AI for
Hospitality
AI is transforming hospitality. However, you need to ensure that any implementation doesn’t put your business or guest experience at risk.
Read the request
Send it to the right team
Record the result
Guests get an answer without staff searching several systems.
Where AI helps on site
In our experience, the best use cases combine automation and human-in-the-loop processes to automate repetitive tasks at each site while retaining control of any exceptions. Here are just a few examples:
Voice booking: Confirming standard requests through a booking or property management system and flagging any it can’t action for site staff.
Forecasting revenue: Taking account of local events, comp set, school holidays and weather to make pricing recommendations directly or as an input to your RMS, such as IDeaS or Duetto.
Reducing no-shows: Sending reminders alongside cancellation terms, measuring response and managing any resulting complaints.
Managing reviews and feedback: Compiling comments from different sources, grouping by common themes, suggesting responses for approval and monitoring trends.
Streamlining back-of-house operations: Suggesting rotas based on predicted demand, organising waste records or flagging likely allergen queries before they happen.
What a generic solution misses
However, implementing them successfully takes more than a generic AI solution. The process needs to comply with applicable legislation such as allergen safety under Natasha’s Law and other compliance obligations, escalate queries such as health, billing, or technical issues and integrate with existing processes. For example, no language model can determine if an individual can safely eat a dish, so an allergen request needs to be referred to site staff and answered from trusted and regularly-updated records.
Likewise, while language models are improving, they’re not perfect. So any solution should be tested and monitored to ensure that it can recognise queries that are out-of-scope or contradicted such as an old menu item, provide current and consistent answers such as cancellation or deposit terms, escalate queries to the right team such as group bookings, alert teams to serious issues such as a sudden increase in complaints and obtain approval before taking certain actions such as amending rates.
The operations review
Which is why we propose starting with a review of operations. During the review, we’ll:
Demonstrate common workflows. We’ll explain how they could help your team and how you’d manage exceptions (for example, GMs can justify rates, finance can spot common costs, allergen queries get passed to food-safety owners and ingredient lists for food-prepacked-for-direct-sale highlight allergens and list every ingredient).
Agree the relevant compliance and safety requirements. Some may already exist, such as excluding card data from prompts using a hosted payment link, while others need to be documented, such as processing certain health-related details that could be a special category of personal data requiring a DPIA and an appropriate lawful basis to do so. We’ll scope compliance into the final solution.
Perform a one-site audit. We’ll interview the responsible person, examine call logs or equivalent and connect to relevant systems. The fixed-fee exercise helps us understand how common queries are handled and identify gaps that could be addressed by AI. For example, a single restaurant could struggle to manage reviews but wouldn’t benefit from forecasting if their booking system already met their revenue management needs.
Build a first workflow at a single site. We’ll identify opportunities for improvement, agree escalation and privacy requirements, configure a solution, agree any required approvals, measure current performance, test the solution, log answers and iterate.
Plan a solution across your estate. We’ll identify sites that would be representative to test and consider local idiosyncrasies to ensure we’ve covered them.
Prepare a proposal for further work. We’ll be clear about any one-off implementation fees split by workflow and site, third-party costs, support options and how we measure performance.
Get in touch
Are any of the use-cases above relevant to your business? Do you know where AI could improve processes within your estate? Could it save money, increase revenue or improve your reputation? Book an operations review today to find out.