Productonomics is my own little invention that applies an “economic way of thinking” to technology product development, and the Fallacy of Sunk Cost is, IMO, a highly applicable economic principle. Proper understanding and application not only benefits the products we make, but it benefits morale and company culture. To give credit where credit is due, Steve Bronstein, a former colleague and a current friend, is really the one who stuck this phrase into my head and made me see the light (he very effectively argued against a position I held by dropping this science on me). I realized at that point how crucial the concept is in decision-making and strategic thinking, but I also realized how counter intuitive it is.
What is it?
The principle is easy enough to understand, and it can probably be summed up most succinctly by the “dont throw good money at bad” idom, although that’s not all there is to it. A short definition that I like is “Once the money (or time or effort) is gone, then it’s gone. There’s no point in worrying about this” (from Plonkey Money). This may be oversimplifying it a bit, but there is a ton of stuff out there on the web that can explain it better than I can (links below), and I want to focus specifically on how it can help product development. Most of what you find out there is associated with money management/investing, but the principle holds for any decision-making situation where spent resources come into play.
How is it useful in Product?
Our decisions now only affect the future and have no relationship to the past – - that’s the jist, and logically, it is not hard to understand this. If I buy a ticket to a bad movie and I can’t get a refund, sitting through the movie doesn’t re-acquire any of the value I have lost – I’m better off walking out and doing something that gives me more utility/return – that’ s the rational thing to do anyway. But now let’s consider a situation that many of us have been in – the team has been working on a roadmap feature-set for months… discovery, design, prototyping – you name it. Its been all consuming and buckets of sweat and tears have gone into it. When the assumptions were made and the numbers were run months ago, it seemed like a home run, but just recently – we either realized we were wrong and nobody wants this thing, or a bunch of external factors changed (and when doesn’t this happen?) or both. What changed isn’t important – - competitive environment, economic environment, regulatory environment, technology stack, legal framework, etc., etc., etc. – - the point is that what to do NOW is what’s important, and what was decided 6 Months ago, while perhaps relevant for knowledge, is not relevant in the decision calculus going forward.
It comes down to two positions:
- We should just kill the entire project and do something else.
- We have put too much work into this and gone too far to quit now – we will have lost all of that work.
Well, assuming that the core assumptions are accepted to be true – (that we were wrong or the landscape is no longer receptive) - choice number 1 is the correct choice. Now, its important to make the distinction between the “we were wrong” vs. “things outside our control changed”, because each situation presents its own challenges to the people on the team, but let’s be clear that in either case – bullet #1 is the way to go. All that hard work is in the past, and without a time machine, we can’t retrieve the resources. All we can do now is make the best decision possible that defines our path from today onward, so continuing to throw resources at something we all feel is doomed is a fools errand.
What about the “Zen” Part?
Well, that all sounds fine and dandy and rational and what-not, but here’s the rub… the rational decision is likely to be unpopular, and if the culture of the company doesn’t allow for failure to be acceptable, there is going to be political face-saving shenanigans, finger pointing, and dogmatic group-think takeover. This quip from the Skepdic Dictionary says it all:
To continue to invest in a hopeless project is irrational. Such behavior may be a pathetic attempt to delay having to face the consequences of one’s poor judgment. The irrationality is a way to save face, to appear to be knowledgeable, when in fact one is acting like an idiot.
Ouch, but true… and I’ve been there… oh, have I been there. What’s the answer? Well, that’s where the “Zen” part comes in – its really about the company culture from the top down. We can’t be afraid of failure or creative destruction (we’ll be talking a lot about ‘cd’ and Schumpeter in the future), and this is the essence of the tech start-up, is it not? One of my favorite business books, Getting to Plan B, is devoted entirely to this concept: adapt and overcome. Well, you can’t adapt if you don’t make mistakes and then recover from them quickly, so understanding this fallacy and injecting that understanding into the culture of the team is important. The “Zen” part is being able to be in the moment, acknowledge that a change in direction is the best course, and let go of the past – “bend like a reed in the wind” (yes, that is a Dune quote). To sum this one up in manner in which most people can relate, how about this quote (relayed by my friend and former colleague, Mike Tatum from Whiskey Media):
Would you rather be rich or right? – a very wise (and probably rich) man
Don’t Take it Too Far
Just like any knowledge, this can be misinterpreted or taken too far, so to be clear, this principal does NOT propose that we do not or should not learn from the past. Knowledge and experience, of course, should shape our decision making. Also, if I am to be fair, the product scenario that I described above is really referencing a “Sunk Cost Dilemma“, and there is a significant distinction. The “sunk cost dilemma” adds a more realistic framework for real-life — specifically introducing uncertainty and multiple events. In my example above, I said “assuming that the core assumptions are accepted to be true – (that we were wrong or the landscape is no longer receptive)“. Well, the reality is that we probably aren’t often that certain that the “we were wrong” assumption is true, and so we are only making the best decision we can with the information we have – thus, uncertainty. In a Game Theory context, these decisions are made multiple times, and if you make the decisions based strictly on a calculus of ONLY looking at open costs (not sunk costs), you actually get into a situation where the aggregate of the each “correct” decision creates a negative result over time.
As decisions are only made considering open costs but not sunk costs, each single decision is computed to be beneficial. But in the end, the overall payoff of the project is negative. While the project progresses towards disaster, the decision not to go on with the project gets more and more unlikely. The project is like a train: once it has been put on a track, it is very difficult to change its direction.
I’m gonna go out on a limb here and say that I think a fundamental understanding of the sunk cost fallacy is still incredibly beneficial, and this scenario, in a way, argues the same broad point that I am making – - momentum for momentum’s sake is not productive, and realistic, honest decision-making is good. Essentially, this dilemma is created by “too much of a good thing”, and understanding the fallacy is just another tool in the box… its not dogma. Oliver Lehmann’s “Visionary Tools” offer some insight into how to avoid this dilemma, and I find them to be good common sense tactics – nothing we wouldn’t already be trying to do anyway (short reporting cycles, defined roles, proper interpretation of data… common sense, in short).
Say It Loud, Say It Proud
The best place to start changing culture is with yourself, and this is a process. Just working in a ”but… i could be wrong” now and then is a good start. Saying “I WAS wrong, and here’s what I propose we do about it” in front of a room full of people is perhaps the ultimate test. Ultimately, if the nature of the relationships with your clients, customers, co-workers, and investors is honest, humble, and thoughtful, then saying “I’m wrong” means that you don’t really have to say “I’m sorry”.