When it comes to designing a new product, there are a number of decisions that need to be made on the basis of questions you answer among your team. Some of these might be things like “Why do we think our customers want these solutions?” or “Are there potential impacts of our product?” While you would like to have an objective answer, many of these require you to lean into what your gut tells you.
To do this, you’ll likely have to make value and viability assumptions. This article will help you understand the different types of assumptions, explain how to figure out which ideas will be successful, and provide you with methods for testing.
An assumption is something (conviction or belief) that you believe to be true or certain, with no proof of it being true. In the product world, an assumption must be true for our ideas to be successful. Because of this, assumptions can be dangerous and costly if they turn out to be incorrect.
That said, ultimately assumptions are often at the center of product decisions. Rather than cowering from them, you can develop an understanding of the different types and work on strengthening the accuracy of your assumptions.
In product development, assumptions usually can be categorized into five different types.
With a viability assumption you answer questions like:
This assumption considers factors like market need, competition, and potential revenue streams.
In this case, you’re answering questions like:
This type of assumption focuses on understanding customer needs, pain points, and motivations. It also explores whether customers will be willing to pay for the solution or adopt any necessary behaviors.
Here, you would answer questions like:
The focus is on the technical expertise and resources required to create the solution. It considers whether existing technology can support the product and if the development timeline is realistic.
When it comes to usability assumptions the conversation would focus on:
This assumption analyzes the user interface (UI) and user experience (UX) to ensure the solution is intuitive and user-friendly.
With ethics, you would deal with questioning like:
This assumption ensures the solution aligns with ethical principles and avoids causing harm to users, society, or the environment.
To figure out which ideas will fly versus flop, you need to know how to write assumption statements.
Then, in step two, you take your assumptions statements and map them on a two by two diagram. You should develop testing for the assumptions that fall within the top right quadrant. As a final step, you analyze the results and make a decision.
Now, let’s dive further into each step:
Does it matter how you phrase an assumption? Well, it does matter a lot because the phrasing will affect how easy the assumption is to test.
A couple of things to keep in mind:
Then, you’ll take all of your assumption statements and map them in a two by two matrix.
To understand the risk and uncertainty of an idea you should ask, “What are the things that must be true for this idea to work/to be successful?” This will allow you to identify all five assumption types underlying a business idea: viability, value (desirability), feasibility, usability, and ethical considerations.
Before diving into any development, it’s important to work with valid assumptions. And, using assumption mapping can make a big difference in increasing the success of a product and its features.
Here’s the skeleton of an assumption map:
On the x-axis, you use the labels “have evidence” and “no evidence” because it all comes down to evidence. It’s about having observable data (a piece of evidence), qualitative or quantitative, to support your assumption.
On the y-axis, you use the labels “important”’ and “unimportant.” You might be thinking, “Well, we only have to write down the important assumptions, so this is redundant.” However, not everything is the most critical. By asking ourselves, “Which hypothesis, if proven wrong, will cause your idea to fail?” you can determine which are the most important assumptions.
Once you have this, focus on the top right quadrant! Your assumptions that are critical for success and yet have the least amount of evidence to support them, focus on those.
Doing assumption testing exercises regularly isn’t easy because it requires a mindset shift. It’s easy to think that if you use a product/feature, others would use it too.
You already know that ​​you make better decisions when you consider more than one option. That’s the same in product development too. That’s why you shop around when looking for a new job, a new place to rent, or a property to buy.
But at work, it’s a bit more challenging. When you hear about a user’s pain point, you usually jump right into the first solution that comes to mind based on your understanding of the world and the problem. After all, you’re problem solvers and like to go ahead with your guts and ideas.
But how do you know this is the best solution we can do if we don’t compare multiple options?
You just don’t know! Imagine if you rented a place like that. That’d be insane! But why do you tend to go ahead with the first solution anyways?
Maybe you don’t think you have enough time to consider more than one option within the constraints you have. Or, maybe you think assumption testing is time-consuming and slows down the team.
However, you can still move fast while considering options. The key is to stop testing whole ideas and just start testing bite-size pieces that your idea depends upon. Assumption testing allows you to evaluate which ideas will work and remove the ones that will not.
If you have a computer science background, the following idea will be familiar because it’s similar to setting up a unit test for your code:
Finally, you need to synthesize the results from the test.
If the data confirms your assumptions, you can move forward with product development. If the data falsifies/contradicts your assumptions, you can pivot or stop focusing on that solution:
The example above highlights how assumption testing with features like personalized recommendations can help e-commerce websites validate or falsify customer behavior before a large investment is made in platform development.
You might be thinking, well, you don’t always have that much time and resources to develop a feature for A/B testing, right? If you’re thinking that, I’m glad, because I want to give you four more methods with low investment.
You know that untested assumptions about an idea, or feature can lead to unsuccessful or even disastrous results. So, how can you test assumptions to increase the chance of success? The answer is ask, show, offer, and deliver — a concept developed by Keith Hopper.
These represent the ways you can interact with your users to validate or falsify your assumptions and fall along a spectrum based on your level of certainty/evidence about the assumption.
Ask and show are ideal for the early stages when the concept is still evolving and when you don’t have lots of certainty about the idea. For example, when you have lots of ideas about what might be valuable for customers, but you have too many options and don’t know what would offer the most value.
Then, offer and deliver are more appropriate when you have a clearer direction and when you have more certainty about the direction, but you need to validate your idea/assumption to be sure users will use it.
The method involves structured interviews with individual customers. Here, the goal is to understand what truly matters to them. By asking specific questions, you can identify customer pain points and opportunities for innovation and what is meaningful for users.
When you’re at the ask stage, you can cover questions such as:
So, if you don’t have answers to these but have a bunch of assumptions about them, this is where you can start.
Go into the session with structured questions, so you don’t end up with casual conversations.
Always try to dig deeper, a couple of layers down, to have valuable data.
During assumption testing, the quality of the information you gather is just as important as the information itself. This is where the concept of evidence strength comes in. Just like in scientific research, not all evidence is created equal, not all evidence has the same strength.
Some insights provide a stronger foundation for decision-making than others. The ladder of evidence framework helps you prioritize user insights based on their reliability:
The weak evidence is at the bottom of the ladder and usually, what users say they do or want. These include surveys, interviews, and focus groups. While valuable for gathering initial ideas and understanding broad sentiment, these insights can be susceptible to biases such as social desirability (users reporting what they think is expected) and memory limitations.
As you ascend the ladder, the evidence becomes stronger and more actionable. Here, you find evidence based on what users do, providing a more objective picture of their behavior and needs.
Examples of strong evidence methods include:
My answer is as early as possible AND definitely not at the end of the product development process.
Here’s why.
Lots of product people go to designers and researchers saying, “Hey, could you validate this [idea] for me/could you run a quick user study, you know, just to validate this assumption?”
That’s user-centered performance. It’s too late to matter. It’s just a performative exercise regardless of the outcome.
When people do this, they aren’t interested in being wrong. Not at all. Because it’s too late for that. The feature is very likely to have already been scheduled for release. What people want is to check the box. And yes, it might be well-intentioned, but it’s just too late. So, do it as early as possible.
If I can leave you with one thing, I want to leave you with, “Don’t validate! Falsify.”
Have a mindset shift, and decide that you want to be wrong with my assumptions! You want to do assumption testing that shows you how wrong my assumptions were in different ways.
That’s how you should approach insights, assumption testing, and research in general.
But lots of people don’t approach research this way. They want to validate, they want to be right.
So, be ready to be wrong! That’s where real learning lives.
LogRocket lets you replay users' product experiences to visualize struggle, see issues affecting adoption, and combine qualitative and quantitative data so you can create amazing digital experiences.
See how design choices, interactions, and issues affect your users — get a demo of LogRocket today.
In this article, we’ll explore the state of no-code design with AI. We’ll then discuss Fronty, its features, and how to use them practically by converting several Figma design frames into live webpages.
As we approach an inflection in the evolution of user research, it’s important to unpack its expansive role in catalyzing innovation.
Alt text describes images for visually impaired individuals and broken images that don’t load correctly. AI can help generate this text.
Teens are an exciting group to design products for, but we need to consider age, lifestyle, and generational differences.
Learn more →
