What makes catalog items non-conversational in premium chat
Summarize
Summary of What makes catalog items non-conversational in premium chat
Premium Chat in ServiceNow uses Now Assist to provide a conversational interface for catalog item requests, replacing traditional form filling with guided conversations. However, not all catalog items qualify for this conversational experience. Certain conditions and configurations determine whether a catalog item can be handled conversationally or must fall back to a traditional form.
Show less
Key Conditions That Make Catalog Items Non-Conversational
- Conversation Disabled by Property: If the
turnoffnowassistconversationproperty is set, the item is excluded from Premium Chat regardless of its design. - Explicit Non-Conversational Flag: The
makeitemnonconversationalproperty forces the item to open as a form, often used for complex items needing structured input. - Excessive Number of Questions: Items with more than the configured question threshold (default is 4, controlled by
snnowassistcr.og.questionthreshold) are considered too complex for a natural conversation. - Use of Client-Side Scripts: Scripts that dynamically change form fields (e.g., hiding fields or running calculations) only work in browser forms and are incompatible with chat interfaces.
- UI Policies: Rules that modify form visibility or behavior based on user input depend on browser rendering and cannot be replicated in chat.
- All Fields Hidden or Read-Only: If no questions are left for the user to answer, there is no meaningful way to enable a conversational flow.
- Unsupported Question Types: Certain complex question types like check box, custom UI pages, list collectors, attachment questions, and others are not supported in chat.
- Advanced Filtering Fields: Fields that allow users to pick from a filtered list of options force the item to revert to the traditional form.
- Pricing or Cost-Influencing Questions: Items where any question affects pricing cannot be handled conversationally as pricing support is not available in chat.
- Repeating Table Inputs: Sections allowing multiple rows of input (e.g., ordering accessories for multiple employees) are not supported in the Premium Chat interface.
- Complex Catalog Item Types: Order guides, wizards, content items, and standard change templates have complex structures that do not support conversational experiences and open as forms instead.
Practical Guidance for ServiceNow Customers
Meeting conversational criteria does not require rebuilding catalog items entirely. Often, minor adjustments such as reducing visible questions, removing client scripts, or changing field types can enable a catalog item to support the Premium Chat conversational experience. Understanding these conditions helps customers design or modify catalog items to leverage agentic AI capabilities effectively, improving user experience by enabling guided, conversational requests.
Let's understand what determines whether a catalog item can deliver a fully conversational experience in the Premium Chat interface.
- Conversation turned off using the turn_off_nowassist_conversation property, preventing the item from appearing in Premium chat mode, regardless of how the item is designed.
- The catalog item has been flagged to open as a form and not a chat using the make_item_non_conversational property. This is used for complex items where a structured form is preferred over a guided conversation.
- When a catalog item has more than the configured maximum number of questions (Using the sn_now_assist_cr.og.question_threshold property, which has a default value as 4), it can be complex to handle as a natural back-and-forth conversation.
- Catalog items use client-side scripts that change the form dynamically as the user fills it in. For example, hiding certain fields based on earlier answers, or running a calculation. These scripts only work in a traditional browser form and can't be replicated inside a chat window, so the item can't be made conversational while these scripts are active.
- Similar to scripts, UI policies are rules that change what requesters see on the form based on their inputs. Because these rules depend on the form rendering in a browser, they aren't compatible with the step-by-step chat approach.
- Every field on the catalog item is either hidden from the user, pre-filled with a default value, or set to read-only. Since there are no questions left, there is nothing for a conversational experience to do.
- Some questions types, such as check box, Custom, Render UI Page, Custom with Label, Lookup Select Box, List Collector, Lookup Multiple Choice, Masked, Table Question, Requested For, Attachment Question, Table Name aren't supported.
- Some fields let users pick a value from a list (for example, choosing a specific laptop model from a catalog). Items with this kind of advanced filtering fall back to the traditional form.
- If any question on the catalog item influences the cost or pricing of what is being requested, the item can't be handled conversationally. Pricing implications aren't yet supported in the chat interface.
- Some catalog items include a section where the user can add multiple rows of information. For example, ordering accessories for several employees at once. This kind of repeating, table-style input is not yet supported in a Premium chat interface.
- Certain catalog item types, such as order guides, wizard, content items, or standard change templates, have a structure that is more complex than a simple request. These types are not eligible for the conversational experience and open as a traditional form.
- Conversational catalog items that don't support agentic AI may be discoverable, but they don't appear within conversation in the Premium chat experience. These catalog items open as forms that requesters can fill out and submit.
- Meeting these requirements doesn't require rebuilding a catalog item from scratch. Often, small adjustments, such as reducing the number of visible questions, removing an unused client script, or switching a field type, are enough to make an item eligible for the full agentic AI experience.