It’s almost impossible to create an intuitive website without knowing your user’s goals and struggles along the way. But how do you find out what they are? Luckily, there is an effective way to do that — task analysis. By following this article, you will master task analysis and obtain the knowledge you need to design an efficient and user-centered product.
What is task analysis?
Task analysis broadly means understanding of the user’s task. It is a combination of understanding the user, understanding their task and understanding their environment (Courage, Reddish & Wixon, 2007). It encompasses a broad range of techniques from observations of the user in their natural environment to documenting how the users perform their task on our existing system. Performing a task analysis will leave you with detailed understanding of the task sequence, its complexity, environmental conditions, tools, skills and information the user needs to perform the task to achieve their goal (Usability Body of Knowledge). A good task analysis leads to actionable insights into user processes which can be directly applied in designing efficient user flows that liberate users from unnecessary work and delegate said work to the system.
Aspects of task analysis
These are often construed as types of task analysis, but we like to think of them as different aspects. To perform a task analysis that leads to actionable design insights, you need to understand:
Contextual task analysis
Contextual analysis means obtaining a user-centered model of the task as it is currently performed in the user’s actual environment. This enables you to understand how the product will fit the user’s environment, needs and other tools they already use. For example, to design for users in a distracting environment, you need to take interruptions into account, provide safeguards against unintentional errors and include options to pick the task up again after a delay (Mayhew, 2007).
Contextual task analysis is indispensable in pinpointing novel business opportunities – “At which point can we design technical solutions that help the user do their task more efficiently?” It also helps us design the product so that it can be seamlessly integrated into the user’s existing processes and it’s easy for new users to pick up. Lastly, understanding how users already interact with existing tools helps us design an interface that’s inherently familiar to the users and fits their cognitive models.
Cognitive task analysis
Cognitive task analysis focuses on understanding the tasks that involve deeper mental processes such as decision-making, attention, memory, and judgment.
Hierarchical task analysis
Hierarchical task analysis studies user behavior by breaking the high-level tasks down into smaller subtasks. The information is then visualized in a form of a diagram that describes the steps taken to accomplish a certain goal.
When to do task analysis?
The short answer: ideally during all stages of the design process, but it’s most beneficial at the beginning (Courage et al, 2007). If you follow the Design Thinking process, incorporate it into the Empathize and Define stages. Doing task analysis at the beginning will ultimately save time and money during the later stages. Understanding how users work makes the design phase move much more quickly, it helps prioritize the features, and saves on testing, as the design will be more informed and fewer iterations will be needed. However, task analysis can be just as successfully applied to updating an existing project and can still drive your updates to be more user-centric.
Before starting task analysis, decide on the scope and granularity – i.e. how much time you have, what user population you want to cover, how many types of tasks and in how much detail you want to specify them. For example, if you are designing a collaboration platform, you may be interested in a larger picture – understanding how work moves from person to person and the users’ general jobs. On the other hand, if your product is targeting single users who don’t interact, you may want to start with the target user’s main goals and sub-goals and move down to the breakdown of specific steps they take to achieve these goals.
Developing a brand new product
To develop a cutting-edge solution on the market, start from scratch and understand user’s goals, mental models and tasks in their natural environment. ‘Empathize’ to understand your users: Who are they? What information do they have and lack? What mental models do they have of the the activities that your product covers? And most importantly – what are their goals?
If you have the time and resources, bring the research to users by conducting site visits. Observing the users in their natural environment will allow you to document their steps and decisions as they solve tasks. Supplement your observations by asking questions about their goals and reasoning and how they may change their steps and decisions under different situations (Courage et al, 2007). You can use this info later to draw flow diagrams of their actions (see analysis). If you have the opportunity, show your diagrams to the users and let them explain and correct your interpretations.
Interviewing users is the next-best thing to direct observation. To start, make the interview behavioral rather than attitudinal – get them to walk you through their process and explain their decisions. Ask them to show you the artifacts and tools they would normally use and let them walk you through how they would use it. Artifacts could be e.g. a calendar, notes, paper form – anything they already produce to help themselves in the task. However, don’t forget to ask users about their goals and attitudes as well – understanding these is just as important and will help you in detecting inefficiencies in preexisting processes and in designing helpful solutions.
You can even do this in the form of a focus group with multiple target users. Moderate their discussion to reach consensus on what the task steps look like, what kind of decisions they have to make along the way and what kind of goals they are achieving.
Controlled study. If you don’t have the time to conduct visits, bring the user to you – physically or remotely. This is great for detailed procedural analysis – getting down the specific steps and decisions needed to complete the task.
Upgrading an existing system
To understand how users already perform a task with a preexisting system, you can even conduct an almost-usability-study, which will collect the data on a specific system. Courage, Reddish and Wixon (2007) recommend using logs of user activity to track your users, their click paths and actions completed (e.g. purchase or download). Conducting user testing remotely, for example with our tools, will help you reach a diverse, international population, which is important especially if the user’s mental models of the task differ between cultures. It also helps to reach people without prior relationship to the system – not only those you already collaborate with, but also potential users and users of competing products.
It is beneficial to lay out the data in a flow or sequence diagram (see below) documenting how the information moves from system to the user. For preexisting design solutions, the diagram will often be surprisingly elaborate and messy. This allows you to see the points where you can make the task flow more efficient for the user and define actionable design recommendations.
Putting the Analysis into Task Analysis
When analyzing the data you collected, let your goals guide you so that you can provide actionable insights to the problem at hand. Pick the right task scope and granularity, whether you are designing a large collaborative system from scratch or iterating on a simple interface for an alarm clock.
A variety of tools can help you to make sense of your data. If you are just starting generative research to prioritize features of a new product, use affinity diagrams mapping user’s needs, goals and preferences. To understand your target users, craft rich persona descriptions that contain user background, goals, needs, knowledge and environment information. Moving closer to the task itself, you can write scenarios – short stories starting with the user’s situation and describing the steps, tools and artifacts the user uses to arrive at happy ending. However, the ultimate method in hierarchical task analysis are the diagrams.
Flow diagrams are the most important method of task analysis. They document the core of the task – how user information and artifacts flow through the system – and they illustrate the dependencies. Depending on your analysis goal, diagrams can easily incorporate the details of the task, including the goals, decisions, tools and interactions with other individuals.
For an even more detailed view of how the information flows between the user and system over time, you can plot the detailed procedure in a sequence diagram. This is very helpful for decision-making in information architecture and design, because it describes all the steps in the system – including menu buttons, and pop-up dialogs. It also lets unnecessarily complicated information exchange between the system and the user pop-out. Larry Marine suggests implementing some of the functions of sequence diagrams by color-coding the tasks in a flow diagram, so that you can immediately see which actions are done by the user and which actions are done by the system.
Usability Body of Knowledge lists these steps to arrive at a task diagram:
- Pick a specific task to analyze
- Break this task down into 4 – 8 specific subtasks which cover the whole task
- Draw the subtasks as a layered diagram
- Decide upon the level of detail into which to decompose: are you only interested analyzing high-level tasks (e.g. during generative stage) or making precise procedural diagrams (e.g. when examining interactions in existing systems)?
- Continue the decomposition into smaller steps until you reach your desired level of detail. It’s helpful to accompany the diagram with a written description.
- Validate the analysis – present the analysis to someone else who has not been involved in the decomposition but who knows the tasks well enough to check for consistency.
Larry Marine adds one step crucial for designing novel, efficient flows: optimizing the task. This means making the task easier for the user by eliminating steps and delegating them to the system. Here, it’s crucial to take into account what the user does and doesn’t know – our brain works better and faster when we have fewer choices, so don’t make users think. Since the company you are designing for probably knows more about the subject matter than the user, you can help the user lower decision fatigue and need to search for information by implementing informed defaults and helpful suggestions.
Task analysis example
To make it easier for you to understand the process, we are going to walk you through a simple example of UX task analysis on an existing system. In this example, we will be analyzing Marco, who loves online shopping and wants to buy a new pair of jeans for the summer.
So, Marco’s goal is going to be
“Purchase the jeans from an online store.”
Subtask 1: Find the jeans on the website
- Open the website
- Go to the “men” clothing section
- Find the jeans on the category page
Subtask 2: Add jeans to the shopping cart
- Click on the product detail page
- Choose the size
- Click “add to cart”
Subtask 3: Proceed to checkout
- Go to the “shopping cart” page
- Click “checkout”
- Register/log in or choose guest checkout option
Subtask 4: Checkout
- Enter delivery info
- Choose the payment method
- Enter billing info and card number
- Review the purchase and pay
Here’s an example of what the diagram would look like:
Looking at the task diagram, how can we optimize the task flow to make it more efficient from Marco’s point of view? Can we reduce the number of steps, decisions and the information he needs to know? This is where your own cutting-edge design solution comes in.
Less is more
Implementing UX task analysis in your design process is always a good choice. It helps you focus on the user and create a seamless and efficient UX for whatever product you’re working on. And when you’ll need some help doing user research at any stage of the process, go ahead and register for a free trial of our usability testing tools.
References and further reading
Courage, Reddish and Wixon: Task Analysis. In Human-Computer Interaction, 2007
Marine: Task Analysis: The Key UX Design Step Everyone Skips,
Mayhew: Requirement Specifications within the Usability Engineering Lifecycle. In Human-Computer Interaction, 2007
Usability’s Body of Knowlege: Taks Analysis, http://usabilitybok.org/task-analysis