London IA mini conference

  • Location: Guardian Offices, 90 York Way, London, W1W 6DS
  • Tag: #iamini

These were notes taken live, so they will contain errors, omissions and ridiculous typos.

Intro

  • Speaker: Ken Beatson

Goals

Take away one thing that you can change in your work

5 minutes of something from your every day work.

Format

  • 5 talks
  • 1 workshop
  • More participation. Shorter, more interactive sessions.

Introducing Information Architecture at The Guardian

  • Speaker: Martin Belam (The Guardian)

  • Presentation: Introducing Information Architecture at The Guardian

  • Guardian just won Newspaper website of the year

  • Joined Feb this year

  • First IA @ the Guardian

  • Process (was): Product Manger -> Designer -> Software Engineers (Agile, 2 weekly sprints)

  • Product Managers: IA = wireframes

  • Designers: IA = Visual Hierarchy

  • Software Engineers: IA = Domain Driven Design (CMS)

  • Embed UCD in product development

  • "Ambush User Testing". Take Silverback out and ask people how they get their news!

URLs should be:

Permanent

Addressable

Discoverable

Open

  • Using tags for content aggregation pages - who tags?

  • Keyword manager! - official job title ** Keywords added by jourrnalists, editors, etc. ** Manages items & keywords moving from one section to another

  • Open Platform - asking people to use their their content rather than erecting paywalls

  • Building - most environmentally friendly building in London (Kings Place)

  • "Producing a newspaper is an information service, and the Internet is an information platform."

  • "Weave The Guardian into the internet."

Ten minutes on Agile user experience

  • Speaker: Cennydd Bowles (Clearleft)

  • Most agile is bad -- rushing headlong into coding w/o understanding the problem space.

  • Devs tend not to understand the big picture

  • "User scented design" - results of agile dev, related to genius design *agilemanifesto.org Agile Manifesto - set of values, no process mandated

  • Model Users in Iteration 0 ** Personas ** Concept maps ** Goals ** Use cases

  • prioritise user stories - designers must be involved ** find out about user stories ** flow from personals and scenarios

  • Jeff Patton - workahead - research ** Research n+2 ** Design n+1 ** support n ** Test n-1

  • Get the structure in place - let devs set up db, get data structure right

  • Design the obvious stuff first - profile / product pages

  • Involve the users in every iteration

  • Needs senior people

  • Good communication - designers can't be separate from developers* Article: Getting Real About Agile Design

Questions

  • Horizon factor - push components into a subsequent iteration
  • Agile is light on documentation - gives time to think?
  • Buying time with the backend - how do you avoid backend constraints? ** Takes time to set up the environment? ** Talk to the devs, DBAs etc. Ask what they're working on
  • Avoid problems by sitting with developers while they're making it.

Why users don't follow instructions

  • Speaker: Phillip Winwood (Usability Tester)

  • Psychology background

Scope of the talk

  • Get out of central London
  • Think in terms of people who don't understand what is going on
  • Software in general (not just web)

Talk

  • People's brains are pattern matching systems, works with tricks - apply known patterns to new things
  • What people know affects how they interact
  • William James: Infants perception involves a "buzzing, blooming confusion"
  • Don Norman & affordances - 'pull me' doors that need to be pushed
  • People have learned that you put the website address in Google b/c it works ** That's the model they have in their head
  • users don't like to read ** of the Active User] - academic paper (see area42.co.uk) ** in psychology *** production bias - people want to get things done, ignore dealt *** assimilation bias - use what you know to solve a problem
  • users aren't a blank slate ** applying what you already know makes it worse ** see Don't Make Me Think - people don't want to think, want to apply what they've previously done to the current situation
  • learn until you can make it work, then you stop learning ** asymptote of mediocrity
  • Word Manual (Word 2 - 1991) - 800 pages! ** User's don't read manuals - never got used.
  • Recall vs Recognition ** Hearing your name at a party. Your brain changes focus ** More difficult to recall than recognize
  • Bottom Line: knowledge is always partial

Questions

  • Q: What happens when people don't have preconceived notions about how a phone / camera will work? ** A: Pitch terminology at the right level. People don't have enough time. Time is the killer.
  • Q: Is there anything you can do to encourage people to read some instructions? ** A: Open up the potential of the product: "READ ME FIRST" manuals - a piece of paper that floats out with vitally important information
  • Q: Quick guide is useful, but pitch it as benefits for the user. ** A:

Response

Good talk, despite the technical difficulties.

Takeways:

  • users are busy, distracted
  • users are different, they all have different experiences
  • very few are or will ever become experts

Spec docs from Axure wires

  • Speaker: Ken Beatson

  • Slides and a software demo all in one. Exciting

  • myvillage.com

  • IA, Design in UK; devs in Belarus

  • requirements are important

  • wireframes are used, very few prototypes

  • muppets how dancing cheese sketch?

  • iterative deployment of modules

Axure demo

  • he uses two fields - display rules & what's changed since last version
  • walk stakeholders through jpgs of wireframes
  • you can export a word document from Axure ** often good enough to send to devs ** sometimes he needs to generate prototypes

Questions

  • Q: In Axure, once you start to add other features, do the spec documents become less useful? ** A: Yes. If you generate a prototype, the spec docs is less useful.
  • Q: (followup) What are other people using for creating heavy documentation (Axure vs. OmniGraffle). ** A: Stop-motion animation output to vimeo. ** A: Mallof has some good blog post ** A: Getting more sketchy. Interactive Prototypes are the documentation.
  • Q: How to capture change. ** A: Capturing rationale is more important.
  • Q: What do you get your clients to sign off? ** A: Talk to your clients and pick reasonable clients.

Great discussion on "locking clients down" vs negotiation. No real resolution, though.

Design Consequences workshop

  • Speaker: Leisa Reichelt

  • Presentation: Design Consequences: A fun workshop technique for brainstorming & consensus building

  • Solution to the problems we were just talking about

  • Used to write beautiful documents w/ annotations ** people didn't read them or didn't have the skills ** we can't ask people to be literate in wireframe reading

  • Get people around the table

  • "Pull problems to the front of the process."

  • She uses this a lot during the project initiation phase

Design your own hat exercise ("Design Consequences")

  • Sketching exercise stops preciousness surrounding design. Nice! I like it.
  • First person sketches a solution
  • Hand off to next person, who picks a part of the sketch and carries it on to the next level. What happens next?
  • Really useful and fun. Generated some useful ideas.

Leisa continues:

  • Two things that happen: *# People realize that they have a picture in their head, and that there are different solutions. *# Good to involved front-end developers who often have ingenious solutions.
  • including people early on improves communication with them
  • afterwards: *# Design one version together (cheap option) *# extract concepts onto sticky notes and affinity sort/rank (KJ Method)
  • recommendations ** do a pilot session ** make sure the design question is clear ** do a design warmup exercise (e.g. something to entertain a dog, irritate an enemy) - get people sketching ** invite a multidisciplinary team (including the people you most fear) ** be creative with materials / stimiulus ** always show example output (show that your sketches are sketchy! show your worst work!) ** bring energy (jelly beans! "spike their energy with sugar.")

Q&A

  • Q: Would you consider spending longer on wireframing. ** A: Different exercise. This is getting ideas out there. Getting it perfect would be a different project.
  • Q: Do you use this to teach people a lesson. ** A: Use it to show the work that designers do and take the mystery out of it.
  • Q: Do you have clients who think you're getting them to do the work they are paying you to do? ** A: Even though I have designer in my title, I find myself facilitating design among my team.

Interacting with forms

  • Speaker: Matthew Solle

  • Big problems with forms.

  • 3 categories *# Non-critical: signing up for email *# Non-critical but critical to fill in: Online share dealing account *# Critical: Tax, passports, visa, etc.

  • non-critical forms that are excellent ** Huffduffer ** Google account (see if your account name is taken) ** Google search ** delicious - very clear ** radar -- is that the electronic business card site? ** Audience comment: recall vs. recognition?

  • non-critical but problematic ** Amazon - log-in ** Google account vs. Gmail account - two different places, very confusing ** Halifax - application form

  • Passport / ID card is coming (critical) ** "Secure data" ** Do online forms communicate with this? ** Will this be done badly?

(More) Audience Feedback

  • Verified by Visa - critical, but crap?
  • horsesmouth.co.uk - you need to be approved and they text you. Jane Austin did the IA.
  • Jonas Löwgren - Ethical design
  • Interesting: TFL has poor wayfinding to prevent bottlenecks elsewhere in the station.

Interactivity - how IAs learned to stop worrying and love designers

  • Speaker: Tom Coombs (www.manwomanandchild.comlink)

  • Flash -

  • dull wireframes result in dull websites?

  • Silverlight demo - "Contoso Fixter" - Silverlight deep zoom - you can zoom in forever and just download what you need

  • Flash demo - timeline & tweens - (me: it's been a long time since I've seen flash. brings back memories.)

  • "Dynamic specification" ** Really easy to design for interaction *** Demonstrate interactions *** Easy to build interactive prototypes

  • Really cool - webcam hacked to see infrared light to create a multi-touch interface.

  • Video: vimeo "Playing with Multi-touch" by IDEO labs link?

  • A new kind of interaction. Not actually happening on the web. Exciting to play with that stuff.

  • 3D - Flash demo of 3d-like behavior.

Audience Questions

  • Q/A: Flex for prototyping? A lot of the interactivity is build in?
  • Q: Where are you going with the touch table thing? ** A: Thinking about it at my desk. NOt sure where it will end up.
  • Q: Tutorial for infrared stuff? ** A: A lot of people working on it. LIbraries, code bases, etc. link?
  • Q: Resolution of Prototypes? What are user testing responses like? ** A: The Flash mockups are conceptual, but can be too high resolution for some situations.