1 of 53

SUMO

Information Architecture Project – Phase 1

February 2012

Susan Farrell – Nielsen Norman Group�Bram Pitoyo, bram@mozilla.com

2 of 53

Overview

  • Project goal and scope
  • What we’ve done, why we did it this way,�and how we’ve done it
  • Existing IA/sitemap structure test
  • The new IA
  • Findings, insights and recommendations
  • What now?
  • What next?

3 of 53

Project Goal

Outfit SUMO with a new design, based on a new Information Architecture / taxonomy that:

  • Is scalable
  • Has a robust structure
  • Has a navigation built to be extensible
  • Is based on research and data

We focus on what SUMO users are doing,�in order to optimize their tasks, by making navigation and search result improvements.

4 of 53

Project scope

Analyze Data and Propose IA Structure

This presentation

Remote Test IA Structure

Test Paper Prototypes

Turn Paper Prototypes into Wireframes, Test

Develop Wireframes into Web Pages

Introduce New Website to the Community

5 of 53

Why we analyze data

in-site and across silos

  • According to search log and FAQ, some key user tasks & searches naturally cross over.
  • Usability improves ease of use in a systemic way. This means looking around the edges.
  • Usability issues often stem from info containment within one service.
  • Websites always have to be optimized to work against many other external contexts.
  • Support websites need to improve information exchange with other areas of the organization.

6 of 53

What we’ve done

Analyze data sources in-site and across silos

Interview three SMEs

Identify top user tasks and top support needs

Test existing IA Structure

Deliver Proposed IA Structure

7 of 53

How we identified

top user tasks

Cross reference the tasks we have built from:

  • Top 1000 search queries
  • Top 100 en-US pages

To see if there are good correspondences between most-found pages, and most-looked-for information.

8 of 53

9 of 53

How we test

existing IA structure

  • Turn our sitemap into an Excel document
  • Plug it into TreeJack
  • Test our top tasks against the sitemap with 85 users
  • Make recommendations that help the new IA

10 of 53

Project goal & scope

What we’ve done, why we did it this way,�and how we’ve done it

Existing IA/sitemap structure test

The new IA

Findings, insights and recommendations

What now?

What next?

11 of 53

12 of 53

13 of 53

14 of 53

Notable findings

from existing IA

  • People mix up “getting,” “installing” and “registering.”
  • People are having a very hard time finding information on saved forms & passwords, and parental controls. This stuff is important.
  • Most people choose sync as a way to move bookmarks.
  • Most people look for search engine customization on add-ons, toolbars and setting the home page.

15 of 53

The new IA

KB Articles

Topics and Categories

Tasks

Products

Products

Topics

Controlled

Categories

KB Articles

Old

New

16 of 53

The new IA

  • Navigation
    • SUMO global navigation
    • SUMO utility navigation
    • Mozilla global navigation
  • News
  • Task-based section
    • Featured issues
    • Common tasks > product picker
    • Knowledge Base > category landing page
  • Support by product
    • Product landing page > topic landing page

17 of 53

The new IA:�Navigation

18 of 53

The new IA:�News

19 of 53

The new IA:�Task-based section

20 of 53

21 of 53

22 of 53

23 of 53

24 of 53

The new IA:�Category-based navigation

25 of 53

26 of 53

27 of 53

28 of 53

The new IA:�product-based section

29 of 53

30 of 53

31 of 53

32 of 53

Project goal & scope

What we’ve done, why we did it this way,�and how we’ve done it

Existing IA/sitemap structure test

The new IA

Findings, insights and recommendations

What now?

What next?

33 of 53

Findings, insights

and recommendations

People were extremely concerned about private browsing and erasing data, contrary to every article you’ve read about how people don’t care.

Peoples lives, freedom, jobs, marriages, and children’s safety all depend on getting this information clearing function right.

This aligns with Mozilla’s mission perfectly.

34 of 53

Findings, insights

and recommendations

Our users run the gamut from prelingual, ESL, low-literacy, etc. Be kind. This fact has to inform our writing level.

We need more scaffolding (examples, tutorials, pictures, screen movies of interactions, etc.) and try to alter the behavior by altering design and messaging.

35 of 53

Findings, insights

and recommendations

People search in the wrong places (Facebook)

Font size turns out to be a giant issue. Can we bring visible controls for zoom back, or make them easy to access? People must be able to read web pages.

When searching, it matters how you�spell “favourites”

36 of 53

Findings, insights

and recommendations

People search for “export bookmarks”,�not “exporting bookmarks”

Articles need to vary verbs they use.

We need to have verb synonyms in the�search engines.

37 of 53

Findings, insights

and recommendations

People search for problems, not solutions.�As a result, MozillaZine articles from 2002 are ranking higher than SUMO KB articles for the same topics.

We need to demote MozillaZine, identify things from MozillaZine that are really valuable, and bring those into SUMO so we can promote and maintain it. Same goes for obsolete contents�on SUMO.

38 of 53

Findings, insights

and recommendations

Users absolutely need to know where to find info and get help before they get frustrated. Every permutation of contact support, email support, get help, unanswered questions, etc. needs to be a fruitful search term.

Most casual software users don’t make the distinction between customization/settings/ preferences/add-on management.

39 of 53

Findings, insights

and recommendations

It’s clear from looking at the Ethnio Survey answers that a lot of people do not separate�the browser from the web. Let’s make sure�that we teach them mental models in the Getting Started area.

Good mental models will help them understand concepts like saved passwords, forms, and knowing when one is logged out. It will also help them when reporting problems.

40 of 53

Findings, insights

and recommendations

User’s tasks frequently cross boundaries:

  • Download, installation, compatibility and version info�(overlap with product marketing websites)
  • Ways to join Mozilla as a contributor�(About us, participate, social media support efforts)
  • Accessibility issues (accessibility info,�missing content, product teams, font size)
  • SEO issues. Google ranks old info higher than�newer SUMO info (MozillaZine, Yahoo!Answers, etc.)
  • Mobile site, Thunderbird site, Newsgroups, forums
  • Add-ons

41 of 53

Accessibility efforts address

many common needs

  • Reading difficulties, ESL / low literacy:�need predictability, easy reading level, font-size controls, good contrast, focused layout, color blindness-aware palette, pictures, in-context explanations and actions, use core vocabulary.
  • Low/no vision: means linear access and keyboards NOT random access and mice; use video and audible descriptions, provide text transcripts, avoid popups.
  • Low/no audio: provide text transcripts.
  • Technical: require device/platform independence,�web standards, avoid hover for anything important.

42 of 53

Findings, insights

and recommendations

Combine search results until search is unified.

Titles, URLs, headings and link text should correspond to each other, be human readable, and indicate something about the IA.

Start tweaking old content right away. Make�a new guideline that applies to new content.

43 of 53

Findings, insights

and recommendations

Do periodic search and data analysis to make and measure improvements.

  • Determine emerging tasks
  • Adjust list of top tasks
  • Adjust synonym ring
  • Present emerging issues to product teams
  • Identify missing content

44 of 53

Findings, insights

and recommendations

WMP plugin, later identified as a malware issue, was getting a lot of visits and searches. This is why support for multiple article tags�is important.

(plugins + malware)

45 of 53

Findings, insights

and recommendations

There is currently a lot of term ambiguity�in product facets. For example:

• Desktop: I thought Desktop refers to *my* computer?

• Mobile: the mobile version of SUMO? Firefox for Mobile? Is my laptop mobile?

• Firefox Home: the top page of any website is a “home”

• SUMO home ≈ Mozilla home ≈ Firefox Home

When naming things, qualify them:�“Mozilla Support Home”, “Firefox Mobile for Android”, “Firefox Home: sync for iOS”

46 of 53

Findings, insights

and recommendations

We need specific and visually identified steps of what new contributors and people who seek help should do to get started:

  • Make it friendly — tone = rainbow dolphins
  • Make it clear — don’t make me think!
  • Use video — tell a story and make it personal

This applies across SUMO: Superheroes Wanted, registration page, ask a question, etc. The goal is to get more contributors.

47 of 53

Findings, insights

and recommendations

Users in the forum who are looking for info are not finding content that exists.

KB articles don't live anyplace in particular and have no way to browse to it.

Therefore people ask the same questions again and again.

This repetition leads to attrition of contributors.

48 of 53

Findings, insights

and recommendations

Compatibility is a very big deal. It can be disempowering and drive people to other browsers. Whenever installation instructions exist, also provide uninstallation instructions and download links to other versions.

49 of 53

Project goal & scope

What we’ve done, why we did it this way,�and how we’ve done it

Existing IA/sitemap structure test

The new IA

Findings, insights and recommendations

What now?

What next?

50 of 53

What now?

  • We identified some specific content�and search engine tuning needs.

  • We can make big strides and see�significant results now.

  • Tuning can be done in parallel with�our IA/UX work.

51 of 53

What next?

Phase 2

  • Card sort to group KB articles into tags

  • TreeJack test the draft sitemap with top tasks to compare against current sitemap results

52 of 53

What next?

Phase 3

Paper prototype development and testing – to determine what goes on which pages and work through some interaction alternatives.

53 of 53

Further Reading