Building a startup used to involve Steve Jobs level intuition and magic. Customer Development and Lean Startup heroes Steve Blank and Eric Ries turned building startups into a scalable repeatable process, through talking to customers and applying the scientific method. Your startup is a science experiment. Learn how to reach product-market fit before your resources run out.
There are no facts inside the building, so get the hell outside. – Steve Blank
Your startup can be treated as a science experiment: you form hypotheses about what customers want, test them as cheaply and quickly as possible, and use what you learn to decide whether to keep going, change course, or stop. The aim throughout is to reach product-market fit before resources run out.
At the heart of this approach is the scientific method applied to business. Blank’s central lesson is that there are no facts inside the building, so founders must get out and talk to real customers rather than building on untested assumptions – a common and often fatal mistake. Ries extends this with the build-measure-learn cycle and the minimum viable product: rather than spending heavily up front, you build the smallest thing that lets you learn what customers actually want and will pay for. Demand can even be tested before anything is built, through a “smoke test” such as running ads for a product that doesn’t yet exist, selling it before you build it. Building something nobody wants is one of the leading causes of startup death, so cheap, early validation matters enormously.
Adaptation and decision-making follow from this. No business plan survives first contact with the customer, so founders need to know when to persevere, when to pivot, and when to piss off, and in the early days, runway is better measured in how many pivots you have left than in cash alone. Good decisions depend on good data, which is why it pays to keep a lab notebook of experiments and results, and to be wary of vanity metrics, the numbers that flatter you without moving the business forward.
Homework: If you don’t yet have an MVP, work out the smallest thing you could put in front of real people right away, even just a smoke test, and ask what’s stopping you launching within days. If you already have a product, identify the next thing you need to learn and design an experiment to learn it.
Transcript:
Kia ora koutou, hello and welcome to Episode 9 of nzangels.com – a guide to raising angel investment in Aotearoa New Zealand. I’m Dave Moskovitz, one of New Zealand’s most experienced angel investors.
This is Episode 9 – The Science of Building Startups
Back in the olden days, before the turn of the millennium, to be a great entrepreneur you needed to have a sixth or even seventh sense about how to hustle a business into existence based on your own beliefs of what the market wanted. It was a bit like alchemy, or magic, at the same time as being potentially high cost and high risk. Steve Jobs was such a magician, but there aren’t many Steve Jobses around.
Then, in the 2000s decade, Steve Blank and Eric Ries burst onto the scene with their Customer Development and Lean Startup Methodologies, which were codified in their respective books, the The Four Steps to the Epiphany, and The Lean Startup. Blank and Ries took a dark art, and turned it into a scalable repeatable process, through the application of the scientific method.
You may remember the scientific method from high school: formulate a hypothesis, construct an experiment to validate or invalidate the hypothesis, analyse the data from the experiment, and come to a conclusion about whether that hypothesis was true. Rinse and repeat, changing your hypothesis and/or the experiment based on what you learned, until you have a solid result.
Your startup is a science experiment. You need to run experiments to validate your hypotheses about what your customers will pay for.
Blank’s central idea is encapsulated in his most famous quote: “There are no facts inside the building, so get the hell outside.” Here we are nearly 20 years later, and I still meet plenty of wannabe entrepreneurs who have built a thing without much reference to actual customers. That only works if you have Steve Jobs-level magical superpowers. It’s much better to get out of the building and find out what your desired customers and market really need and are willing to pay for. You can’t learn that with any accuracy without talking to a significant number of them. If you don’t validate your hypotheses by talking to customers, you run a significant risk that you’re wasting time and money building something that the market doesn’t want.
Eric Ries took these ideas much further, building a methodology around how to test your hypotheses by building a minimum viable product (MVP), and using it to learn as much as you can as quickly and cheaply as you can, without wasting resources. He says, “The goal of a startup is to figure out the right thing to build – the thing customers want and will pay for – as quickly as possible.” Before you get to product-market fit, any resources you spend on anything else are potentially wasted. Validated learning happens through a build-measure-learn cycle, echoing the scientific method.
You don’t even need an MVP to test some things. You can do a “smoke test” before you build anything to determine if there’s any interest at all in what you’re proposing to do. Spend a small amount of money, and run some google or facebook ads advertising the product that you haven’t built. Getting any clicks? No? Maybe think about framing your ad, or your whole concept differently. This tactic has been called “sell it before you build it”. It would be a real pity to spend a lot of time and money building something, and then finding out that nobody wants what you’ve built – this is a leading cause of death of startups.
In the previous episode of this series, episode 8, we talked about the Lean Canvas and how you can use this as a tool to discover the assumptions – or hypotheses – inherent within your business model that you need to validate. In the early stages, this is a great place to start applying the scientific method to building your business.
And you need to adapt, iterate, and pivot as you validate and learn. Paraphrasing von Moltke the Elder, Steve Blank says that “no business plan survives first contact with the customer”.
Ries says that you need to know when to pivot, when to persevere with your existing strategy. I’ll add, you also need to know when to piss off and do something else entirely. Pivot, persevere, or piss off. In the early stages, the real measure of your runway isn’t how much cash you have left, it’s how many pivots you have left before you either get to product-market fit, or have to shut down.
Like any good scientist, keep a lab notebook – this could be a spreadsheet – of the experiments you run and the results. You can then go back over the data and learn from the patterns that you observe.
Making good decisions, especially business-changing decisions like whether to pivot, persevere, or piss off, depends on having good data on which to base that decision. We’ll have a detailed discussion of metrics in our next episode. Spoiler alert: it’s tempting to choose vanity metrics – things that are easy to measure or make you look good – but these are not the numbers that are going to move your business forward.
Your homework for this session is, if you don’t have an MVP yet, to figure out what the smallest thing is that you can put into market to start testing your assumptions with real people. If you have an idea for an MVP that you’ve been delaying building, can you strip that back substantially? It might even be as tiny as a smoke test. In all seriousness, what’s stopping you from having this in market in the next few days?
If you do have a product in market, figure out, what is the next thing you need to learn? What experiments can you run to learn that thing from actual interactions with real people? Can you instrument the capability to run experiments directly into your product?
Let’s finish up with Eric Ries: “Startup success is not a consequence of good genes or being in the right place at the right time. Startup success can be engineered by following the right process, which means it can be learned…”
Just before we go, we have some bonus content for you on the this nzangels.com episode page. It’s an eight minute crash course on problem solving for startups I made in 2021, which reinforces many of the concepts in this episode. This will give you some tools for punching through those difficult issues on your way to startup success, including divide and conquer, the five whys, rubber ducking, and more.
I look forward to seeing you in our next episode about metrics.
Until then, ka kite!
Bonus segment: A Crash Course in Problem Solving for Startups
Leave a Reply