Saturday, March 05, 2005

Document Genres - The Hidden Workhorse of Information Architecture (Peter Merholz)

Session Description

Slides (PDF) [via Peterme's follow up on his blog, 3/10/2005]

My thoughts: creating _meaningful packages of content_ based on _groupings_ of metadata, rather than focusing on individual metadata
- packages of content that serve different uses (e.g., book, guidebook, map, listings)
- use suggested by content and _form_

what's the difference between these genres
- form, structure

making explicit the genres of info we're offering so people know what to do with them
- what affordances can we provide?
- what cues

- "genres emerge as a response to purpose"
- genre is a _communicative action_, can be book, video, prezo
- digital doc genres akin to content types

- content inventory -> content map (e.g., janice fraser's work on PeopleSoft docs)
- challenge: what is most meaningful level of granularity to define genres?
- mapping content genres again user tasks -> leads to IA

- genres serve as important "trigger words", pretty much all a user has to figure out where to go

good examples
- TrendMicro

document genre not= content type (doc type?)
genre not= template
- genres drive templates (layout)
- genre also include content and purpose

separating presentation from content
- but for genres, presentation is key to communicating what to do w/ content

*-content as object, the "thingness" of content, the shape of info
*- genres are medium/device/channel-specific: maynot be easy to just pour content once, and automatically reuse it over multi-platforms

*- evaluating channels (PC, PDA, mobile phone, paper, phone call) by _genre needs_: portability, interactivity, detail/depth, multimedia, familiarity/trust, reflection, responsiveness, multitaskability, solidity of record -- then evaluating the importance of those qualities for a task (e.g., buying a house)

- meaningful divergence vs convergence (one size does not fit all)

- genres are fluid, innovation occurs when developing the right genre, in the right context, for the right medium
*- IAs need to think about how people _use_ use info, not just how to find it