Ends do not dictate means
OK, so now you ruthlessly compared and validated alternative missions (together with your team) and found a worthy problem to solve and a solution concept that shows interest/traction. You have a rough "what" and a really good "why".
Now it’s equally important to avoid getting fixated on a certain way of achieving the goal, of solving the problem. The "how". And actually a big part of the "what". The confirmation bias monster is always lurking just behind your shoulder.
In finance 101, we learn that investments should aim at the best expected value for an investment at the lowest risk.
What if I told you that, in a startup, the most valuable return for any investment is NOT money?
But isn’t a company even legally required to maximize profits, or more broadly “shareholder value”?
Yes – but in the long run. And a startup does not maximize long-run profits by maximizing the financial profitability of each investment.
Data is black gold, understanding is rocket fuel
Execution, especially in the beginning, is mostly experimentation: The primary aim of product development, production, and marketing should be to give you as much understanding and data that brings you closer to finding that business model. Striking gold on the first shovel swing is unlikely.
One is tempted to bring up the Edison cliché: “"I have not failed. I've just found 10000 ways that won't work.” Yes, he did fail, but he did not waste his time. “Failures” in terms of the direct end goal strived for can be very valuable if they are properly documented and allow better chances of achieving the main goal in the future! (I'm tempted to write a separate piece on 3 distinct type of "failure" and the value of each and distinguishing between them. No promises!)
As a general takeaway point for startups entrepreneurs: Prioritize the experiments/investments of which you learn the most, most quickly, and with the smallest investment. You should especially prioritize investments that help you to decide what to do next.
This was one of the best tips we got from Martin, one of the first angel investors in my previous startup.
So, sometimes it’s actually worth postponing or completely skipping investments that would be profitable in terms of financial return – especially if those investments are not scalable and do not help you towards finding a sustainable business and growth model.
Don’t follow the money, blindly
For example, in a previous company of mine, we had found a working and cost-efficient process for launching the service in small Finnish towns. So, if we wanted to maximize the certainty of short-term revenue growth and profitability, it would make sense to launch the service in all Finnish small towns as our next actions, postponing everything else.
However, the market of small Finnish towns is very limited. There’s no guarantee that the same process will work in small towns in other countries. (In fact it was quite likely that it wouldn’t, since we were quite aware that the process was dependent on certain “home field advantages”.) Also, with no potential competitive threat, there was no hurry to launch in those small towns. What is more urgent in terms of maximizing the company’s long-term growth potential as well as proving the case for future investment rounds is making riskier though not necessarily bigger) investments in trying to find a scalable expansion model for larger markets (bigger cities, bigger countries, new types of target industries for the marketplace we were developing etc.).
Also in science, failures are discoveries
Bringing back the analogy to academic research, if you wanted your experiments to “succeed”, as in yield the expected results, you would be repeating the same, obvious experiments until eternity. And that would not be very productive science.* As Enrico Fermi said: “If the result confirms the hypothesis, then you've made a measurement. If the result is contrary to the hypothesis, then you've made a discovery.”
Science does not progress by proving things right, but by proving them wrong, falsifying previous hypothesis. The same holds true for startups, though it might seem painful: You need to constantly try to falsify your assumptions and business model hypothesis.
Even though we shouldn’t aim to maximize direct returns, it does make sense to continue the iterative search process by developing and testing further something that shows signs of working – repeatable and scalably.
Martin’s other good hint was: “try many things with small investments, and when you find something that works, double down on that”.
Then there are of course experiments that can only work at a bigger scale or with a bigger initial investment. Marketplaces, social networks, and other services with “network effects” often have this kind of critical mass dynamic. But more on that in another blog!
For now, go out there and fail fast!
*PS: ..but DON’T degenerate into a scientist!
One big difference between startup entrepreneurship and scientific research is that repeating the same experiment multiple times is much more desirable and valuable in science. Many researchers agree that repetitions of previous studies are not done often enough nowadays, as academics have realized that journals and other media favor “novelty” in research papers.
In startups its much more OK to assume, draw fast conclusions, and take other shortcuts, especially in the beginning when one needs to also be constantly questioning the profitability and worthwhileness of the whole “research field” (the industry or customer need), which academia does not do a lot of.
In startups the end goal is real added-value (for customers and employees) and profits (for investors). Theories and understanding of the customers’ behavior and the functioning of the market are just means to this end.
In science, more solid theories are the end goal. Experiments and measurement equipment are means and tools.
E.g. the multi-billion-euro particle colliders in CERN are not great achievements of science, but merely tools by which some scientists believe we can notably improve and clarify our understanding of the universe. E.g. Tesla's broad charging network is not primarily a research device – though they likely collect a lot of data.
The final goal in Facebook aiming to acquire and commit most of the Earth’s population into its users is not to help it develop and prove new theories of human online social behavior (which it undoubtedly also does lots of), but to produce more added value for its paying advertiser clients and to be able to capture more of that value.
PPS: The easiest way to follow this blog is probably on Facebook ;) http://www.facebook.com/paradigmslayer