Ontario.ca IA and search

What our team is about

Our team is dedicated to improve how users find what they need faster and easier online. We analyze how users would:

organize information (including taxonomies and hierarchies)
label information
navigate from where they are to where they need to go
search to find the content they need

We research and set standards for

Labels

How we write and structure content such as:

titles
ledes (summary of page)
headings
links

Organization systems

How we categorize, create relationships and present content such as:

taxonomies and hierarchies
hyperlinks
organization schemes (for example: alphabetical, chronological, topic or task schemes)

Navigation systems

How we browse for information such as:

menus
breadcrumbs
headers
footers

Search systems

How we find information such as:

search interface (search bar and results)
query builders (spell checks, autosuggestions)
filters and facets

Hidden components

Components that feed into other systems such as:

vocabulary and thesaurus lists
metadata
index terms

User goals

As a user of ontario.ca, I need to:

land on the page I need or want the first time
trust the content I find
know what options are available to me

So that I can:

complete my task
answer the questions I had
discover relevant information that I didn’t know about
have reassurance that I had all the information I need

As our fairy tales suggest, getting lost is a bad thing. It is associated with confusion, frustration, anger and fear.
_{— Louis Rosenfeld, Peter Morville and Jorge Arango.}_{Information Architecture for the Web and Beyond (4th edition).}

How users browse for programs and services

We’ve conducted research sessions to learn how people of Ontario would categorize programs and services (such as renewing driver’s licence, applying for student loan, or registering a business).

We discovered that users:

usually organize services by topic such as driving, health, or travel.
may sometimes organize services by role such as seniors, students, or parents.
rarely organize services by region such as Northern, Central, or Southwestern Ontario.
do not organize services by ministry or government body such as ServiceOntario, Ministry of Transportation, or Consumer Protection.

The findings guide how we structure content on ontario.ca.

Ongoing research activities

Tree tests and UI patterns

We conduct tree tests to evaluate how well users can find answers to relevant tasks.

We also explore UI (user interface) patterns that will reflect how the tree is structured.

A clear structure improves usability as it helps users discover relevant content that they may not have searched otherwise.

Tree structure of driving and transportation in Optimal Workshop. Driving expands to driver's licence. Under driver's licence, it shows: getting a driver's licence for cars, motorcycles, trucks, and bus and ambulances.

Testing our structure

Prototype of an ontario.ca page for Driver's licence. Options include getting, renewing, replacing, exchanging a driver's licence.

Exploring UI patterns

The structure of information environments influences more than how we find stuff: it also changes how we understand it.
_{— Louis Rosenfeld, Peter Morville and Jorge Arango.}_{Information Architecture for the Web and Beyond (4th edition).}

Example of an open card sort test

Open card sorts

We conduct open card sorts to understand how users would group and label the content.

With clear labels and a logical structure, we can help users locate specific information.

Search data and behaviour

We analyze top search queries and prototyped ways in which we can design the experience to complete tasks as fast as possible.

Search query with "St" and there is an autosuggest box with a few selections: Start a business in business and economy; Start a family in courts, justice and the law; Student aid in education and apprenticeship; Study abroad in education and apprenticeship.

Example of search prototype

Organization terms vs. user terms

How government addresses a program or service can often be different from how users would refer to them. For example, vehicle permit is the same as green paper, green permit, or registration papers.

We interview ministries and gather user data to create a curated list of terms that help us understand:

what the official term is
what the French equivalent is
what it means in the context of Ontario
words that users use to refer to the official term

Vocabularies and thesauri can enhance the search experience as users do not have to type the exact words to get to the results.

How IA supports productivity and efficiency

Easier to content audit

When structures are clearly mapped, this enables ministries to evaluate content better as it is easier to spot where everything is.

Hierarchical structure of information organized in multiple groups with its children.

Example of how we organize content

Cut out of an outline of a document. Image shows philosophy or rationale, maintaining the IA, content standards and patterns for titles.

Work-in-progress standards document

Easier to maintain and scale

As we analyze information, we document standards and systems that can apply across ontario.ca so that it will be easier to add, modify, or archive content in the future.

Reduce technical and design debt

With strong content standards and systems in place, we can have a leaner CMS (content management system).

This means we can spend more time building user-centred services and less on content that no longer serve user needs.

Takeaways

Standards, systems and training are more important for our team than categorizing each content on ontario.ca, as this can help us scale faster and better.
There are many components to IA that impact each other and it can be very overwhelming as we work. Rather than focusing on one thing at a time, we document our ideas as we come across insights.
There is a lot of government content to organize as it encompasses everything from birth to death. We constantly have to limit our scope and implement in small cycles to move ahead.
Good IA is not possible without collaborating with stakeholders, teams, and user research.
There is no such thing as “perfect” when it comes to IA as we all interpret information differently. To keep the momentum going, we aim for “good enough” and allow room for modification when needed.