Structure and Function IA for Web Applications

Structure and Function: IA for Web Applications

Structure - IA with content • In a content-only site, the user interface is easy, • the information architecture is hard Lots of things topics with different sub-structures grows, hard to know how it’s going to expand • but it’s understood Innovate - For What’s Next™ © 1999 Scient, Proprietary and Confidential Page 2

Function - UI for applications • For desktop applications, the information architecture is easy • the user interface is hard lot’s of different actions a user can take they interfere with each other effect of actions needs to be clear to the user • but it’s understood Innovate - For What’s Next™ © 1999 Scient, Proprietary and Confidential Page 3

How is UI different on the web? • Supports more tasks at once lots of domain-specific tools, not one general purpose tool • Supports different tasks shopping, communication, decision-making • Combines traditional tasks with more things Innovate - For What’s Next™ © 1999 Scient, Proprietary and Confidential Page 4

Typical Web applications • Shopping, for simple and complex products • Decision-making • Auctions and marketplaces • Verticals - applications embedded in portals • Process tracking, workflow, negotiations • Status tracking Innovate - For What’s Next™ © 1999 Scient, Proprietary and Confidential Page 5

Structure of this talk • Team structure for these projects • How users and their intentions are different • Common IA challenges in a Web application Object structure Navigation Other • Basic Advice Innovate - For What’s Next™ © 1999 Scient, Proprietary and Confidential Page 6

Teams and Collaboration

User Interface Designers • Design the task flow, and thus the page flow • Design page-level interaction • Bring a lot of knowledge about human-computer interaction Collaborate on: • information display • page layout • navigation Innovate - For What’s Next™ © 1999 Scient, Proprietary and Confidential Page 8

Are you being a UI Designer? Are you: • choosing where to put info and buttons on a page • deciding when to show users info and when not • designing task flow • If so, learn the user interface domain as well Innovate - For What’s Next™ © 1999 Scient, Proprietary and Confidential Page 9

Building a team • Typical design team: UI Designer, Information Architect, Visual Designer • Useful people: Ethnographer, technical writer, usability tester • Be flexible in ownership of tasks • Be collaborative in the design process • Be clear about inputs and outputs Innovate - For What’s Next™ © 1999 Scient, Proprietary and Confidential Page 10

Different Audiences, Different Goals

Who are the users? • People at work • Trying to make money • Trying to save money • Users often aren’t the people who buy the system • People at home doing something important Innovate - For What’s Next™ © 1999 Scient, Proprietary and Confidential Page 12

Same questions, different answers • Frequency of use all day, every day • Level of domain expertise often deep • Language jargon is extensive and important • How optional is it, what happens if it fails not very - either it’s important or they are forced Innovate - For What’s Next™ © 1999 Scient, Proprietary and Confidential Page 13

Common Challenges

Information limits task Structure of things and their attributes sets what is possible • IA needs to see how the info interacts, flexes • Know what users should change, when, why • Guess what tasks the information allows that haven’t been thought of • Overlaps with a DBA role Innovate - For What’s Next™ © 1999 Scient, Proprietary and Confidential Page 15

Information limits task If the user can’t enter it here, it can’t be chosen,

Information affects understanding The user’s mental model is made of things, their attributes, and what can be done to them • Make relationships between attributes sensible, obvious • Know what attributes will be compared in trade-off decisions • Have the UI surface interaction between attributes, the effects of actions Innovate - For What’s Next™ © 1999 Scient, Proprietary and Confidential Page 17

Information affects understanding Innovate - For What’s Next™ © 1999 Scient, Proprietary and Confidential

Keeping users in a task Ubiquitous navigation increases the chance for mistaken moves • Collapse general navigation • use sequence nav • Avoid related links on the pages for a task • Use pop-ups for honest side tasks • Try to make tasks short On a web app, users WANT to stay on task Innovate - For What’s Next™ © 1999 Scient, Proprietary and Confidential Page 19

Keeping users in a task sequence navigation with collapsed global nav- Meta. Design Innovate

Navigating between tasks • Some tasks need instant access at all times need to understand the user’s day and mix of responsibilities • Some tasks are related and grouped need to know the user’s more general intention to decide connections between tasks • Tasks are less likely to expand than lists of things horizontal navigation often works Innovate - For What’s Next™ © 1999 Scient, Proprietary and Confidential Page 21

Navigating between tasks Switch active orders and adding new ones, rarely email Switch between sourcing, buying, and looking for partners Innovate - For What’s Next™ © 1999 Scient, Proprietary and Confidential Page 22

Flexible navigation Different situations demand different navigation • Looking for a task vs. completing a task • finding a thing vs. finding a task • different users, with different roles and permissions • Some of this is in the “global nav”, some is an issue of links that appear or don’t appear Innovate - For What’s Next™ © 1999 Scient, Proprietary and Confidential Page 23

Flexible navigation Buyer has a different navigation than the agent Innovate - For What’s

Navigating tasks and sub-tasks • Use pop-ups judiciously • If the sub-tasks are optional, highlight the typical next step • Design a good multi-level sequence navigation • This is tied to the UI area of task flow, but determines pages and structure Innovate - For What’s Next™ © 1999 Scient, Proprietary and Confidential Page 25

Navigating tasks and sub-tasks Effective use of pop-ups helps support sub-tasks -Meta. Design Innovate

Look for a thing, then choose a task • The e-tail model, but often have more tasks • Have the right tasks available at the right level in the object hierarchy • What can I do to a class of objects? • What can I do to one object? • At what point do you have matrix navigation Innovate - For What’s Next™ © 1999 Scient, Proprietary and Confidential Page 27

Look for a thing, then choose a task List of things with actions attached

Maintaining tasks over time Many tasks extend over days and weeks • Have workbench for user’s tasks, what they are responsible for • Display status and provide access, due dates • Notification of events, with email or on that workbench Innovate - For What’s Next™ © 1999 Scient, Proprietary and Confidential Page 29

Maintaining tasks over time Workbenches monitor and provides access Innovate - For What’s Next™

Searching an application • Users are looking for how to do a task, not for a piece of info • Many pages should not be searched at all • Heavy use of meta-tags rather than full text search • Search is often part of a task, not just navigation Innovate - For What’s Next™ © 1999 Scient, Proprietary and Confidential Page 31

Basic Advice

Task-focused research • Can’t rely on card-sorts • Listening in context is often the only way to find jargon • Different groups of users are different Collision repair shops differ in how they order parts, who does it, what they call the tasks Innovate - For What’s Next™ © 1999 Scient, Proprietary and Confidential Page 33

Wallow in the information • Things, attributes, and relationships have a huge impact • Things define tasks • Tasks determine things Innovate - For What’s Next™ © 1999 Scient, Proprietary and Confidential Page 34

Start IA and UI at the same time • UI (page flow) controls IA below the top level • Both need to learn the same stuff, work on the same design problems • Knowing the things requires knowing the tasks Innovate - For What’s Next™ © 1999 Scient, Proprietary and Confidential Page 35