Information Architecture on a Large Scale
Presented by Dimitri Glazkov
at IPSA Meeting on July 11, 2008
- I am not an authority
- Just a guy who built a lot of large sites
- This is my story
- 1 hour?!
- Way to little time to talk about Information Architecture
- Will try to cover only a tiny subset
- Big points:
- Admit it's a losing battle
- Embrace organic growth
- Purpose is king
- This is not a rehash of the Polar Bear Book
- You should read the book though
- It's not perfect
- Not an instruction manual
- An inspiration
- What is Information Architecture?
- Mythical beast
- Wicked magic
- A balancing act
- Forces of a Web Project (diagram)
- Client
- User
- Technology
- Business
- Resources
- Organizational politics
- Proportions are rarely held at perfect ratio
- All forces want more pie
- Some have more leverage
- Some get sacrificed
- Information Architecture is a Misnomer
- Architects build houses
- Houses change very little over time
- Information architects build sites
- Sites change a lot over time
- Organic growth
- Changes to site that are outside of the original architecture
- Organization objectives are not homogeneous
- Spurred by org units attempting to pursue their (unsatisfied/unanticipated) Web needs
- Only select few (if any) are privy to the big picture
- Not because it's secret, but because it's so complex
- Those who are not privy are behind the organic growth
- Anthill
- Completely organic growth
- Planning is overrated
- Iceberg
- Begin from the top
- Lose interest/faith/money after a few first layers
- The rest grows organically
- Kudzu
- IA Project successfully completes
- Overrun by organic growth in the first few weeks
- Frankenstein
- Results of IA projects randomly applied and stitched together
- Don't fight organic growth
- Harness it by looking for internal Web needs
- Anticipate it by providing growth opportunities
- Stimulate it by encouraging user-generated content
- Don't look for perfection and completeness
- Perfect architecture is a mythical beast
- Aiming to create a complete IA of a site is the perfect recipe for disaster
- Instead, take care of important things
- Important things
- Take care of a select few pages (iceberg)
- Look for them across the site
- Do not assume that the most important pages are at the top of the hierarchy
- Build streets, not houses (kudzu)
- Specify content areas, not content in your architecture
- Even if it means to give up on labeling/vocabulary systems
- Modularize (frankenstein)
- Add flexibility to your wireframes
- Anticipate content areas to change
- Build solid organic infrastructure (anthill)
- Structural navigation
- Tag/taxonomy-driven navigation
- Build global navigation upside-down
- Keep it in flux until most of the architecture is complete
- It will come to you: start seeing patterns
- The Purpose Ballad
- Each chunk of content
- Must have a purpose.
- Though your user may not know,
- What that purpose is,
- You should.
- If you don't know,
- Then let it go.
- Chuck that chunk of content
- Oh yeah,
- Chuck-a-chuck it and stop your whining,
- Yeah baby.
- (harmonica solo)
- Answer the "why" question:
- Why is this content here?
- Answer preferably backed by research data
- Simplify deliverables
- Don't waste time preparing complete site blueprints
- For instance, for wireframes, all you need is
- Whiteboard
- Cameraphone
- Flickr Notes
- Use online word processing tools (Google Docs, Writeboard) to collaborate