In many instances, Triggers are set to fire immediately based on very specific criteria. There may be times however, that you may want to use suggestion to identify times where a Zing may be needed (when intent is ambiguous.) 

Example
Let's say a country club wanted to set up a Keyword Zing to automatically respond with the club's hours of operation (8-5 M-F) based off of the keyword of "hours." We would use the following trigger to automatically respond to any incoming texts that equal just the word "hours" as we'd assume that these contacts were inquiring as to the operating hours of the club. 

We could also add a secondary trigger with the contains filter instead of equals. This would then fire the message to go out any time a member inquired of any hours. Consider these questions that may come in:

  • "How many hours are the tennis lessons for?"
  • "What are the hours for the restaurant?" 
  • "How many hours does a round of golf typically take?" 

With our contains trigger live, each of these questions would immediately yield the answer "Our hours of operation are Monday to Friday, 8 AM to 5 PM," clearly the wrong answer to all of those questions. This then is the perfect time to use a Zing suggestion by changing run to suggest.

This changes the behavior of the Zing to prompt users with an option to run it if the keyword matches. 

As shown in the image above, the interface now offers a dialog box to run the Zing if appropriate. Users logged in to the service can opt to Run the Zing or select 👎 to decline.

AI Intent Recognition

Note: This is currently a premium trigger for Hospitality customers only.

This triggers when intent is recognized for an inbound message using Zingle's Artificial Intelligence (AI) engine. Zingle has trained an AI bot to recognize a few hospitality focused "intents". Many Zingle customers currently use "keywords" to recognize intent; however, using this trigger will be more accurate because it uses AI and NLP (natural language processing).

Instructions:

  1. Select a "confidence level". The Zingle AI model is built to err on the side of caution and is conservative by design; however, we understand our customers may want some control over the confidence thresholds.
  2. Select an intent. As of the time of publishing, Zingle supports the following intents:
  • Broken - Reporting something is broken
  • Bugs - Reporting of bugs
  • Checkin - Requesting checkin time info
  • Checkin - Requesting early checkin
  • Checkout - Requesting checkout time info
  • Checkout - Requesting late checkout
  • Cleaning - Request room clean
  • Cleaning - Requesting cleaning schedule
  • Extension - Requesting stay extension
  • Gym - Requesting gym info
  • Info - Requesting hotel info
  • Lost - Reporting of lost item
  • Luggage - Requesting hold luggage
  • Parking - Requesting parking info
  • Plumbing - Reporting of plumbing issue
  • Room - Requesting item to room
  • Room - Room ready request
  • Taxi - Requesting a taxi
  • Thank you - Guest saying thank you
  • Wakeup Call - Requesting a wakeup call
  • Wifi - Requesting wifi access

3. Select to automatically run the Zing when the intent is recognized, or choose to "suggest" the Zing. Suggesting a Zing will automatically pop up a dialog box above the reply input (in the Inbox), giving the Zingle user the option to run.

For more information around what is possible with AI, reach out to your Customer Success Manager.

Did this answer your question?