3 weeks ago - last edited 2 weeks ago
“This tool is so much more helpful than running reports and feeding them into another AI program. For example, I wanted to see sales numbers comparing Oct 24/25. I realize those numbers are easily accessible in reports, but then I told it we opened an extra day in April 2025 — what was that impact? It responded easily with numbers, percentages, and divided it into categories driving higher sales. We added DoorDash in 2025 — what was that impact? Again, it easily provided numbers and percentages.”
— @tinathom
3 weeks ago
I have been dipping my toe in the waters of the Toast IQ portion of the software. I have been enjoying the ability to take a look at what major events and weather trends are playing a role in sales that I am seeing in the restaurant. Im looking forward to the insights others have in regards to how they are using what seems to be a really powerful addition to Toast!
3 weeks ago - last edited 3 weeks ago
Hello Friends,
First, happy Thanksgiving! The season of giving and spreading kindness!!!
Thanks, Rob, for initiating the dialogue, a great way to get engagement and ideas bounced from the user community!
ToastIQ is an excellent addition to the Toast ecosystem. I actually built something very similar about six months ago because my wife didn’t like having to open the ToastNow app and manually change dates and filters just to see daily sales or slightly more detailed views. Now that ToastIQ is live and functional, I’m excited to retire my homegrown solution (screenshot below) soon. The ability to interact with our data using simple, natural language is genuinely invaluable.
That said, I do have a few suggestions that I believe would make ToastIQ even more powerful and user-friendly:
Session history retention
When I navigate from ToastNow to ToastIQ via the bottom prompt, then go back to ToastNow and return to ToastIQ, my previous query and results are cleared. It would be great if the query history and results were retained at least within the same session, as ChatGPT does with conversational context.
Command/query history
A visible command history would be very helpful. Being able to quickly rerun or edit previous queries instead of retyping them from scratch would make day-to-day use much smoother. (See my sample below)
Preconfigured and adaptive queries
I’d love to see some preconfigured “quick queries” at the bottom of the interface, ideally adapting over time to my most commonly used questions. This kind of self-learning shortcut bar would make repeated interactions much faster. (See below in the example screenshot)
API access
An API interface for ToastIQ would be fantastic. I want to be able to send queries and retrieve results programmatically (e.g., from a command-line tool or scripts) so I can integrate ToastIQ insights into my own workflows.
Reward redemption details
Reward redemption queries don’t yet give exactly what I need. For example, when I ask:
“Give me the total number of orders, total reward dollars redeemed, and the total sales from those orders.”
ToastIQ returns the total orders and total sales generated, but not the total reward dollars redeemed. Having all three metrics in one response would be very useful.
Top customers by reward balance
I’d like to query something like: “Show me the top 5 customers with the highest reward balances.” This currently appears to be unsupported, but would be a valuable feature for targeted outreach and loyalty management.
Reward card balance queries
Similarly, the ability to query and report on reward card balances would add significant practical value.
Ideally, we shouldn't need to go to the orders hub and make ToastIQ my main interaction. The ToastNow app would serve as a secondary interface, and most usage would migrate to the natural-language query versions.
Overall, ToastIQ is already a strong step forward, and with a few enhancements like these, it could become an indispensable daily tool for operators like me.
my 2c 🙂
Please play with it and give feedback to make it even better! You'll love it.
Cheers, Ashok
a week ago
@ashokr how have you noticed the accuracy of the data on some of the requests? I had some prompts I used to ask about how many of certain items were sold in a time period, and they seemed off. When I double checked them with the product mix, I saw a large discrepancy with the data I got off of the AI versus checking the product mix.
a week ago - last edited a week ago
You are correct, I have noticed that discrepency too. It seems to match sometimes, but also be wildly off next time. I would suggest you take a screen shot of the correct and wrong result so developers can know to track this down. Data is key 🙂.. but I should have captured.. was in a rush when I noticed and didn't get to it after.
But to their credit they did say "Answers provided by AI can be wrong" 🙂