How you define ‘good enough’ is contextual – but not in the way you might think. Almost all problem solving has an implicit nature of diminishing returns. Progressively more investment results in progressively less benefit. How do you know when to stop making those investments? The best way is to combine your outside-in assessment with an inside-out analysis.
I have a bit of a history with the concept of ‘good enough’, from satisficing in sprint planning (2008) to Kano analysis for product managers (2009) to trying not to miss the mark in problem solving (2012) to understanding intent and satisfying customer needs (2014). I explore the concept from different perspectives in product workshops I run, and in my Competitive Product Strategy class. In this article, I want to dive into combining internal and external perspectives to make investment decisions.
I first learned about diminishing returns as a formal concept in economics class. The concept was easy to grasp, as there are so many practical examples which match intuition. An example I use in my class applies utility theory from economics to eating bananas.
Imagine you are hungry, and would benefit from eating a banana. The benefit you get from eating that banana is a reduction of your hunger. How much benefit would you get from eating a second banana? You get less benefit from the second banana than you got from the first banana. This makes sense – you’re less hungry than you were before. A third banana would provide even less value than the second banana. The amount of benefit which comes from each additional banana diminishes.
The reason this works is because the problem you are trying to solve is getting progressively smaller. With each banana you eat, you become less hungry. Therefore the value to you of becoming less hungry becomes smaller. Starvation is a large problem and the value of solving it is large. Feeling a bit peckish is a small problem and the value of solving it is small.
We have to think about diminishing returns in product management because people keep assigning value to the banana. The value is in solving the problem.
Laws of Physics
Every problem reaches the point where continued investment to reduce the size of the remaining problem is not worth it. And this comes from two insights I think of as ‘laws of physics’ when it comes to problem solving. These two laws act as a double-whammy, forcing you to find a definition of ‘good enough’ for every problem you choose to solve.
- Diminishing Returns: It takes progressively more objective improvement to realize a subjective benefit.
- S-Curves: Objective improvement becomes progressively harder to achieve.
Diminishing Returns
These graphs highlight the nature of the model of diminishing returns. We start with a goal for how much we want to increase value for the customer. As a function of how much problem remains, we identify how much problem-solving is required. If the customer is very hungry (the problem is large), then we can create disproportionate value for the customer with a modestly sized solution.
If however, there is not much remaining problem, the customer is approaching being satisfied, then conditions have to get notably better to create an equivalent increment of value. We have to make a bigger dent in the problem to achieve an equivalent increase in utility for the customer.
Problems rarely have binary characteristics, they tend to follow what Kano called “More is Better.” He was looking specifically at feature-level analysis, and when abstracting up to utility or value, I believe diminishing returns is necessary to acknowledge.
S-Curves
S-curves are a model for considering the costs associated with actually making something which solves problems. In the diminishing returns graphs above, we are looking at the impact of somehow solving the problem. In the S-Curves, we are looking at what is involved in solving the problem. We’re double-clicking into the “somehow.”
Think about your ability to perform some function as a capability. The way you should measure a capability is always context specific, as a function of what you’re trying to accomplish. What you can do, to avoid that complexity and keep going is simply think of it this way – you solve a problem (with a capability) by improving the capability. You could invert this as well, to say the better the capability is, the less problem there is remaining to be solved.
You could judge a city’s traffic system in terms of how many people can commute from outside the city to a stadium for a sporting event. The more people you can get to the event in a given timeframe, the better the system is.
When coming at this from the point of view of solving a problem, you would define the problem with something like this problem statement.
The Problem of… The stadium averages 80% capacity for sporting events due to traffic congestion preventing people from attending
Affects Whom… The owners of the sporting team and stadium, and city residents who benefit from tax revenue
The Impact of Which is… 20% of seats are unfilled on average at events, resulting in $100M in lost revenue per year
The Benefits of a Solution are… increasing revenue by $50M per year through reducing congestion enough to convince an additional 12% of people to attend
In this situation, you might define the capability as being better when it has higher throughput, or you might define it when people’s perception of congestion is improved. Let’s go with throughput for now, because it is easier to imagine how those investments work. The main idea about an S-curve is that the better you get, the harder it becomes to get even more better.
It could be hard to add a lane to an existing road to increase its capacity. Adding another lane becomes even harder. Roads tend to be built with easements and shoulders, allowing for “easy” expansion in the future. But continued expansion beyond what was anticipated can require relocating utilities, or even businesses and structures. The more you improve things (capacity), the harder it becomes to improve them. The top of an S-curve looks like diminishing returns, because it is the same pattern, but considered from a different point of view.
What makes an S-curve different is how it begins.
There is, in most of the situations I’ve encountered, some initial work which must be done to lay the foundation for building something meaningful. A software developer may have to invest in their development environment, an organization may need to build their pipeline for deploying software. A city adding high-capacity transport with a light rail system has to build a lot of infrastructure before even the first train can run.
Those investments enable future investments to complete the work of delivering something which is potentially valuable. A manufacturer who assembles drones may need to make all of the logistics investments to get all of the drone components into one location before the work to assemble those drones can begin. Once that up-front work is done, the model highlights where modest incremental investment can yield large increases in capability.
Once you’ve built the rail line, the incremental expenses of adding trains or train-trips to the schedule yield large increases in capacity. This is where the bang for the buck really pays off. There is real potential value in making these investments, tied to a measure of the capability as a presumed source of value. Many engineering teams still struggle with the decoupling of potential value from realizable value.
Escaping this struggle requires combining the S-Curve (familiar to engineers) with the diminishing returns model (familiar to product managers).
The last portion of the S-curve is where the compounding effects put you at a disadvantage.
It simply becomes harder and harder to become incrementally more capable. Adding the initial rail line and train schedule creates an improvement in capability. As you saturate the schedule, it becomes harder to add additional trips, or additional trains, or additional tracks. The more you increase capacity, the harder it becomes to increase capacity. The crux of the S-Curve.
The Double Whammy
Now combine the increasing difficulty of becoming more capable with the decreasing incremental value of becoming more capable. This is the double whammy. These laws of physics are one of the two reasons why companies taking an inside-out approach to developing products are at a disadvantage competitively. Without considering these two principles, you are more likely to overinvest.
You are more likely to enhance the capability beyond the point where it is a smart improvement. You will make unjustified investments of time and money without yielding sufficient benefit to customers.
Conclusion
Combining these two concepts allows us to shift our thinking. We can reorient from an inside-out question of “What do we expect it to cost?” to an outside-in question of “How much should we be willing to spend?”
Every meaningful investment decision we make should anchor on the value of the investment, not solely the cost of making it.
