liveBook · Manning

Data modality

As explained in Effective Conversational AI

The communication mode through which a user interacts with a conversational system, such as text, voice, email, or push notification.

A user setting up a printer asks where to find a latch. Repeated text directions leave them searching; a video can show the exact part. The problem is not just what the assistant knows, but how it communicates the answer. section 9.2.3

Data modality is the mode through which a user and virtual assistant communicate. Text, voice, visual interfaces, and combinations of them shape what users can provide, how they receive help, and how easily they can act on it. Modality is also part of the context an assistant can use to personalize an interaction. section 9.2 section 9.1.2

Choose the mode around the user’s task and situation:

  • Text supports precise answers that users can skim, revisit, and reference, though typing can be slow. section 9.2.1
  • Visual elements such as buttons, images, and video can make choices or physical steps easier to understand, but need a well-designed interface. section 9.2.1
  • Voice helps when hands-free use matters. Keep spoken answers brief and clear; noise and varied speech patterns can make recognition harder. section 9.2.2 section 9.2.4
  • Multimodal interactions combine modes—for example, a spoken request followed by a visual result—when one mode alone is less useful. section 9.2.1

Show the step instead of repeating it

For printer setup, text instructions fail to resolve a question about the cover latch. A visual assistant offers a setup video, then returns to the cartridge-installation segment when the user asks. The user can inspect the step rather than continue a frustrating exchange. section 9.2.3

Keep voice help useful

A flight-status voice bot that sends a caller to a website for the gate number adds effort. A better interaction gives the status and gate, offers directions, and asks whether to send a map to the user’s phone. section 9.2.3

Evaluate each flow in its intended mode. Check whether users can understand the response, complete the task without avoidable back-and-forth, and move to another mode when that would help. Adapt the flow as those observations reveal friction. section 9.2 section 9.2.3

How Effective Conversational AI covers it

How deep it goes
Defines four interaction modes and compares their strengths, limits, and design implications without giving implementation code.
Examples
Contrasts text and video printer guidance, then poor and improved voice interactions for flight status.
What you can do after
Evaluate/understand it by choosing a suitable interaction mode and spotting avoidable friction in an existing assistant flow.
How it connects
Connects modality with contextual information, personalization, accessibility, voice recognition, and multimodal interaction.

Also explained in

Compare how 4 books explain data modality →

Related concepts

This explainer was generated by Manning's knowledge graph from the text of Effective Conversational AI. The section links point to the parts of the book it draws on.