A Trick Question

Posted 2025-March

Recently, I was coaching some entrepreneurs and I asked a question that I later realized is a bit of a trick: no matter how it's answered, there's a very difficult question to follow. Let's explore.


A question that entrepreneurs infrequently ask themselves is , "Why didn't anyone else think of this?" Or, in a business context, "Why hasn't this been done before?" (IMPORTANT NOTE: "This" does not in any way refer to technical implementation! "This" refers to delivering a specific value to customers.) It's a simple question without an easy answer, because any answer leads to additional tough questions. Answering those, in turn, can lead to a lot of insight into the business. But beware the rabbit hole! Here's a quick outline:

Why hasn't this been done before?

  • This has been done before. Then why did the previous players fail?

    • They did not fail. So, competitors are out there. How are you differentiated? How will you take on their moats? How will you build your moat? If moats are shallow, can you really out-execute experienced players to win? How will they respond to your market entry?

    • They failed. What caused their failure? What is different now? How does that difference apply/accrue specifically to you, so that there are not a bunch of others also trying the same thing right now? How can you out-execute/out-compete all of these new players? HINT; "We're first to market" is not good enough.

  • This has not been done before.

    • Was the barrier technology?  What has changed? It may be that you have a true innovation, which is great if you can protect and monetize it. (The latter is often easier said than done, as you need a whole ecosystem of partners, all of whom get a cut.) But true innovations are rare - new technology applied to an old problem is often simply better/faster/cheaper and not different in the sense of value delivered.* Then the question is better/faster/cheaper enough to break through the previous failure.

    • Was the barrier the market? What has changed? It may be that the market was just not ready and/or had too much inertia to be disrupted. If it’s market readiness, why won’t there be many other entrants just like you? If it’s inertia, realize that current players in any market have a vested interest. Whether it's capital deployed, union contracts, the regulatory regime - the current market can fight very hard. (As an example, read up on containerization in the transport industry... and how long the longshoreman's union slowed its introduction.) Being the disruptor is great; being the disrupted is not, and people will do much to avoid it.

    • Was the barrier the business case? Technology can do a lot of things, but the economics work for only a subset of those. Can you deliver the value for less than the price customers will pay?† Hint: you really should be testing pricing early with customers, before you build a great tech solution no one will buy at the price you need to charge.

Conclusion

"Why has this not been done before?" It's a simple question. The wrong, but default answer is "because we are smarter than everyone who has ever tried". The right answer requires a much deeper understanding.

Notes

*  This question of if you've made a true innovation or are simply using new technology to try to be better/faster/cheaper is an excellent way to understand hype. Currently, everything is "AI" and "AI changes everything."  Nah. While some AI is itself innovative (some... but not LLMs, which can trace themselves to the ML algorithms that I studied 40 years ago!), the application of AI to most things is not. That's OK - really! It's better to be demonstrably better/faster/cheaper at something than amorphously improved... unless you're incented to use "AI" to get ahead.

And guess what? You can substitute "blockchain" for "AI" above and it still reads correctly. Or, for we software people, "object oriented". Few new technologies have been truly transformative. Only the Web comes to mind. (And even that had a Hype Cycle or two.)

† I can offer two case studies here, one real and one speculative. The former is "medical transcription and coding" - the process by which a doctor's oral report of your visit is translated to text, including the special billing codes needed by insurance. This is a very difficult and expensive task. (There are 200,000 certified medical coders - humans - in the US alone .) Massachusetts General Hospital created a whole company - CodaMetrix - to do this automatically in a computer. It is a technical marvel and has good accuracy. But when push came to shove, the hospital chose instead to send the recordings to India, where they are encoded by medically trained humans at a lower cost.

The other case is this one: seeing some of the latest Generative AI for video, a friend asked "how long before this is used on OnlyFans?" My answer was simple: creating AI video at low latency may be possible in a year or two. However, the cost - in processing power - to do so will far exceed what you have to pay a human performer, perhaps indefinitely.  Just because something can be done with technology does not mean it makes economic sense to do so.