In this post, I want to explain what it is, how it helps, and how to run a lightweight version of it on projects of almost any size.
Let’s jump in.
What is Top Task Analysis and How it Can Help
Top Task Analysis is an approach you can use to identify and prioritize actions users want to take on your website or in your app.
That might include tasks users want to complete or information they want to know.
For example:
- Building your information architecture, structuring navigation and content around what users actually want instead of guessing.
- Finding the objections holding people back, so you can address recurring concerns before they cost you a sale.
- Understanding the questions people care about most, giving you a ranked list of what your audience wants answered rather than what your stakeholders assume they want answered.
- Deciding what deserves easy access, based on the features people actually use every day rather than burying them further down the page.
- Prioritizing your dashboard, so the most important information sits at the top and everything else gets tucked away further down.
- Planning a redesign, so you know which tasks to protect and don’t quietly break the essentials while rebuilding.
However, Top Task Analysis doesn’t just gather these tasks; it also helps you prioritize them.
The need to prioritize cannot be overstated. You will find that most users are primarily looking for only a tiny subset of what your website offers. These are known as top tasks.
Unfortunately, users often struggle to find these top tasks because they are hidden by the plethora of tiny tasks (those only a few people are interested in). It becomes like seeking the needle in the proverbial haystack.
To solve this problem, we need to know what matters to users and prioritize that within our site or app with clear signposting.
For example, I often use top task analysis as one of the first steps in creating a site information architecture. Once I know the top tasks, I can carry out an open card sorting exercise using them. I then run closed card sorting to integrate the remaining tiny tasks. In other words, I build a site structure around the top tasks and fit the tiny tasks around that.
Hopefully, you can see the value of this approach. However, Top Task Analysis can feel a little intimidating.
Doing Top Task Analysis the Right Way
Gerry McGovern’s process for running a top task analysis is roughly the following.
- Gather a long list of potential tasks.
- Rationalize this list into something more manageable.
- Get users to vote on what tasks matter the most.
- Analyze the results
Without a doubt, this is the right approach for achieving the best results. It draws on an extensive range of sources for identifying potential tasks, including:
- Customer feedback.
- Stakeholder interviews.
- Competitive analysis.
- Social media.
- Site behavior.
- Search analysis.
It also puts considerable effort into rationalizing the list through techniques such as:
- Eliminating overlapping tasks.
- Avoid audience-specific tasks.
- Removing vague concepts such as “self-service” or “knowledge base.”
Not only does this improve the quality of voting, but it also makes later analysis easier.
The downside is that this can prove time-consuming and puts some organizations off the technique.
Fortunately, there is a lightweight alternative.
Lightweight Top Task Analysis
The secret to a lean approach to Top Task Analysis is to combine the gathering of tasks with their prioritization.
This combination can be achieved by running a survey that allows users to share their top tasks but also votes on the suggestions of others.
Unfortunately, there aren’t any tools out that support this particularly well. However, as we are now in the age of “vibe coding” I built my own that you are welcome to use completely for free.
What makes this app unusual compared to other survey platforms is that it allows users to submit ideas that others can then vote on.
Adopting this approach significantly shortens the time it takes to run a top task analysis by removing the need to identify tasks upfront.
So let’s run through the steps.
1. Create a new survey
To get started go to the website and click “use for free” and login with Google SSO. Once in, you will be able to create a new survey in under a minute.
Start, by entering the name of your survey. If you like you can duplicate an existing survey or start from scratch.
2. Configure your settings.
Once you have created your survey, you have a couple of settings you can configure.
- You have the ability to control if you are still accepting submissions or not. Top Task Analysis, created by Gerry McGovern, is one of the most used tools in my arsenal as a UX designer and conversion optimization specialist, yet few others seem to use it.
- You can decide if people can add new suggestions.
3. Seed it with some initial answers
I have learned that it is not always immediately evident to participants what type of tasks they should be submitting.
To tackle this problem, I tend to pre-populate the survey with a small number of sample tasks to show users the kind of thing to include and the best way to format any ideas they submit.
When looking for ideas to include, I typically brainstorm a few top tasks with stakeholders, especially those who work with the customer daily.
I’ll also run some deep research with AI to surface patterns I might otherwise miss, and I always talk to the customer support or sales team, since they hear the same questions and complaints on repeat.
Digging through analytics, chatbot transcripts, or anything else that shows what people are actually doing on the site rounds out the picture nicely.
Once I have the ideas, I add them to the survey, simplifying them by avoiding the inclusion of verbs. For example, instead of “find a price,” I would include the term “pricing.”
Formatting ideas in this way encourages others to do the same, making the whole survey easier to scan.
People won’t always follow your suggestions, but that is okay. We will have to analyze the results later anyway.
4. Define your audiences
Sometimes it’s worth segmenting your top task analysis by audience too. Not every audience matters equally to your organization, and knowing how each group voted can be genuinely useful. A task that ranks high with casual visitors might barely register with your best customers, and that gap tells you something worth acting on.
To support this, add a list of audiences to your survey, just like you added a list of tasks.
5. Send the survey to your audience
Once your survey has been pre-populated with some tasks, it is time to send it to your target audience. The url for this can be found by clicking on the button at the top right of the settings page.
How you reach out to your audience depends on who you are trying to reach and your available resources.
If you are attempting to reach an audience you are already engaged with, that makes life much easier. You can post your survey to an existing website, send it via email, or use your social channels.
Things become a little trickier if you do not have access to your audience. Unfortunately, unlike many survey tools, my app does not help recruit participants. However, we have limited options with nothing else I can find on the market that offers voting and suggestion functionality.
The best option is to go to a recruitment agency like Centiment. However, this has associated costs.
Services such as Centiment can help with recruitment if required.
Whatever approach you use for recruitment, you also need to get an adequate number of responses. In an ideal world, you should use a sample size calculator to determine how many responses you require. However, this can lead to an intimidatingly large number.
Testing with smaller samples is OK if you know that it may come with inaccuracy. After all, we are trying to be lean. Usually, I aim for above 130 responses to at least counteract any outliers.
That said, something is better than nothing, so even less is acceptable.
The user’s experience
The survey participant will be asked to complete three activities.
- Select an audience that they most associate with.
- Select 5 tasks that they consider most important.
- Order those 5 tasks by priority.
They can either look down a long list of tasks, search for a particular task, add one of their own.
You can see for yourself what the participant experience will be like.
The ranking of the tasks that they select is especially important, as this affects the number of votes attributed to each task. A task that is placed in the first place position will be given five votes, second place four votes, third place three votes, and so on.
Once the survey is sent and the responses received, it is time to look at the results.
5. Analyze Your Results
Your survey results will look like this: a series of user suggestions for tasks prioritized by the number of votes they have received.
Your results should be a prioritized list of user-suggested tasks.
Depending on the survey, this could be a pretty messy list of results.
For a start, you will find duplication and overlap. For example, the following two tasks are similar: “make a one-off donation” and “give a single donation.” These can be combined, and their associated votes added together.
You will also find some ideas that are either spam or do not make sense. You can remove these entirely.
Finally, you might want to reword the tasks to be more readable by stakeholders. You can do this, as I mentioned earlier, by removing verbs. For example, “to find out about pay structure” could be summarized as “pay structure.”
Once you have cleaned up the results, you will quickly see that a small proportion of the tasks receive the vast majority of the votes. These are your top tasks and are clearly marked in the app.
These are the tasks your website needs to make the most findable, and your information architecture needs to support primarily.
Pragmatic, Not Perfect
So there you have it. A lean approach to top task analysis. It is far from perfect, but it is quick to do and proves massively helpful in helping to prioritize what matters to users in your app, information architecture or landing page. If nothing else, it is an excellent alternative to stakeholders just making an arbitrary decision.
The whole exercise, including analyzing results, can be done in a few minutes and depending on your audience, you can get survey results in less than a day.
The only real challenge of the approach is recruiting participants. If you are trying to reach a new audience and need help with recruitment, then you might be better off using Gerry’s approach and replacing my app with a survey tool that helps with recruitment, such as Pollfish. The only downside is that you will need to run one survey to collect ideas and then another to vote on those ideas, so slowing the process down somewhat.
