The Bubble of Deep Tech

The Pivot to Deep Tech is Due to Bubble

It Starts with Bacon

In the 17^th century, Sir Francis Bacon discovered the scientific method. Its magic is quite simple: hypothesis, test, analyze.

The problem with the scientific method is it’s really expensive. Ideas are free. Research can be done on the cheap.

But testing an idea requires a lot of capital. 

Let’s say you wanted to build a small restaurant in your town. Maybe you believe your town needs a sushi restaurant because your community doesn’t have one within 50 miles.

At a minimum, you’ll need to buy items, build out a kitchen, and find customers.

You also need to make great quality food. Once those customers come, they need to return. There’s also the cost of regulations, real estate, and hiring staff – after all, preparing sushi is harder than burgers!

This is all the cost of testing a hypothesis, a rather daunting task just to see if you’re correct. Since most of this would take months to develop, your cost could be many tens of thousands of dollars. In many instances, it’s +$100k.

A Better Way to Test:

If the true goal of a small business is to make money, we should opt for easier money to made than harder.

That varies by markets, gaps, profit margins and scale.

Let’s start with two factors to consider:

1. Is it physical or digital?

2. Is it B2C or B2B?

With this in mind, winners tend to get to market faster. They do that by reducing the cost to build their MVP (version 1). In the old days of tech, it would cost millions of dollars to do this.

This made it a really hard market to compete in.

Today, the opposite is true. Most of the gate keeping is on physical work – whether that’s restaurants, construction, or a nail salon.

Alternatively, a great coder can spin up an app in a few weeks or months. Of course, they need to know how to code and understand the full stack (front end, backend, and database).

Until recently! With the advent of no-code companies, it’s almost too easy to build a tech company. What used to take months and cost $50-100k, now takes a week or two.

Moreover, you don’t need to learn to code. Yes, the thing that people spent 4 years studying – those people are largely being replaced. Software engineering roles are down 50%!!

Bubble, the most prominent no-code company, is less than a hundred dollars each month.

The learning curve is low and you can build a minimum viable product for less than $500.

That’s a 2-3 order of magnitude difference from beforehand (and you need zero knowledge of coding). The ability to build is in the hands of the 98% of people who cannot code!

What’s the tradeoff?

Well, cost.

If you run an app, you need inputs and outputs. Using real code, an input has a linear and simple line to get you an output.

With Bubble, it’s very staggered. Each input needs a condensed and exact output before getting you the final output. It’s a non-linear path to execute a command, which ends up being more costly in the long run.

The shortest path is always going to be cheaper. But code is incredibly cheap only once the upfront creation exists.

The cost to service your first customer is maybe $100k. Customers 2-100 are nearly nothing more. The risk is ensuring customer 1 is:

a) Found

b) Interested

c) Purchases

d) And can be reproduced to many more

With Bubble, servicing customer 1 might be $1,000. In fact, we are doing this now with two of our clients. To actually test this out,

Ongoing COGS may be 5-10x more than with real code. But still, it will take hundreds of customers before that breakeven matters.

The Pivot to DeepTech Venture Capital is Inevitable:

Founders don’t particularly want an overbearing venture capitalist, breathing down their neck. The coexistence occurs mostly because founders have no money - VCs enter to help you test a hypothesis and build that MVP.

All for a small percentage of the company. 100% of nothing is still nothing!

If the product finds a market gap, wonderful! You are looking at 80% margins since the marginal cost to service a new customer is ~$0.

But with such a low cost to test an MVP, why would you need venture money?

The response would be, ‘we provide expertise, guidance, and advice to actually add far more than just money.”

But is that really true?

Famed VC Vinod Khosla spoke to Sam Altman recently and disagrees:

This is not some schmuck, he’s one of the most successful VCs of all time. He has a net worth of $7.4B by investing in small businesses like Instacart and DoorDash. His words are gospel.

While I’ll write more about the pivot to DeepTech (and what it is), the obvious repercussions of a test that drops by 90% is founders don’t need to raise $500k.

VCs are smart and ahead of the curve, they see the writing on the wall. They have two options:

1. write bigger checks to more established companies

2. write similarly sized checks on businesses that still require an ‘expensive hypothesis’ to test out

In part, they have been funneling into investments that satisfy #2.

This is mostly really expensive equipment and physical products. Think rocket ships, airplanes, wastewater treatment plants, robotics factories, and nuclear facilities.

Take a look at the largest venture capital firm, Andressen Horowitz. They are heavily positioned in their new “American Dynamism” fund which is a fancy way of saying physical tech.

More on deep tech and the quick pivots, coming soon!

Stay Nimble.