How to Implement Intranet Task Testing

  • Bauer Media has seen 69% of all questions intranets receive feedback from a variety of departments
  • McKinsey estimates that by adopting intelligent social intranets, companies increase efficiency by 25%
  • At IBM, 52% of employees are more satisfied because of information obtained on their intranet.

Intranet task testing helps build a user-friendly navigation for your employees.

It is easier, cost-effective, and yet very fast, often following a card sorting process, to create a intuitive and user-oriented intranet navigation using the task testing method.

Screenshot depicting individual task being analyzed

Screenshot depicting individual task being analyzed

Task testing: what is it and when to use it?

Intranet task testing is just like the card sorting process, which helps improve user engagement on your new intranet, building a strong foundation for its long-term adoption.

Task testing is also an evaluative process: it tests how well the new navigation helps users find the information they’re looking for.

Usually, task testing begins right after content audit (that help define the scope of an intranet), content sorting (a design tool used to draft a navigation), and other activities designed to collect information about content and user’s needs.

So rather than create information architecture, task testing tests and improves it.

Here is an example:

An employee has returned from a business trip and wants to document all the expenses and submit for reimbursements. Where should he go to find the necessary information within the site navigation to do this?

The user navigates through the draft navigation looking for necessary information that’ll help him/her complete this task. So task testing is nothing but a process of finding out how well the organization of the navigation and labels helps users complete important business tasks.

Note:  Task testing tests the navigation design, not the participants – meaning low success rates of testing will always convey a “poor” navigation design.

Why task testing is important?

There are many benefits of conducting task testing:

  • Task testing validates an IA and gives useful feedback on how to improve it.
  • Task testing helps find out how easily participants can find information using the navigation, rather than performing searches.
  • Task testing improves the quality of information and design, while also sharpens team’s focus and help them connect to real-world uses.
  • Task testing also helps improve user engagement: Intranet users feel they were involved in building the new tool. The more you engage your participants in building new intranet, the easier it becomes to adopt, because the engagement builds a sense of collective ownership and excitement.
  • The task testing makes the new intranet application easy to use.

Now that you know all the basics and the benefits of using task testing, let’s look at how to perform intranet task testing successfully:

STEP 1: Involve employees from all levels

There’s one major goal of task testing: to find out participant’s – employees’ – viewpoints on how easily site navigation helps them find content or perform a specific task.

Of course, not all participants will have the same viewpoint.

The way they use and see a sitemap may vary, depending on their locations, job levels, departments, etc.

This is why you need to involve as many employees as you can throughout the hierarchy – from executives to administrators as well as people from all locations and departments – to capture full range of perspectives.

How many users should you involve?

We suggest gathering at least 20 or more user testers for intranet task testing, as fewer users can lead to inconclusive and vague results.

STEP 2: Confirm draft navigation & note down assumptions

The whole point of implementing intranet task testing is to confirm draft intranet site navigation using all the assumptions that resulted from the card sorting process.

So to implement intranet task testing properly, you’ll need to document all the assumptions that resulted through the card sorting process, and use those assumptions to build site navigation.

Remember that the point of draft navigation is not to get everything right at the first time, but rather to execute a set of assumptions to test.

So no matter how you drafted it, note down all assumptions and finalize your first official draft site navigation, and get ready for task testing.

Simple draft navigation created from Spreadsheet

Simple draft navigation created from Spreadsheet

Tip: Create at least two-level deep navigation for each top-level navigation item, with one or two 3rd-level-navigation for few items.

STEP 3: Recognize navigations problems to test

Next, document all the assumptions (about how the content must be grouped), questions, and uncertainties that result through card sorting process.

You’ll test and validate all these issues in the task testing process.

In this way, task testing is related with card sorting, as you’ll be validating and testing issues that results from card sorting process.

For example:

The card sorting process may have suggested you that the top-level items like “Admin Tools” and “Company Profile” would make sense to all users, however, the process may also have shown that “Branding guidelines” did not fit conclusively in any of these groupings.

Your job is to find which location suits better for the “Branding guidelines” section.

For this, you could place the “Branding guidelines” section under one of the top-level navigation – let’s say, under “Admin Tools” – and create a task to find that information.

The task testing results will then reveal whether the location – “Admin Tools” – was appropriate for that information.

How to begin task testing?

Begin task testing by listing all major issues with the draft site navigation that you feel most unsure of. In addition to those issues, list all key assumptions you’ve made how the content should be grouped together.

For example, during card sorting process, you may have assumed that “since most users grouped Human Resources related topics together, we need to create a separate top-level navigation section for all HR related content.”

That’s an obvious assumption, but you still want to test this hypothesis by creating a task.

We advise that you create at least 8-10 issues to test from your list of questions and assumptions.

You’ll be creating tasks based on these key issues.

Why test 8 to 10 issues, you ask!

For two reasons:

  1. Completing more than 10 tasks may be too much for user testers; they’ll likely lose their interest.
  2. And, 8-10 tasks are sufficient for a user tester to become familiar with the navigation.

STEP 4: Create scenarios for task testing

For each 8-10 issues, you’ll write a task testing scenario.

What is a task scenario?

It is a real situation, where users are asked to complete a specific task using information found on the Intranet.

For example, to test “HR” in the global navigation, you can create a task related to common activity like “annual performance reviews.”

The task scenario could be written like this: “You have to meet with your manager for the annual meeting to review your work. Where would you go to find this information on the Intranet?”

Remember, each task scenarios you create have to resonate with all the participants, and not just few. You cannot create “marketing” related scenario if one or more users don’t know what it is.

The main objective of task testing is to test the top two and three levels of navigation that makes sense to all users.

STEP 5: Find the location of the “correct” answers

After you’ve written down at least 8 to 10 scenarios, your next step would be to find the location of the “correct” answers, and for each scenario.

In the “HR” example in STEP 4, you’ll find that the “correct” location for the common activity “performance management” falls in the section: Home > Human Resource > Performance Management.

This is how the correct answer looks like in a simple navigation:


  • Staffs
  • About ABC
  • Projects
  • Sales & Marketing
  • Offices
  • Admin Tools
  • Human Resource
  • Performance Management
  • Annual Performance Reviews
    • You’ll discover many holes and flaws in your navigation as you complete STEP 5.

      This usually happens because of card sorting results, which doesn’t always lead to final site navigation.

      But that’s all fine!

      The process of building an intranet-based navigation is repetitive, and discovering weaknesses and holes in your navigation is quite helpful for task testing, too.

      STEP 6: Check scenarios for language

      Because task scenarios often include topical keywords that are directly related with sections in the draft navigation, you run the risk of revealing “correct” answers to test users.

      For example, consider this scenario:

      “Your annual performance report is due soon and you must meet with your HR manager to review your work. Where will you find forms to do that?”

      Tip: Remove or re-write those navigation topical keywords, while still keeping the language simple and easy to understand by every participant, like this: “Your annual performance report is due soon for review of your work from the past year. Where would you find this information?

      For more information, check out Jakob Nielson’s excellent article on Terminology matching.

      STEP 7: Prepare for online task testing

      Use TreeJack – an online task testing tool – to build draft navigation, quickly and easily.

      First, build your draft navigation, then create tasks around it.

      Look how the top-level navigation looks in our online software:

      E.g. Top-level navigation that’ll be displayed to the task tester

      E.g. Top-level navigation that’ll be displayed to the task tester

      By clicking on any one of these top level navigation tabs will reveal you all the sub-page that falls under it, while other tabs are hidden automatically:

      top-level-site-draft-nav-when-clickedThen add tasks to test.

      Here’s an example:

      For each task, type the text of the task scenario, and choose the location in the site navigation where the “correct” answer lies.

      STEP 8: Start online task testing and invite users

      Discuss with your team how long the task testing exercise should run, and then, set it “live” in the TreeJack software.

      If users are responsive, typically, you’ll collect enough responses within few days.

      On the other hand, if users are located in different time zones, you’ll have to run the task testing exercise open for a week or more to allow all participants to complete the task.

      To notify users, send an email with a link to the task testing.

      Include this information in your email:

      • Introduction to the new intranet
      • Explain how each participants (employees) will help craft the structure of the new intranet
      • Link to the task testing software
      • Expected time required to complete the task; deadline for completion
      • Point of contact for more information

      It’s fine to use an email-template like this, but we suggest you to use your own language and style that matches your company’s policy, values, and cultures.

      Tip: Send this email to all participants a week or two before you begin task testing. This way, you’ll know in advance who’ll be available for task testing exercise and who’ll not be in office.

      STEP 9: Analyze results and update your site navigation

      Although the online task testing tool makes it super simple to conduct a task testing, you still have to make assumptions and guesswork about what the results mean and how to analyze and solve issues.

      In short, you’ll analyze “agreement between users”: How many users completed the task successfully – how many selected the correct answer within the navigation for the information needed to complete that task?

      This is how the task testing result looks in a simple table format:

      the task testing result looks in a simple table format

      The task testing result looks in a simple table format

      At the top of the table, 10 tasks are listed, while on the left of the table is the site navigation.

      Since the data is presented vertically, you’ll need to read vertically to understand what it means:

      • The green cell refers to “correct” answers with correct number of responses.
      • The white cell refers to incorrect answer
      • While a red cell indicates incorrect answers selected by 10% of more users.

      For task 1, all participants (17) selected the right answer. This result proves that the organization of the site navigation and terms used for task 1 are 100% correct and sensible for all users.

      For task 4, on the other hand, there were three different responses, with 1/3rd of participants choosing the “wrong” location. This result suggests that this area of the draft navigation needs deeper analysis.

      So for task 4, your analysis should begin like this:

      • Why did so many users chose the “incorrect” answer?
      • Were the terms used in the site draft navigation closely related?
      • Are the “incorrect” locations just good as the location for the “correct” answer?
      • Did users navigate through the top-level correctly or they made a wrong turn from their very first click?

      For task with low success rates, you could either do of the following:

      • Change the term (topical keyword) in one or multiple sections within the site navigation
      • Move a section.
      • Change an item or add a new in the navigation.
      • Re-group the content.
      • Or, simply cross-link sections that hold closely related content.

      After analyzing and adjusting your site navigation, you can either do other Intranet projects, or conduct another round of testing. Online task testing is simple, easy, inexpensive, and very fast to implement. Within few hours, you should be able to run several rounds and make adjustments to your site navigation.

      Once you’re done with several round of testing, finalize your new intranet navigation and move on to content migration.

We will be happy to see your thoughts

Leave a reply

Login/Register access is temporary disabled
Compare items () compare