and writing session descriptions are the most common uses noted by third parties.
Some third parties have started to incorporate AI into the process around proposal responses and destination copy. This is where we heard clients saying they are perfectly fine with this approach, as long as they can’t find the same program or experiences themselves using ChatGPT or a similar tool, and agencies are using protected versions of Generative AI when responding to RFPs to protect their privacy.
“Clients are asking for ‘wow’ options. Experiences people cannot do by themselves or ‘anything I can find on chat GPT.’ Clients in general are not specifically asking, but we expect they will increasingly be curious and specific about it.” – DMC
Clients expressed a desire for the agencies to continue to innovate and bring content that ChatGPT and the like have not yet been exposed to. Even if an agency can’t find something unique, there is some upside. As one client noted, agencies can show their value by being prepared to discuss why they would or would not recommend something based on their experience.
“If you use ChatGPT to find unique things to use in the Bahamas and I did same, you can now say “we used ChatGPT too and wouldn’t recommend that activity or activation because of X” – that establishes expertise.” – Corporate Program Owner
But program owners want to ensure that anything generated from AI is something they can have.
“For me, I don’t have an expectation that they are using it, but I do have an expectation that (particularly on DMC side) what is proposed actually exists. I want to make sure what I see exists and is not a reasonable facsimile. I want to know if it’s AI and not available to me for my program. If it’s an artistic rendering, let me know that it’s not exact.” – Corporate Program Owner
The suppliers we spoke with expressed some concern about program owners ultimately choosing to either slash DMC budgets or choose to take more planning internal based on the capabilities of Generative AI. They question whether AI will become informed enough to be able to find specific activities and experiences locally.
The program owners we spoke with certainly expect the suppliers to know more and deliver more than they can find on an AI platform to avoid diminishing their own value but did not see themselves going to a self-managed model for large-scale events and incentives. However, the challenge of small and simple events or incentives has plagued many companies for years. These smaller incentives do not often meet the minimum threshold for a third party. That leaves the program owner on their own to design and deliver. Generative AI may be a solution according to one corporate program owner.
“We use it for smaller events to create itineraries. I would never use Chat GPT and assume I got it right for my President’s Club trip. We use it only for things too small for a third party to touch. We’re solving for a gap – finding new venues that have opened, new experiences.” – Corporate Program owner
Suggesting and improving itineraries is another use for Generative AI. One third party is sharing itineraries with AI and asking it to optimize the education session times based on her audience profile. Again, this requires human review, but can help a planner think differently about the structure of an incentive. Some even ask the engine to review the program and make recommended changes based on published best practices and research.
Onsite, AI is readily available for things like push notifications in mobile apps, facial recognition in event photography, and writing emails or other communications to attendees to update them on any mid-program changes due to weather or other disruptions.
The internal nature of recognition programs opens up many possibilities for the use of Generative AI, particularly in organizations with internal AI engines. In those cases, program owners can run data through a walled-off system to gain insight into potential program improvements.
The potential for AI to positively impact employee recognition programs is exciting to program owners and third parties. From leveling the playing field when peers are writing award nominations, to determining the recognition styles of individuals, to communicating the program in a more compelling manor, AI is already being introduced in ways large and small to enhance recognition programs.
To help ensure employee-submitted peer nominations are more compelling, one company is suggesting their employees turn to AI for help crafting those nominations.
“We just closed the cycle for peer nominated awards. I have encouraged employees to use AI to strengthen their nomination write-ups. It levels the playing field for those that are not great writers. I have seen some of that pay off in the nominations being more equal.” – Corporate Program Owner
Another helpful use for Generative AI being used today by some of our respondents was identifying opportunities for program changes. For example, using Generative AI to create a survey to understand how people prefer to be recognized, and the scale of recognition that is meaningful to them.
“Our clients are interested in collecting employee preferences like how to recognize people publicly but they’re not about recognition in front of a crowd. I used AI to come up with the survey in the first place (including the scale of recognition). It’s also helpful to best communicate changes to programs.” – Third Party
AI can also be used to help managers write more meaningful thank you notes by putting a few key pieces of information about the task performed or goal achieved into ChatGPT or another tool, then asking it to write a thank you. Again, this is a place to start. The expectation would always be that the manager would review and refine the thank you note to ensure it is personal and conveys what they intend.
And for third party and tech providers, there is a host of opportunities from using Chatbots to help individuals better navigate recognition platforms to increase their use or include easy ways to remind managers of upcoming service anniversaries, remind leaders to use their monthly point allocations to recognize employees, and more. These adaptations will help recognition become more meaningful and more consistent in organizations.
Third parties and corporate program owners have been working through opportunities to refine their reward programs for years. From the early days of merchandise catalogs to today’s self-service redemption sites, the more personalized the rewards, the more impact they have on the recipient.
By asking a few simple questions, an AI-driven rewards engine could help serve up a curated selection of merchandise, helping to drive rewards redemption. This is particularly beneficial in organizations where points do not expire, so the organization continues to carry the burden of the impending points redemption. On a broader scale, consumer loyalty programs are working on these improvements as well to reduce the financial liability on their books associated with loyalty programs.
One program owner shared that their organization would benefit from more personalization.
“I would really like more personalization and understanding of how people like to be rewarded so you know if they want to be rewarded differently. We found 85% of our people don’t redeem their points. We did a push to see if people wanted to donate to a cause. The majority redeemed their points to support the cause, which was great, but not what was originally intended. So, we would benefit from a system that could use segmentation and personalization to help get it figured out and make it easy to use.” – Corporate Program Owner
The trick will be for these systems to advance beyond the basic “you also might like this” recommendations already available in some of today’s platforms. Third parties have an opportunity to use AI to analyze PII-protected redemption data to quickly refine award offerings and make recommendations for changes.
Measurement and reporting are areas where AI enhancements are already making an impact and are certainly something on most program owners and third party wish lists. A few shared how they are using AI today for reporting and measurement.
“We had a team impacted that evaluated our survey data. We don’t have those people anymore to support us, so we are using AI to run our surveys through to give us direction on sentiment, themes, areas of happiness or dissatisfaction. Also using it to ask things like ‘How do I take this data to put it into a chart for those not excel or PPT savvy?’ That is helping me get information to business partners faster.” – Corporate Program Owner
“I use it when analyzing data. For example, I don’t always know the right formula to get the data I’m looking for, so I can ask Generative AI “What’s the formula to extract this data set out of this data? Or I need to extract this code from this text, how do I do that?’ Then I can apply the result to the data set and get where I need to go faster.” – Third Party
Program owners and third parties are increasingly using Ai to analyze patterns and trends in performance of incentives against specific KPIs. That can help with rules structure analysis and help course-correct unintended consequences in how an incentive is structured.
“From a non-travel perspective, I’d love to start using Microsoft CoPilot with Excel to analyze data and program reviews for our sales and channel programs. It would make providing data to clients more efficient.” – Third Party
“I’d like to ask how to improve my incentive program that is structured in a specific way and adjust it based on available data and research, particularly in my industry.” – Corporate Program Owner
Other thoughts around how to use Generative AI include building more robust audience profiles based on past survey responses to understand the differences between past winners and first timers. Are there generational differences or different motivators that could influence how programs are designed and incentive travel programs are executed? It’s a real possibility to begin feeding historical incentive program surveys into an AI engine for deeper collective analysis. The caveat is, of course, ensuring the data does not contain any PII.
If you’re not using AI today, taking the first step and experimenting with prompts and how to have a “conversation” with AI is a great first step. You won’t need to upload data, simply use a few basic prompts and then continue the conversation to help refine the results. Try some of these examples below to get started.
“From the individual perspective, the concern I hear is that AI is a specific program they have to learn like Microsoft office or Cvent. But it’s really just an electronic discussion, not a big training or a system you have to learn. We gave our team an exercise where we gave them a destination and they had 10 minutes to come up with a theme, menu, gift, and itinerary for a program. It helped break the barrier.” – Corporate Program Owner
Remember to protect confidential data and PII when using Generative AI tools.
Thank you to the incentive industry professionals who contributed to this report:
Thank you to our Research Advocacy Partner
A use case for leveraging generative AI in brainstorming reward ideas for frontline hotel staff is explored using prompts and actual ChatGPT 4 responses.
During this IRF webinar, we examine key findings from our new report 2024 Attendee Preferences for Incentive Travel. Our panel discusses how the right destinations and good program design boost employee motivation. They review data from an IRF survey of North American employees, Cvent, and 3D Cruise partners, and talk about the top employee destination preferences…
