Why Cutting Losses Early Is the Hardest—and Most Crucial—Skill in Startups and Venture Capital

Cutting losses is one of the hardest decisions in startups, investing, and leadership—but it’s also what separates winners from those stuck in the sunk cost trap. Here’s why mastering this mindset is essential.

Recently read this Forbes article on Igor Tulchinsky, a Billionaire quant trader who runs the hedge fund WorldQuant. In particular, this section on cutting losses caught my eye:

Source: This Billionaire Quant Is Turbocharging His Trading Models With ChatGPT-Style AI

While I don’t come from the public markets world, I have taken a series of major risks as a founder, operator, and investor. Of course, now that I am a full-time venture investor, I live in a world where I take and manage risk every day, including macro, business, tech, portfolio construction, and people, among others.

Based on my journey so far, I can’t emphasize enough the importance of developing the ability to quickly cut losses. Interestingly, before making a major decision, most people are fairly good at identifying & mitigating key underlying risks. However, I have learnt with experience that even after executing the best risk management process, things will still go wrong. And once things go wrong, even the most intelligent organizations & individuals easily fall prey to the sunk cost fallacy (“throwing good money after bad money”).

Let’s take the classic example of finding your next job. As part of a thoughtful risk management process, an intelligent candidate consciously tries and figures out mutual fit during interviews, gathers feedback on the company’s culture, perhaps speaks to customers & competitors to evaluate the product, or, in the case of startups, even does a 1-2 week part-time project before commiting full-time.

A similar scenario is also playing out on the employer’s side. Most hiring managers give high weightage to candidates who come recommended from trusted connections or with whom they share a past history. The interview process consists of multiple rounds to stress-test skills & personality. The company does rigorous reference checks, often also focusing on off-sheet checks to eliminate bias.

So both employers and candidates follow a fairly rigorous risk management process. Yet, as most of us have seen in the real world, leadership hiring has a 50 %+ failure rate in Corporate America. Here are some summarized stats from ChatGPT on this:

In this case, even the most rigorous upfront risk management process can’t account for a variety of post-decision risks, including process weaknesses (a great hiring process can be undone by a weak onboarding & training process), uncontrollable externalities, and random one-off events.

In these scenarios, a willingness to quickly cut losses & limit further damage of time & money on both sides is the best way forward. And make no mistake, it requires a lot of courage. That’s why I found Starbucks firing their last CEO in less than 18 months of tenure to be a very bold move, especially for a company of that scale & history (you would expect them to be sluggish).

While exec hiring missteps can be major setbacks even for large companies, they can often become matters of life and death for an early-stage startup. A wrong hire for a critical role can do strategic & cultural damage that might be irreversible with the existing runway. That’s why the best founders believe in the “fire-fast” philosophy.

Zooming out from hiring, startups succeed by taking calibrated risks on top of a technology change that an incumbent would just find extremely hard to do. This requires running a bunch of iterative experiments with very limited upfront data, but balanced by an asymmetric risk-reward profile (if this works, it will massively move the needle).

By the very nature of these experiments, a majority of them will fail. Combine this with a very limited cash runway that even the best startups get at each stage to get to the next set of milestones, and founders need to combine controlling the cost of each such experiment with an active intent to cut losses once it’s clear that the experiment is not working.

Essentially, a mindset to cut losses early till you get to something that is clearly working is a key requirement for startups to successfully emerge from this maze of early experiments with real product-market-fit. Windsurf CEO & Co-Founder Varun Mohan framed this idea brilliantly in his recent interview with 20VC:

Never fall in love with your idea…

One of the weird thing about startups is that you don’t win an award for doing the same wrong thing for longer.

Coming to my world of venture capital, I have seen many instances where the aversion to cut losses has come back to bite the investor. The context I have seen this the most over the years is in ill-conceived bridge rounds.

Classic scenario – the company has exhausted most of its last round of capital, has created just enough progress to keep existing investors somewhat interested, but if looked at with rigor and intellectual honesty, is nowhere near product-market-fit. Combine this with a founder who is good at storytelling and can pitch “if we get just this much more money, we will break through”, and existing investors are highly likely to cave in & bridge the company.

Unfortunately, in my experience, a majority of these types of bridge rounds don’t end up working. Peter Thiel said this uncomfortable truth a few years back about what he has observed in the Founders Fund portfolio over the decades (paraphrasing):

Once something starts working, people often underestimate it. And when things aren’t working, people often underestimate how much trouble they are in

Everytime a company raised an up round done by a smart investor, it was almost always a good idea to participate…

Steeper the upround, the cheaper it was…

In flatrounds and downrounds, it was almost always a bad idea to participate…

This behavioral weakness is perhaps why Michael Kim of Cendana, a major LP in emerging managers, recently said in an interview that the biggest mistake he has seen GPs make is deploying reserves poorly. My logic is that reserves deployment, especially in rounds without quality external signaling or real business progress, is particularly prone to multiple human biases kicking in, including loss-aversion, likability bias, optimism bias, and overconfidence bias.

Funds with relatively large reserve ratios should think deeply about potential solutions to this problem. One thing I have seen a few funds do is have a dedicated GP whose sole job is to evaluate each reserves-deployment situation like a fresh late-stage deal from the ground up. This can help counter the personal biases of the lead GP on the original deal.

To summarize, the ability to avoid the sunk cost fallacy & cut losses early is critical not just for entrepreneurs & professional investors, but also for each of you as every contact with the real world exposes you to risks big and small, whether you realize it or not. Getting out of sticky situations early enough ensures that you stay in the game and keep compounding your advantages.

Top 15 insights on how to operate as a startup leader

Recently came across a great conversation between Keith Rabois and AngelList, back from Aug’18. So many tactical insights for operators, founders, big co./ startup teams, or anyone who is interested in understanding how leaders should operate on-the-ground. My key takeaways below:

  1. Talent can be classified into “Barrels” (can independently execute end-to-end, from idea to product-in-market) and “Ammunition” (require supervision, execute only specific elements well). The number of Barrels in your team governs how many parallel things you can do.
  2. Every business can be ultimately distilled into an “equation”, with key revenue & cost variables that ultimately drive profit. Founders need to understand their business’s equation really well, which is what drives strategic insights that lead to better decisions.
  3. A key job of a founder or CXO is to compress “time” for the business, via a communication strategy of “simplify” and “clarify”.
  4. In the majority of cases, larger engineering teams tend to slow execution down. Paraphrasing a quote by Eric Schmidt — “one of the most powerful things is 2 engineers working together”.
  5. Put your best people on the most challenging problems, irrespective of what it does to your org. chart.
  6. The more transparency around data and information that the CEO can create, the better everyone else can make day-to-day operating decisions that align with the company goals and strategy.
  7. There is a saying in sports that a particular team has been “coached to play fast”. This is what startup leaders need to do to increase the speed of execution — coach their teams in a way that they can take fast decisions & react instantly, and in high fidelity to company goals.
  8. As a leader, it’s important to speak in “Whys?”, and not “What we are doing?”.
  9. As a leader, it’s important to change your management style as per the kind of individuals or teams you are working with at a particular point in time.
  10. The CEO is the “Chief Editor” of the company. You aren’t actually doing a lot of the functional work yourself but your key job is to a) simplify things for others, 2) create consistency across teams, and 3) create a coherent narrative & voice, internally & externally.
  11. As a founder, it’s important to understand the difference between a “bad” team and an “incomplete” team. Both require very different strategies.
  12. Best way to onboard talent (from intern to exec) -> start with as narrow a scope as possible, let them succeed at it, and then keep expanding their scope & pushing their range.
  13. Hiring is a muscle — you get stronger as you do more of it.
  14. An important question to answer while hiring: are you hiring for upside creation (is there a spark?) or downside protection (rigorous value creation role)?
  15. A simple best practice to improve hiring is to borrow your network to vet candidates and do comprehensive reference checks.

I already started implementing a bunch of these at my startup Workomo. Would love to know if you have used some of these tenets in the past, and your experience/ key learnings from it.