There are two ways to start a chat in Ask TrustYou: ask a free-form question or run a preset. Once a chat is open, you can rename it, download it, share it, and track its status from the History section.
The Start Screen
Click Ask TrustYou in the main navigation; you will land in the Chat tab. The start screen has two sections:
- Ask anything about your reviews: A free-form input where you type your question directly.
- Use a preset: Preset cards for frequently run analyses. Click All presets to browse the full library.
Your monthly chat usage is shown in the top-right corner of the page. Click More to open the AI usage page in Settings.
Setting Your Data Scope
Below the input field, you will see your current data scope displayed as three filter chips: Date, Source, and Entity. Click any chip to change the filter:
-
Date: Select from a list of predefined date ranges or set a custom date range. The earliest available date is January 1, 2021, or later if your plan's review limit is reached first.
-
Source: Select one or more review sources, or choose all.
-
Entity: Choose the entities in your scope to include. You can use rules, manual selection, or a combination of both. Maximum selection: 6,000 entities.
The data scope cannot be changed after you send your first message. If your selection exceeds your plan's dataset limit or returns zero reviews, the chat cannot start. Narrow your entity selection or date range until the count falls within range.
Starting a Free-Form Chat
Set your data scope, type your question in the input field, and click Send. The chat opens, and the AI begins analysing your reviews. For tips on phrasing, see Writing effective questions in Ask TrustYou.
Running a Preset
Click a preset card in the "Use a preset" section to run it directly. If the preset requires variables, such as a brand name or time range, you will be asked to fill them in before the chat starts. To browse the full preset library, click All presets. Each preset run uses one chat credit. See Presets for more.
During a Chat
Each chat gets a title generated automatically after your first message. You can rename it at any time using the options menu (⋮) in the chat header.
Answers can come back as a short summary, a table, or a chart, depending on your question. You can also ask for a specific format, like "Show this as a table." To save a chart, use the download option on the chart itself.
From the options menu, you can also:
- Download the chat as a PDF.
- Share a link to the chat with a colleague.
- Rename the chat.
- Exit or delete the chat.
To review the data the AI is working with, click the ⓘ icon in the header. This opens the Data scope panel, which shows your filters and the types of data the AI can query.
Note: AI might make mistakes. Always double-check important responses before sharing them.
Chat Status Messages
You may see one of the following messages in the chat footer:
- This chat has run its course. The chat reached its 10-message limit. Extended exchanges reduce AI accuracy, so starting a new chat is the best way to continue.
- Waiting to start. System demand is high. Your chat has been placed in a queue and will start automatically when capacity is available. While waiting, you cannot send messages. Once the AI processes your request, the chat will appear in your History with an Unread status. Any chat already open is not affected.
- Time for a break. Your team has started chats at a high rate in a short period, and throttling has activated. New chats cannot be started until the countdown finishes. Any chat already open is not affected.
- You've reached your organization's chat limit. Your subscription has used all its monthly credits. You can keep reading the current chat, but new chats cannot be started until the limit resets on the 1st of next month. See Monthly chat limit to learn how to get more chats.
- Something went wrong. A system error ended the chat. Start a new one to continue. It will not count against your monthly allowance.
Chat History and Statuses
To find, manage, and understand the status of past chats, see History.