IA Summit 12: Better Cross-Channel Experiences with Metadata

Another 20 minute session that I enjoyed was Adam Ungstad’s Better Cross-Channel Experiences with Metadata. He talks about how defining your data can improve communication by ensuring that standards are used. These standards then enable exchanges between all channels.


Metadata enables:

  • Consistency
  • Context
  • Interoperability

Metadata is information about information objects. There are three types of metadata:

  • Descriptive
  • Administrative
  • Structural

Descriptive data allows you to discover and identify objects. Examples of this type of data are: date, color, version, make, origin etc. Descriptive metadata enables context. In UX it allows things like faceted search.

Administrative data is used to manage information objects. Examples are: status, owner, access, property rights, etc. It helps you get the right context, at the right time, using the right channel for the right user. In UX it allows a content strategy. Administrative metadata enables consistency.

Structural data is the composition of compound objects. For example your name is really a combination of 3 objects, first name, middle name, and last name. In UX things like good form design come out of structural data. Structural data enables interoperability.

Information Architecture of Metadata

Moving between multiple communication channels requires exchanging information. There are 3 sides to the IA triangle for information exchange:

  • Presentation – How information is presented to the user.
  • Storage – How that information is stored.
  • Exchanges – How the systems exchange that information.

When you have two or more systems talking to each other using metadata standards enables those exchanges, which enables data linking.

Data Standards:

  • Semantics – The meaning of data.
  • Syntax – How the data can be expressed.
  • Lexical Rules – Rules that qualify the data.


A users is an information object, but experiences are not one size fits all. Experiences must be personalized. Each channel must know the user, the user isn’t the exchange mechanism. You should never have to ask the user for the same information twice.

Next Steps

  1. Make metadata a priority. Link it to ROI through reduced inefficiencies, and create loyalty.
  2. Define your data. Identify what object need to be defined by any object that: is part of, is delivered by, or is used by another object. Create your ontology using mental models, diary studies, and contextual inquiry etc. Discover what needs to be defined, then create the data standards.
  3. Identify where exchanges are happening. Use tools such as journey maps.
  4. Build the infrastructure. Start small, borrow and reuse.