In the world of fantasy films, a character might discover a mysterious lamp and release a grateful genie. This genie then grants three wishes, each with its own positive benefit and negative consequence. Despite this, the individual doesn't hesitate to make all three wishes. This captivating story encourages contemplation about the power of wishes, their consequences, and the importance of cautious decision-making. You might wonder how this concept relates to the real world, technology, and hospitality.

Modern society increasingly relies on technology to overcome challenges and enhance productivity. This is why everyone is excited about the promise of artificial intelligence (AI). It stands out as a new and powerful innovation with the potential to disrupt industries. Generative AI and ChatGPT can be considered the new genie, but with a significant difference: The wishes, or "inputs," are unlimited, and the granted wish takes the form of a displayed output.

It’s All About the Prompt

AI’s efficacy, much like that of a genie, heavily depends on the thought and precision put into crafting the input prompt. A poorly worded prompt may lead to irrelevant responses that fail to satisfy the person posing the question. For the best output, a prompt should be well-thought-out, clear, and specific, addressing the desired outcome with relevant and specific information. The prompt should use natural language that’s easy for the AI to understand and provide the necessary background information to help it generate an accurate response. Lastly, the prompt must include clear instructions on how the AI should approach it. Poorly worded prompts, much like poorly worded wishes, yield unsatisfactory results.

Here are a few examples:

Vague prompt: "How can we improve our hotel?" This is too open-ended and doesn't provide enough direction or specific areas for improvement.

Incomplete prompt: "I need to increase revenue." This one doesn't provide enough details or context, such as the specific revenue stream or target market.

Confusing prompt: "How can we make our guests happier?" This is subjective and might not be something that ChatGPT is equipped to answer.

10 Ingredients for Perfect Prompts

To achieve the best output, a prompt should encompass up to 10 components:

Intent: Clearly state the intention of the prompt and what the user wants ChatGPT to address for a relevant response.

Clarity: Make the prompt specific, clear, and unambiguous to prevent confusion or irrelevant responses.

Details: Offer relevant, specific information to help ChatGPT generate a helpful response that addresses the desired outcome.

Personalization: Use natural language and provide context to make the prompt more personal and engaging, improving user experience.‍

Language: Use natural language that’s easy for ChatGPT to understand. This will help create accurate responses.

Context: Offer relevant background information to help ChatGPT comprehend the prompt and produce more accurate responses.

Tone: Adopt a tone suitable for the intended audience and context, while ensuring the prompt remains professional and respectful.

Set up AI thinking: Give clear instructions or guidelines for how ChatGPT should approach the prompt. This will help it generate more relevant and useful responses.

Length: Keep the prompt concise and focused. This prevents ChatGPT from being overwhelmed and helps it create a helpful response within a reasonable timeframe.

Avoid bias: Refrain from making assumptions or judgments that could introduce bias into the prompt. Be aware of potential biases in the data that ChatGPT uses to generate responses, ensuring the prompt remains fair and unbiased.

Sample Prompts for Different Teams

These examples of well-crafted prompts for various hotel departments show why it’s important to instruct the system on output generation. Including several examples of what the output should look like lets the AI follow your guidance, replicate the described format, and generate the output to the best of its ability.

MARKETING:

"As a hotel marketing team member, can you recommend a marketing strategy to increase bookings for our new property? Specifically, what messaging and channels will resonate with our target audience, and what unique value proposition can we leverage to stand out from competitors? Please use data-driven insights based on guest demographics and behavior to inform your recommendations. Please provide guidance on how to measure and track the strategy’s effectiveness. Please be mindful of potential biases and ensure the strategy is fair and equitable for all guests."

OPERATIONS:

"As a hotel operations team member, please provide recommendations on improving our hotel’s check-in process. The goal is to reduce wait times and improve guest satisfaction. In the recommendations, identify the technology, staffing, or process changes we should make to streamline the process. Also, identify the specific metrics we should track to monitor progress of the recommendations. Please provide context on what has worked well in other hotels or industries and how we can adapt those best practices to our situation. Any identified process must be fair and equitable for all guests. Any potential biases generated should be identified and addressed in the output."

SALES:

"As a hotel sales team member, can you recommend strategies for increasing group bookings and maximizing revenue during low occupancy periods? Specifically, what value propositions or incentives can we offer to attract group bookings, and what channels should we use to reach out to potential clients? Please provide data-driven insights based on guest behavior and trends to inform your recommendations. In addition, please provide guidance on how to measure and track the effectiveness of the strategies and how to adjust them based on performance. Please identify any potential biases and then ensure that the strategies are fair and equitable for all clients."

The only missing component from the above prompts is instructing the system on how to generate the output. You can include several examples of what the output should look like, and the AI will follow your guidance. This could be the style of the report, press release, and marketing campaign. The AI will follow the described format and replicate it to the best of its ability.

Can It Really Work for Hospitality?

At first glance, AI’s highly impersonal technology might seem ill-suited for the highly personal hospitality industry, which revolves around creating and delivering personalized, warm and welcoming experiences. However, when used correctly, generative AI has the potential to transform the way hospitality businesses operate and make them even more personal, from streamlining operations to providing personalized experiences for guests.

Prompt engineering and writing prompts are just small aspects of how generative AI can transform hotels and brand teams to drive more revenue, reduce costs, expedite projects, and scale faster. While properly writing a prompt is crucial, much like contemplating a wish, the same principle applies to an organization's overall generative AI strategy.

Developing a comprehensive AI strategy is quite challenging as it involves multiple steps. You have to document and understand the business processes both internally and externally. Then you must analyze and make recommendations on how to apply the available functionality to solve problems, save money, and generate revenue. The proposed AI strategy must be adaptable to withstand changes in the technological landscape and evolving customer expectations.

Since this technology is new, it can be difficult to determine where to begin. Bringing in a fresh perspective from outside the organization can help identify blind spots or suggest innovative solutions. Through collaboration and idea sharing, hotels and hotel brand teams can find faster ways to drive more revenue, reduce costs, and scale more effectively.

Implementing a successful AI strategy isn’t a one-time project. Any AI implementation must be monitored and optimized. As technology evolves and your organization's needs change, the AI strategy must remain adaptable.

Crafting a prompt that delivers the desired output requires time and effort, much like making a wish. The more thought and time invested in the input (prompt), the better the results (output) and, ultimately, the wish itself!