Hacker News Reader

Comment by ttul | Hacker News Reader

What rings true for me in this article is the notion that, as a founder, it’s better to pick a problem space than it is to pick a solution. Amazon chose to live in the problem space of how to bring consumers more products, more conveniently, at a lower price. Geico chose a similar goal but in insurance.

Technical founders often focus on a particular cool way of solving a problem and burn lots of capital building that thing. Sometimes the thing is the right thing and everyone makes money. But sometimes it’s not. Yet if you stick with the same problem space for long enough, and aren’t too connected to your particular solution, I think you have a greater chance of succeeding.

Okay, go ahead and poke at my pontifications now. I’m ready.

View on HN

I think maybe the fundamental hard thing about startups is that you can’t ignore the problem space or the solution space. It’s not like Bezos was passionate about selling cheap books for years before Amazon, he saw that this new Internet thing was going mainstream and building on top of this new tech was the premise of the business more than selling books cheaply and quickly was.

Most of the startup opportunities in the last couple decades were built off the back of several deeper tech breakthroughs like mass mobile phone adoption and virtualization. I think more founders have been successful just riding these tech waves than sticking to a problem space. Brian Chesky didn’t work in hotel management and Travis Kalanack didn’t work in transportation. They were just familiar with cutting edge consumer tech and applied it to problems they had no familiar with before they started.

Note that choice of starter product is important. Amazon started with books - you know exactly what you're getting, they always fit, they're hard to damage, they have pretty high markups and low shipping costs. People were willing to risk buying books online, back when online e-commerce basically consisted of the USENET misc.forsale groups.

Tesla and many other EV companies started with low-volume expensive sports cars, with high margins so they could be basically hand-crafted. (Tesla actually bought the chassis/body/interior from Lotus) If they had started with the Model 3, they'd have gone bankrupt long before they sold any.

I think Bezos also mentioned that he started with books because no single physical shop could have all the books, but Amazon could. It is a product with so many variations that Amazon had a clear competitive advantage.

Amazon's competitive advantage was not paying any tax. Amazon's selection in technical books was quite poor for quite a while.

However, the tax discount drove ALL the small, local technical booksellers out of business. It might have happened anyway since local sellers had to pay retail rent and so Amazon had a structural advantage. However, Amazon also gained a straight 8% advantage for not having to pay tax.

> Amazon's competitive advantage was not paying any tax.

I'm assuming you are saying "not paying sales tax". Amazon was operating like every other mail order bookseller that did not operate stores at the time. Retailers have complained for a century about mail order catalogs not collecting local sales tax. This was not new, or novel.

> the tax discount drove ALL the small, local technical booksellers out of business

Demand aggregation is a real thing. Localized, there's not a lot of demand for any specific niche technical book, and that leads to sub-optimal inventory. Why carry something you'll only turn once a year? You'd be better off putting that floor-plan money into gum, mints or batteries at the cash register. Quick access to product is why I shifted to buying technical documentation online. The local electronics store or bookshop could order books and it would often take 4-6 weeks for the book to arrive. Even early Amazon did 3-4 days. Demand aggregation is how Amazon won.

I remember at the time there was a lot of talk about how Amazon didn't have the costs of high street stores etc and was 'just' a web site.

That was never true - sure they didn't have high street stores, but they invested big time in warehousing. That's why you could get a book from Amazon in days, whereas a traditional seller it would take weeks if it needed to be ordered.

They also invested in the wider distribution network.

Another now-classic example of solving this chicken and egg problem, I think: Facebook starting with just Harvard, then ivy league colleges, then colleges, then colleges and high schools, and only finally the general public. At each stage the arena was small enough it was realistic to take the whole segment over, and then that became the seed for the next phase.

And their second product was CDs/DVDs, which have a lot of the same properties. Then they opened Amazon Marketplace, allowing other stores to sell through their site, and used the data of what was popular there to expand into other products – they were consistently very deliberate about what to sell next.

Bezos built Amazon to solve the traveling salesman problem.

But that spawned the traveling delivery truck problem.

That was solved by making employees pee in bottles. Need a restroom? Think outside the box, pee inside the bottle.

Then to be shadowed by the traveling delivery drone problem.

I think this also points out the timing aspect. The big "breakthrough" companies (breakthrough in terms of capitalistic successes) happened since people were in the proverbial right place and time with the right idea.

It's a falacy. Most of the time, what defines the "right time" is our post event bias. Apple launched the iPhone in 2007(?), so from our 2022 perspective, "they were just the first to do it but in 2007 everything that was needed to build an iPhone was already there". The fact that Apple succeded in creating an easy to use, touch screen based, infinitely expandable via software general purpose computing device is 100% their merit. Many companies had tried before, but noone were able to get it right. After you see it done, this concepr of " right time" arises, not before it. It's the same with the age of discoveries. Boats capable of crossing thr ocean had existed for centuries, but only around 1500 it became "the right time".

Well, except what you’re describing fundamentally leaves the hardest problem unsolved.

Where to start (concretely) and what you’re actually going to do, and if it will work for customers in a way they will pay you more for it than it costs to do.

Discovering that, and if it is going to work, is what a startup does.

You can’t do it without deciding on something concrete eventually to deliver, and pursuing it with enough vigor to get a workable product in someone’s hands.

Tech startups that build an entire product without understanding the space are leaning too heavily in one direction.

Sales led startups that get a customer base and start collecting money without a concrete product plan are leaning too hard in the other (and even more, add big red flags for scams and fraud).

Doing all the things and integrating it into a workable collection of customer needs they’ll pay to solve, along with an actual solution that works to solve it, while not going bankrupt in the process is why startups are hard.

Not being a scammer or committing fraud in the process? Harder still!

It’s the unknown unknowns and navigating all the bullshit on the way that also make the risk very high. Fundamentally, it’s impossible to know everything until you’ve already done it, and by then, it’s already changed under you.

Totally agree. For a negative example see all the companies desperately trying to find sone problem for blockchain to solve.

I generally agree with your statement, but it depends on what your goal is. Ultimately, do you want to build a successful business or bring a (specific) new technology into the world? I suspect the former is true more often and it can lead to innovations along the way, but there is nothing wrong with the latter.

I was once at a talk with several medical technology founders. One of them namechecked specific legislation that enabled their solution to come to market. And so I asked, "How do you survive if only 1 law is prying open the opportunity for your business?"

They replied that their business exists to solve a need -- a need that exists regardless of the specific regulations. Should the regulations shift, a company motivated beyond a specific solution will adapt to keep meeting its customers' needs.

Depends on perspective. What is one thing and what is a problem space.

Is a microprocessor or a mobile phone one thing in many different problem spaces?

Businesses are not built on things they sell, they are built on the customer problems they help solve.

If you build microprocessors, you better understand how your clients are using them to create electronic devices, and provide all the services and documentation they may need. A less potent microprocessor paired with a website that perfectly explains how to use it and what use cases it supports will win any day over a more powerful chip that no one understands.

Yes, but the investment in things can be so great that it becomes impossible to change according to customer preferences. The business then has to find new problem spaces for their things.

Sometimes quoted as "Fall in love with a problem, not a solution."

> Technical founders often focus on a particular cool way of solving a problem and burn lots of capital building that thing.

This nicely summarizes the current hype in Bitcoin or deep learning.

Isn't Amazon's cash cow AWS which is not in that problem space?

AWS is Amazon's most successful business, but they were wildly successful even without it. AWS only launched in 2006, 12 years into the company.

Today, most estimates I could find put AWS at 25%-50% of Amazon's worth.

So 25-50% of the company is not in the problem space, at least. Doesn't sound good for the stay in the problem space theory.

this is a great strategy if you already have near infinite resources

Great observations!

Reply on news.ycombinator.com