This Concept Changed the Way I Think About Ideation
Ideation is the process of generating multiple ideas to solve a particular design problem. Like all processes, ideation can be done well or it can be done poorly. Of course our goal is to do it well. In this article, I share a seemingly simple guiding principle for ideation and a deeper but related critical concept that changed the way I think about and approach ideation.
A Guiding Principle for Ideation
The following quotes capture an important guiding principle – the critical need to thoughtfully generate various design options before converging on just one to develop fully.
From two-time Nobel laureate Linus Pauling. “If you want to have good ideas you must have many ideas. Most of them will be wrong, and what you have to learn is which ones to throw away.” [1]
From my textbook on product development: “The best design can only be identified in the context of multiple designs.” [2]
From the Harvard Business Review (paraphrased): Generally the first idea is not the final idea. Abandon the belief that you already have a good idea when you start designing. Allow the best idea to emerge through thoughtful ideation. [3]
Successful design professionals are guided by this principle. As a result, they treat ideation differently than novices do. To the professional, this principle has deep meaning that this article begins to uncover.
If Superficially Embraced
While this guiding principle is useful, it can lead to two problems if only superficially embraced. Either problem will likely result in bad ideation.
Problem 1: The casual designer agrees with the principle, but finds it sufficiently abstract to not really do anything about it.
Problem 2: The designer who transcends Problem 1 falsely concludes that they should focus on generating a large quantity of ideas.
When one of these problems is present, the deeper meaning of the principle is lost and ideation effectiveness drops. We can uncover the deeper meaning by visualizing the ideation process in a particular way, which is described below.
Visualizing the Ideation Process
For any particular design problem, there are various ideas that could meet the design requirements. Such an idea is called a design concept. To visualize the ideation process, Imagine that a specific design concept occupies a unique point in a geometric space, as shown below [4].
Further imagine a collection of design concepts as a cloud of points in the same geometric space.
The image above shows only a small number of design concepts, but it is useful to imagine the feasible design space, which is the full set of design concepts that could feasibly meet the requirements. Because there can essentially be an infinite number of solutions to any one problem, the feasible design space is often represented as a solid nebulous blob in the geometric space.
Within the feasible design space, there are many specific feasible designs. All the infeasible designs such as the ones that break the laws of physics, or those that meet all the needs at zero cost, for example, lie outside the feasible design space. It is common to consider the feasible and infeasible space as the design space.
Design Space Exploration
Visualizing the design space as described above, unlocks a powerful way to think about ideation. Ultimately the goal of ideation is to identify a great design that – if implemented – would have the greatest chance of meeting the design requirements.
In the context of the design space, this means that the great design is hidden somewhere in the feasible design space, and it is the designer’s job to find it. The process of searching for and ultimately finding that design is called design space exploration or design exploration, for short. Exploration is an apt word since ideation is facilitated when those who are ideating have an attitude of curiosity, imagination, searching, and/or surprise.
Judging the Quality of Ideation
Visually, there is a relatively easy way to judge the quality of the design exploration and as a result the quality of any ideation activity. The visual evaluation is based on four attributes originally articulated by Jami Shah [5] and studied by various others:
The Quantity of concepts generated
The Variety of concepts generated
The Novelty of concepts generated
The Quality of concepts generated
Importantly, each of these attributes is visually obvious, when you know what to look for. The image below shows the meaning of Quantity as it relates to the design space.
The image below shows the meaning of Variety as it relates to the design space.
It’s not as obvious as it is for quantity and variety, but the image below shows the meaning of Novelty as it relates to the design space. One way to interpret this is to assume that the design concepts in the middle of the design space are safe – they are nowhere near the edge of feasibility. On the other hand, those near the edge are getting close to pushing the boundaries. Those over the edge are infeasible, and frankly there are likely to be many novel ideas in those regions. The main point is this: if all the design concepts are safe within the feasible design space, it can be assumed that the novelty of the ideas generated are low, while on the other hand those near the boundary, on it, or over it, are more likely to be novel.
Many designers encourage “wild and crazy” ideas during the ideation process. These are often infeasible (or lie outside the feasible design space), so why would anyone encourage them? This is a good question that can also be answered by visualizing the design space, as shown below.
Although we imagined each specific design concept as a point in the design space, the reality is that a specific design is one of many that belong to the same basic concept. For example, there is the red cargo bike, the orange one, the pink one – these are all small variations of a particular design concept. Likewise, there is the bike with a wicker basket in front, and another with a steel basket in front, and yet another with a nylon basket in front. These are also variations of a particular design concept. Variations of a design concept exist spatially very close to each other in a design space. This is represented by the circle drawn around a given point. The circle represents a small space wherein a collection of specific designs is found that are small variations of the design at the center. One reason why designers encourage wild and crazy ideas is that there may be a variation of that wild idea that is actually feasible. This is represented in the image above as an infeasible point whose variant circle extends into the feasible space.
It’s valuable to recognize that designers have a stewardship for the design space. Not only do the requirements they identified during the early stages of design define (in part) the feasible design space, the designer’s ability to effectively ideate will determine how well the design space is explored. A well-explored space results in a concept set that has high quantity, variety, and novelty. The image below depicts what designers should do with the design space they have stewardship over (they should explore it!).
Concept Quality
While the quality of the concept set is important, we ultimately only need one design concept to develop fully. So why all the emphasis on generating many ideas? At this point, Paulings quote should make more sense and help us answer the question.
“If you want to have good ideas you must have many ideas. Most of them will be wrong, and what you have to learn is which ones to throw away.”
What will be thrown away are low quality concepts – or concepts that occupy undesirable regions of the design space. This includes infeasible designs and undesirable feasible designs. This will become more obvious if we acknowledge that up until this point of the article, the design concepts were placed in a geometric space that we didn’t define. If all we care about is distinguishing between feasible and infeasible – and we want to maximize quantity, variety, and novelty – no further definition of the geometric space is needed. But if we want to begin evaluating the relative desirability of each concept, we will need to place the design concepts in a cartesian space relative to dimensions of interest, as shown here.
These dimensions will form the axes of the geometric space. For example, one dimension of a bike might be cost, and another dimension might be weight, as shown below. Given these dimensions (against which all concepts can be compared), one would likely find lower-cost lower-weight bicycles more desirable than higher-cost higher-weight ones. In this case then, it would be important for the design team to focus more on generating design concepts in areas of desirability relative to the design space.
Specific design concepts that lie on the boundary of the feasible design space – in the quadrant that optimizes desirability – are called Pareto optimal designs. The main characteristic of a Pareto optimal design is that in order to improve desirability in one dimension you must give up desirability in another. The set of Pareto designs is called the Pareto frontier and more aptly: the trade-off curve. For two objectives that the design team wishes to maximize, the trade off curve would look like this:
Of course the regions of the feasible design space that will be more valuable than others will be determined based on the goal for each objective, be it to maximize or minimize them. The image below indicates where the trade off curve would be for this design space, depending on the goals for the objectives.
Given that it is more valuable to explore the region near the tradeoff curve, the designer’s stewardship for the design space will more efficiently be met when explored as shown on the right versus the left of the figure below.
How to Use This Notion of Design Space in Your Work
A non-computational approach: Understand that when you are ideating you are exploring the design space. You are either doing it well, or not well. Remember, your job is to do it well, which means to generate a set of design concepts with high quantity, high variety, and high novelty – with thoughtful focus in areas of high desirability (areas near the trade-off curve). If you are on a team and the ideation is getting stale, suggest deliberate jumps in the design space: “lets try to create some ideas that super low weight, regardless of cost.” Or “all of our ideas have both wheels in a line, let’s get some ideas where both wheels share the same axis.” By doing this you are encouraging the team to explore a different part of the design space. When allowing an image of design space to influence your work early in the development process, it will mostly be a notional or theoretical application, since there is little math to make it computational. That’s ok and still incredibly useful.
Computational: The design space is very real and very discoverable when you have models that define the design requirements and design desirability. Various methods exist in the literature for doing this, including work I published as part of my PhD and in many subsequent journal articles. Below are a few examples of what a computationally derived design space looks like.
From [6].
From [7].
From [8].
Concluding Thoughts
I’m not exaggerating when I say that the notion of design space exploration has changed the way I ideate. After having learned about design spaces, I view its exploration as a fundamental responsibility I have for any project I am on. It has helped me curb feelings that “I already know how to solve this”, and has allowed me to more thoughtfully look for the best design concepts among the many that I know are present in the design space. Without a doubt, in my experience, a thoughtful exploration has always resulted in a better outcome than jumping to a solution.
References
[1] L. Pauling, “Linus Pauling Quotes”, BrainyQuote, 2001-20, https://www.goodreads.com/quotes/7225259-if-you-want-to-have-good-ideas-you-must-have
[2] Mattson, C. A., and Sorensen, C. D., Product Development: Principles and Tools For Creating Desirable and Transferable Designs, 2020, Springer, Cham.
[3] L. Nordgren and B. Lucas, “Your Best Ideas Are Often Your Last Ideas,” Harvard Business Review, https://hbr.org/2021/01/your-best-ideas-are-often-your-last-ideas#:~:text=Prior%20research%20has%20shown%20that,reaching%20your%20most%20creative%20idea., January 26, 2021, accessed 22 April 2024.
[4] Bard, a large language model from Google AI [Personal communication, April 22, 2024] was used to generate all of the bicycle images used in this article. See the full AI conversation here: https://gemini.google.com/app/301a507a86128cbe.
[5] Shah, J.J., Kulkarni, S.V. and Vargas-Hernandez, N., 2000. Evaluation of idea generation methods for conceptual design: effectiveness metrics and design of experiments. J. Mech. Des., 122(4), pp.377-384.
[6] Mattson, C.A., Pack, A.T., Lofthouse, V. and Bhamra, T., 2019. Using a product’s sustainability space as a design exploration tool. Design Science, 5, p.e1.
[7] Richards, D.C., Salmon, J.L., Dickerson, T.J., Mattson, C.A. and Neff, W.J., 2023. A decision support system for multi-stakeholder exploration of the airship design space. The Journal of Defense Modeling and Simulation, p.15485129231164416.
[8] Mattson, C.A., Lofthouse, V. and Bhamra, T., 2015, August. Exploring decision tradeoffs in sustainable design. In International Design Engineering Technical Conferences and Computers and Information in Engineering Conference (Vol. 57113, p. V004T05A046). American Society of Mechanical Engineers.
