MBA Mondays Reblog: Sunk Costs

The Gotham Gal and I made a decision recently where we had a bunch of sunk costs. It reminded me of this post and I am going to reblog it today.


Sunk Costs are time and money (and other resources) you have already spent on a project, investment, or some other effort. They have been sunk into the effort and most likely you cannot get them back.

The important thing about sunk costs is when it comes time to make a decision about the project or investment, you should NOT factor in the sunk costs in that decision. You should treat them as gone already and make the decision based on what is in front of you in terms of costs and opportunities.

Let’s make this a bit more tangible. Let’s say you have been funding a new product effort at your company. To date, you’ve spent six months of effort, the full-time costs of three software developers, one product manager, and much of your time and your senior team’s time. Let’s say all-in, you’ve spent $300,000 on this new product. Those costs are sunk. You’ve spent them and there is no easy way to get that cash back in your bank account.

Now let’s say this product effort is troubled. You aren’t happy with the product in its current incarnation. You don’t think it will work as currently constructed and envisioned. You think you can fix it, but that will take another six months with the same team and same effort of the senior team. In making the decision about going forward or killing this effort, you should not consider the $300,000 you have already sunk into the project. You should only consider the additional $300,000 you are thinking about spending going forward. The reason is that first $300,000 has been spent whether or not you kill the project. It is immaterial to the going forward decision.

This is a hard thing to do. It is human nature to want to recover the sunk costs. We face this all the time in our business. When we have invested $500,000 or $5mm into a company, it is really easy to get into the mindset that we need to stick with the investment so we can get our money back. If we stop funding, then we write off the investment almost all of the time. If we keep putting money in, there is a chance the investment will work out and we’ll get our money back or even a return on it.

Even though I was taught about sunk costs in business school twenty-five years ago, I have had to learn this lesson the hard way. Most of the time that we make a follow-on investment defensively, to protect the capital we have already invested, that follow-on investment is marginal or outright bad. I have seen this again and again. And so we try really hard to look at every investment based on the return on the new money and not include the capital we have already invested in the decision.

This ties back to the discussion about seed investing and treating seed investments as “options.” Every investor, if they are rational, will look at the follow-on round on its own merits and not based on the capital they already have invested. But the venture capital business is a relatively small world and reputation matters as well. Those investors who make one follow-on for every ten seeds they make will get a reputation and may not see many high quality seed opportunities going forward. Our firm has followed every single seed investment we have made with another round. In most cases, those investments have been good ones. But we have made a few marginal or outright bad follow-ons. We do that for reputation value as much as anything else. We measure that value and understand that is what we are doing and we keep those reputation driven follow-ons small on purpose.

When it is time to commit additional capital to an ongoing project or investment, you need to isolate the incremental investment and assess the return on that capital investment. You should not include the costs you have already sunk into the project in your math. When you do that, you make bad investment decisions.


#entrepreneurship#life lessons#management

Comments (Archived):

  1. Rick

    Do any of the sunk cost losses get reduced via tax deduction or other benefits?

  2. Twain Twain

    I’m wondering how founders need to mirror this too. If we imagine sunk costs as a pool, founders need to spend enough to be able to swim and do some diving in it but not spend too much that makes the company drown.

    1. Joe Cardillo

      Interesting, was thinking about that exact thing reading Aaron Harris piece the other day (… )…obviously he’s talking mostly about communication there but I wondered how much founders and VCs talk about sunk costs, particularly when facing major change in product direction.

  3. William Mougayar

    When you “do that for reputation value as much as anything else.”, doesn’t that give a negative signal to the startup that you’re less of a believer? And how does the “other” lead feel then?

    1. Matt Kruza

      I would assume fred and his partners don’t tell the companies he is doing it to save face…correct me if I’m wrong Fred. Basically if there are 10 companies and 8 they believe in they say they believe in all 10… right? Can’t imagine telling the other 2 here is a pity / our reputational saver investment..

      1. William Mougayar

        well, the outside world doesn’t know it, but the company itself will know if the follow-on is a reduced amount.

  4. Chimpwithcans

    Chronic gamblers are always trying to recover their sunken costs. This approach makes each investment less of a gamble, more of a considered decision – that’s my take!

  5. mike

    Biggest issue for VCs IMO. It’a a tug between ego, loyalty to company, and the right decision for your investors …. and the former typically wins.

  6. WA

    Great commentary on weighing an anchoring bias. Too much ocean to ignore a good tide out. Thanks.

  7. LIAD

    was having this exact discussion with someone this past weekend about a real-estate investment which wasn’t going to plan.when you’re not directly involved intellectually rationalising sunk costs is the easiest thing in the world, but when it’s your effort, pride and treasure at stake its a whole different story.Sunk Cost Fallacy is the quintessential Emotion over intellect

    1. fredwilson

      our sunk costs that led me to reblog this are on a real estate project

  8. MikeSchinkel

    I learned about sunken cost in Engineering Economy class at Georgia Tech in the 80’s. In hindsight, probably one of the best lessons I got during my entire degree.

  9. Sebastien Latapie

    I see this all the time in industry. Many of our clients start these innovation projects, invest time and money in them, and keep giving them a follow on budget for the next year even when there hasn’t been any meaningful progress. People feel invested and committed and are scared to offend people.Instead of being outright killed, they linger and become zombie projects – no life and slowly going nowhere

    1. Vivek Kumar

      Sometimes meaningful progress is in learning – which could be a hard thing to measure.

      1. Sebastien Latapie

        Very true, and that is something that should be encouraged no matter the endeavor. Instead of beating a dead horse, these zombie projects should be terminated and followed by an after action review to extract all the best learnings!

  10. JamesHRH

    A great post would be – things I was taught at school that made sense only to find out that ACTUALLY HAVING TO MAKE THAT DECISION was totally different kettle of fish.So many people “know what to do” when they are talking to you – or lecturing – but are unable to do it when the situation arises for them. Every criminal law prof I ever had practised criminal law for less than 5 years……..Great post.

    1. LE

      I was thinking about this with relation to “case studies” in business schools. Case studies are really just a way to get students talking about the various issues surrounding business decisions. Kind of like kindling. They aren’t supposed to be actual advice simply because the actual nuance [1] of decisions can’t be expressed in a case study.[1] ie life is analog not digital it’s not yes or no but “it depends”.

    2. Twain Twain

      Also known as “back seat drivers”. Here is the thing, though, SOME investors are phenomenal back-seat drivers and can help founders avoid the potholes. Some simply don’t have the frames of reference / domain interest and expertise to help (even when they like and believe in that founder so they’d rather give the opportunity to someone else to steer that founder). Meanwhile, some back-seat drivers just aren’t worth their weight and don’t pull their weight to move the invested-in company forward.I know from times when I’ve consulted rather than made operational decisions that the emotional investment and value is different.Waiting for accounting sunk costs to factor in those emotional investment and values!

  11. Supratim Dasgupta

    Fred,even if the investment is considered ‘sunk’ the equity is still considered intact right? And the follow on investment brings in ‘additional’ equity? Am i correct? What happens when a company pivots or restarts afresh ? Is the sunk cost and equity both written off?

  12. Ed B

    Its one thing to manage Sunk Costs when they are primarily dollars allocated in a diversified portfolio…its another thing entirely when its primarily your time in a completely non-diversified career.

  13. pointsnfigures

    Huge psychological block and it’s often hard to be unemotional. Sometimes it helps to write it down and look at the numbers on paper. One thing to remember is keep your discipline, and apply the same principles to every situation-even though they may be different.The classic business school case is a steel company in Britain. I can’t find the case online, but they didn’t approach sunk costs or opportunity costs the correct way and wound up building a plant that was never used.In USV case, I would imagine that the decision to invest in seed means that capital is also carved out for the next round. Don’t know what happens after that. With angel investors, it’s hard to know if they don’t follow. Other things may be going on in their lives that make them decide not to follow. You are correct, funds are different.True story. I invested in a seed round of a startup. I invested in the option round. The company wasn’t doing well. Their hypothesis failed. The CEO was burnt out, and some investors found a new CEO to take over. But, it was going to take another round of financing that was basically another seed round.I could have invested in that round. But I didn’t. Initial investments are always sunk costs and you have to determine new investments round by round. Here was my thought process:1. The risk/reward of the company is exactly the same as the first round-and I have evidence of the problems in execution which I didn’t have prior to my first investment. (Bayesian probability)2. I can invest-and not take a dilution hit OR, do nothing with the same risk/reward, accept the small dilution and free ride on the new investors that are putting money in. If the company makes it, there will be another round that I will be able to invest in. (not often the case, but it was here)I took the dilution. Company failed.

    1. Richard

      One counter example is if time 2 – time 1 events are not independent, but instead are correlated

      1. pointsnfigures

        yup. then you are making the entire decision in time 1. no choice in time 2

        1. Richard

          You might even have a compelling reason to invest at time 2, if the poor performance event at time 1 is negatively correlated with time 2 performance.

  14. Matt Kruza

    As other comments here mention, the principle makes sense, but mostly in terms of financial dollars. Doesn’t make sense in terms of time spent. 1. Humans are emotional (which is mostly good) so if you have invested years of your life into something not going to give up on a whim.. 2. It is very hard to know if its really a sunk cost or building a base (the whole 5 year bamboo tree analogy 3. The whole Malcolm gladwell 10,000 expertise concept. In terms of time, “sunk cost” may just mean paying your dues. Also, the way I look at it is (and this applies 100% to what I am doing now), is there are literally few if any humans on this earth who have as much invested / risked into what I am doing right now. Sure, doesn’t mean I will succeed, but what it does mean is that no little obstacle will phase me compared to the vast majority of my competitors. I am just getting that fly wheel moving. 🙂 So when it is human capital / experience that may seem like a “sunk cost”, it actually could be the strongest, most defensible human capital skill and motivation that will make the difference for your business

    1. Vivek Kumar

      I had similar thoughts. Some times “sunk cost” == education and learning.

  15. Richard

    Reframing sunk costs as an “option” for future investing removes the sunk costs fallacy. Example: Buying a ticket in April for a mets game for the last home game of the season as an option that is will be the game that determines if they get a wildcard birth.

  16. Alex Murphy

    It is amazing how much money & time is lost because one feels “pot committed.” You see it in Vegas all the time, that’s why their buildings are so big!!There is an example though where it should be considered. If you have invested $500k into a company and hold 20% of the business, and you reasonably believe that another $500k will get you to the next stage, and this new investment will give you 20% of the post investment share, making your total stake 36%, then you are in fact giving yourself better odds than if you put that $500k to work in a different investment.All things being equal, which they never are, you have more upside potential following on with the initial investment. This should obviously be balanced against the fact that the initial investment didn’t work out as planned, and there is not guarantee that the new plan will work out any better.

    1. pointsnfigures

      Except, instead of looking at the money, look at the investment. What’s the probability of investment success-failure? Also look at the opportunity costs. What if I could invest 500k into a different business that I think has a higher probability of success and/or greater upside? Or the opportunity cost of doing nothing?

    2. Jon

      Another example of valuable life lessons being taught at the poker table. The fallacy of being pot-committed is a well documented one. It’s well understood that the chips you’ve committed as sunk costs are gone no matter what happens when you have a hand that can’t win. But that doesn’t mean that it doesn’t factor in at all does it?Consider this example:Being pot commitment is a function of pot odds and your equity in the hand. You are “pot committed” if a call would show an odds profit (but it’s usually stated only when you expect to lose a majority of the time, but still win often enough to make an odds profit give your pot odds).For example, you’re given 5-1 chance at winning a hand, and only need to win 1/6 of the time to break even.You expect to win about 20% of the time, so you call (you only need to win 18% of the time to break even).In the case of an investment in a company or a marketing campaign for example, were you be able to quantify odds with such precision, wouldn’t you recommend allowing the sunk costs to play to factor into the decision?

      1. André DeMarre

        You see this also by people in the blinds. They make bad calls when someone raises pre-flop because they don’t consider their mandatory blind bet as sunk.

  17. ZekeV

    What about the additional information that you learn in the course of sinking a cost? A prof at Stern once used the case of waiting for the F train as an example of the sunk cost fallacy. Regardless of how long you’ve been waiting on the platform, if the train hasn’t come and you’re going to be late, then you might as well hail a taxi instead. This logic never sat well with me. B/c, the longer you wait, the greater the odds of the worthless, effing slow F train actually arriving. And the only way you know this is b/c you’ve been standing there for half an hour waiting. Of course, the longer you’ve been sitting on a seed investment in an unsuccessful startup, the less likely it is that the company’s going to be a rocket ship. But at the same time, perhaps there is some informational value to this activity, and if you’re in the not unheard-of scenario where the startup has an option to make a successful pivot, the only way you’d be able to see that is from the inside perspective.

    1. Matt Zagaja

      Sunk and opportunity costs are not independent variables. Because I have the sunk cost of law school and taking the bar exam, my opportunity cost to become employed as a lawyer is significantly different from someone who has not. The problem with the train example is that it’s incomplete, you do not know when the train will arrive. If you have lots of time to spare you can afford to wait for the train and save the $20(?) cab ride. If you do not have the time to spare then you go with the more certain option. If someone had a sensor and app that told you when the train was arriving and you knew it would make it in time, then suddenly the value of that data to you is $20! Go ahead and split the consumer surplus with the app maker and you can buy yourself a drink :).

      1. ZekeV

        Good point! And also, why the hell don’t we have that universally deployed in the subway yet? Just think of the value of that info multiplied by all the millions of riders.

        1. kev polonski

          Universal deployment in subway? Because the sunk cost of a government run enterprisesis … error computing value. When you are spending “other people’s” money, normal financial prudence goes out the window.

          1. ZekeV

            well shoot, i would think that the evil bureaucrats could at least figure out how to sell the rights for this convenience to some infrastructure company, and then lease it back from them at taxpayers’ expense! but no, i must guess when the train will show up, even while my $20 burns a hole in my pocket.

  18. awaldstein

    Rereading this reminds me that when I take early stage capital the difference between F & F and from VCs or syndicates is the followon.Why it is never a great thing to have dollars mostly from small investors who can’t pony up.I talk about this with investors as we discuss the first bit of capital as a matter of course now.It’s one thing to insure that you don’t take money from someone who can’t afford to loose it, another to see if they can follow on.

  19. PhilipSugar

    This actually is a bigger issue for big companies than small ones. I cannot tell you how many big companies I have seen screw this one up.

    1. Matt A. Myers

      You mean larger investment firms or larger companies that have internal projects?

      1. Nicolas Rodriguez M

        They have a bigger ego to protect. Sometimes you dont need to abandon your sunken cost, you can recycle it and lose some value but not all of it.Also, big companies make a louder bang when they miss but smaller companies die if they miss they single shot.

  20. Thor Snilsberg

    Beware of commitment bias. As my grandfather always said, “don’t throw good money after bad.”

  21. Matt A. Myers

    If you’re not supposed to account for the sunk cost, isn’t really what you’re taking into account is current team and current product? I’ve spent maybe ~$100k over the past 3 years on I Live Yoga – including my own expenses – with the majority of that over the last year and we have traction, revenues I wasn’t expecting, a great product that our core users love, and I have a roadmap of how to move forward – improving speed and UX even further.I’m wanting to completely re-do the frontend to be a high-speed single page app, and the product that will exist will be taking into account everything I learned in developing the product so far – those previous sunk costs were educational.What I’m trying to get at is I don’t think you really shouldn’t ignore sunk costs, unless it’s a matter of using them to determine how much of a return you should be getting/happy with – and if that fits into your spreadsheet or not. The value of education is invaluable though, it’s a matter of trusting who’s being educated – the core team, founder, co-founder(s).I’m about to apply for a Canadian government funds matching program which should be enough to take current iteration to the next evolution. It won’t be enough to hire a solid full-time team that I’d want to work with expanding into where I see the product moving 2-5 years from now but going to start reaching out to potential investors to see.I’m reminded of something I saw somewhere recently – maybe a comment on AVC actually:”Fast, good, and cheap” – pick two.A lean startup IMHO should be “good and cheap” or “fast and cheap” – “fast and good” is only what you do once you know you have product-market fit, the direction and roadmap figured out.Being where I’m currently at, myself and my co-founder, and a freelance outsourced/non-local developer who I’ve been working with for a year now – we’re in the good and cheap mode and readying for fast and good for our next evolution.

    1. LE

      “Fast, good, and cheap” – pick two.There are variations of that in many businesses (I’ve mentioned that as “price, quality and speed, pick any two”)

      1. Matt A. Myers

        Quality and speed = higher price, if I’m understanding correctly – so it’s the inverse? “Speed” doesn’t equate to quality unless it’s a slower speed. Or fast a lower quality could work – or fast and higher quality but higher cost, so not cheap. 🙂

        1. LE

          Take a look at this:http://sinekpartners.typepa…Let’s take the hypothetical example of someone building houses. Someone can have the best quality and deliver on a good fast schedule but they will be expensive. Someone can offer the lowest price but then the quality or the speed of delivery will suffer.Let’s say you offer yoga lessons. You can charge a high price and give people all sorts of attention and longer than “x” hours. Or you can choose a low price but not give as much attention (maybe a group lesson?) Or you can have little knowledge (quality) so you can be both cheap and offer a long lesson. (Note the speed has been inverted). And so on. Make sense? In other words generally you can’t defy physics and given competition you can’t end up without great difficulty breaking these principles.

  22. Valentin Gutknecht

    Thanks for the great post! I think the crucial thing when forgetting about sunk cost is being aware of what mistakes and false assumptions caused them at first place. It’s actually pretty damn hard to be objective with this and not to repeat yourself!

  23. LE

    But we have made a few marginal or outright bad follow-ons. We do that for reputation value as much as anything else. We measure that value and understand that is what we are doing and we keep those reputation driven follow-ons small on purpose.Ok so we can call it the “de minimis” follow on?I could argue either side of this. But I will take this side:I could say that your reputation matters but it only really matters going forward. Not for money that you have invested for others where decisions have already been made. If someone gives you $1,000,000 to invest why should they care about your reputation? Only reason would be if they plan to invest with you again. If they don’t plan to invest again then that is money that is being spent and literally thrown out the window.This reminds me slightly of a time that an attorney billed me a fedex charge for sending a letter overseas to another lawyer. I said “why are we paying for this?” (implying that the other side should have). He said “well that’s what lawyers do” or something like that. While I could justify or rationalize that as somehow being a benefit for me I couldn’t get beyond the fact that he did it for his own reputation at my expense.What I found out was interesting. After pressing him further on this. He told me that a lawyer who couldn’t “control” his client was not thought of highly or something like that (further evidence that I was paying for something that wasn’t to my benefit but to his benefit).Look, I do “work” for people sometimes and put my name on the line. But I always make it clear that in the end the person that I am doing the “work” for is the one that ultimately decides what happens and I’m not going to hold someone to a decision that is not in their best interest because I need to look good to outsiders.

    1. Rick

      “But I always make it clear that in the end the person that I am doing the “work” for is the one that ultimately decides what happens and I’m not going to hold someone to a decision that is not in their best interest because I need to look good to outsiders.”.Are you making millions of dollars a year?

  24. curtissumpter

    This post sounds like a post about emotional discipline. Which is hard to learn from a post because a post is intellectual. You’re trying to impart emotional wisdom. In my experience that sort of thing has to be felt. You’re talking about lessons learned from a break up. When you’re talking about that you’re talking about pain. I think you’ve reached the upper limit of what a post can teach.

  25. Ari Lewis

    Great post. I’m learning about sunk costs in my Corporate Finance class. Theory vs Practical makes such a difference. We are building a product and we always try to make spending decisions on it’s own merit opposed to what we previously did. However, going against human emotion is so hard.

  26. Dave Rothschild

    Not letting go of sunk costs is a classic case of “loss aversion.” It is people’s tendency to strongly prefer avoiding losses to acquiring gains. Great insights that any innovator should understand about behavioral economics from Nobel award winner Daniel Kahneman.

    1. Matt A. Myers

      Potentially very harmful in any type of relationship.

      1. David Gobel

        Ground crew holding on too long to the anchor ropes of a rising hot air balloon is the image I use to remind myself of sunk costs.

  27. Vijay Venkatesh

    Thanks for this post – incredibly timely, and is one of the best things I’ve read. In the midst of building an MVP and getting a startup off the ground, so is a very useful way of looking at investments and time…

  28. ShanaC

    YOu can have emotional sunk costs too – #thingsIvebeenworking on. Don’t get trapped by them

    1. fredwilson


      1. Vineeth Kariappa

        foursquare ?

  29. Dan T

    ok, you almost had me convinced until you revealed your reasoning for not always following what you are suggesting is the right thing to do – reputation. I think this is problem, almost everyone will have a reasonable rationalization for throwing good money after bad, I just bought my first boat, so I am sure I am going to be learning more about this.

  30. Nathan Guo

    Sunk costs make a lot of sense to me as a former professional poker player, but what’s confusing to me is how they play into venture investing. How do you build these sunk costs into the model when looking at follow-on investing?For example: Let’s say you’ve already made a Series A investment in a firm, and you are deciding whether to invest in Series B. Is this the way to model that?Scenario A: Follow-on Investment, firm continues to growSeries A NPV: +$10M (incremental, change in value only not total stake)Series B NPV: +$5MNet NPV: +$15MScenario B: No Follow-on Investment, firm doesn’t find financing and fails, VC invests in some other opportunitySeries A NPV: $0M (equity investment worth 0 without follow-on investment so the equity is not actually decreasing)Invest in some other opportunity NPV: +$15MNet NPV: +$15MThe numbers are obviously contrived, but I’m trying to get at how to quantify the changing value of the Series A investment. Any clarification would be very helpful!

    1. cavepainting

      The idea here is that every dollar of incremental investment needs to be justified against all other alternatives and the best one be chosen irrespective of the performance of previous investment (s). Theoretically, it is the right thing to do but in the real world, it gets complicated for all the reasons mentioned in this thread.In your example, the real comparison is that of $15 M NPV in Scenario B to $5 M in Scenario A for the same cash outlay. Typically, follow-ons come from a different fund vs. the first seed investment so that is another factor that needs to be kept in mind.

  31. cavepainting

    I guess there are at least 3 scenarios where a seed investor contemplating a follow-on investment is considering the decision of moving on and not put “good money after bad”.1. Seed investment doing reasonably well. Company has hit milestones (product-market fit, early deals etc.). But the market dynamics is tricky and there are questions around whether this can really be a home run, or more likely to be a single or double, or worse, an acquhire.2. Good team and interesting market that can grow to be really large, but company is still searching for product-market fit and the right approach. With more capital, they may find the right direction. But they also may not.3. Seems like the company is solving the wrong problem, or has the wrong team. Lots of red flags.With 3, especially when these are fundamental issues with team / DNA / market, the investor’s decision to write-off is obvious.1 and 2 are more tricky. If investor really believes in the team, market and its potential growth and the X factor within, there can be a strong case to follow-on.As some one said recently, a dud is a dud until it becomes a star.Is the struggling investment a real dud or one that could transform to a star and eventually out-perform the alternative investment ?It really comes down to the quality of these judgements and decisions.

  32. Pranay Srinivasan

    Are the “sunk costs” a function of the people or the situation or the project? And what factors would make you “throw good money after bad” – the people, a changing market, a trend, an ability to scale the founder cant see (so bring in ‘professional’ management), departure of a key troublemaker that had created the situation in the first place or just a punt on the company?I know I’m optioning a lot of my own answers but I’m interested in knowing whats been the statistically most relevant reason you’ve followed on in a sunk cost scenario..Thanks,Pranay

  33. Gemmy Tsai

    When you say “ignore sunk costs,” are you allowed to factor in the “sunk cost” as a potential gain?Take an example:1) I’ve invested $10,000 in a company2) Company is about to go bankrupt3) I’m evaluating an additional $100 investment in the company. 4) If I make this investment, I will have a 10% chance of returning my $10,000 investment.Are we saying that 4) shouldn’t be considered?

  34. Mfj L

    Reading the discussion is informative and insightfulFrom my perspective I agree with the writer. and for the simple reason he is making 2 simple points. First you made the inital decision to provide seed money after reviewing a number of opportuities Yes it takes time and toil to fleshout a notion…… The smart seed investor knows that sometimes you have to prime the pump to bring the water up……. when that happens you can now invest in the bottling plant. So forget what was firts invested.. focus on moving things forward……….. Years ago I worked in what were “Skunk Work” Projects….. Keep in mind there is always more money chasing opportunity than there are opportunities…… To all you people out there…………. keep up the good work “Nose to the grindstone” develop your ideas………… keep the faith…………. and MBA Programs came long after the infrastructure was created……………. Plus “wall Street” has never picked a winner beforehand………….. It’s you guys who make WallStreet sucessful, without you they would choke and die