A quantum of product-market fit

I used to be a dumpling mogul. Ok, maybe mogul is a bit bold. I did however hold court in Chinatown NYC and take groups of startup folks around for dumplings and conversation.

What started as a way to build my network after my startup eventually turned into a full-fledged food tour. I had a website tied to my ticketing system, social media followers, corporate sign-ups, and a regular schedule of tours taking visitors to taste the best of Chinatown.

Then few weeks before Christmas, sign-ups exploded. As my phone was buzzing from new orders, I reached out to a few guests to ask them how they found the tour. The all said the same thing, from the most recent New Yorker magazine in their guide to unique holiday gifts!

To be clear, I had no intention of making this a business. It was simply a fun activity, one that brought me joy and give everyone that joined the tour an appreciation of Chinese cuisine and full bellies.

I went into panic mode. I immediately took the website down, sent messages that the tours were cancelled, and returned funds back to everyone. While there were many disappointed folks, I also spared myself the misery of doing outdoor food tours in the dead of NYC winter.

While I never treated the food tours as a serious enterprise, it actually saw impressive growth. In the two years from first tour to the magazine piece, I collected tens of thousands of dollars in revenue from hundreds of guests, all as a one-person operation running regular tours.

The growth story of my actual startup however was anything but up and to the right. It took over a year to collect our first revenue. We had a strong pipeline, but deals were slow to close. Most months we had no revenue. This went on for three years until we finally agreed to sell the company.

This was certainly a better outcome than the majority of startups. We also had the deck stacked against us, selling complex HR tech to large enterprises during the 2008 financial crisis and without the benefit of VC dollars in a market when many VC’s were closing shop.

At no point along that journey did we feel we were hitting the proverbial inflection point. We would have one win only to run into major roadblock. We were running as fast as we could, but never felt that we were gaining momentum to move forward.

There are a ton of articles in the startup world that talk about product-market fit. It’s for good reason, because it is the one number thing every startup needs to achieve in order to have a fighting chance at success. Marc Andreesen, Founder of A16Z, had this to say about the topic:

If one of the world’s top VC investors is saying this, then it matters. You can speak to any successful startup founder however and they will tell you that their success only came after the hard work of finding that elusive product-market fit.

The process of achieving product-market fit will be a slog, much like it was with my startup. Mostly it will feel like lots of false starts, failed attempts, and meandering dead-ends. Nearly every day as a startup founder, I woke up questioning if everything I had done before was a mistake.

Contrast this with my burgeoning dumpling tour empire. It just seemed to grow without my doing much of anything. Then when the media latched on, it completely changed the trajectory of the tours. I think Michael Seibel of Y Combinator said it best about what product-market fit feel like:

So what is that process from going from zero to rocketship? The first thing to realize is that product-market fit not a binary event. It is a journey where you achieve a quantum of product-market fit. By this, we mean you inch ever closer to product-market fit in stages based on unlocking a series of key insights about your market and customer need.

You can think of this journey as five steps:

1. Have a hypothesis of the customer / market need

2. Build a useable product / mechanism to test your hypothesis

3. Establish metrics that demonstrate evidence of need

4. Gather quantitative and qualitative feedback from customers

5. Iterate or pivot to new hypothesis based on feedback

This is a constant loop of getting directionally correct to meet the market. Sometimes you may end up scrapping past work. Sometimes you get some positive responses and understanding something new about the customer or market that allows you to create a more engaging experience. Over time, you will get greater quantum of product-market fit where traction starts to accelerate.

Often would-be entrepreneurs get stuck on the first step, developing the hypothesis of customer need. We introduced in a previous post the four questions to ask when choosing an idea to pursue. You can use this framework to help land on a worthwhile idea, but we also recommend you speak to potential customers in order to get initial validation that you are on the right track.

A common misconception is the idea that the process of product-market fit requires building a fully baked product. We work at AWS, so we interact with many technical founders that want to jump into immediately into code. There are situations where that is necessary, but there are also successful startups did not start with any product like Airbnb and Groupon which were just static websites, and Dropbox that started as nothing more than a video showing how Dropbox would work.

As you progress on the journey, beware of hitting a product-market plateau. These are false signals of market adoption that keep you busy enough to believe you are on the right track, but not the type of crazy and chaotic lift of true market need as Andy Rachleff, CEO of Wealthfront, points out:

Should you survive through all the iterations, mistakes, false signals, and pivots, you will eventually pass enough stages of product-market fit to see the hypergrowth acceleration that feels so drastic that you can barely keep up. Just remember that you can also lose product-market fit if you stop innovating, so always keep the steps above as a core principle of your culture.

What was the journey for your startup in finding product-market fit? Where have you experienced the most roadblocks along the way?

Mark & Basil

Is consumer tech dead? Garry Tan, CEO of Y Combinator, was asked this question recently in a fireside chat and his response was absolutely on point:

He then introduced the concept of the “idea maze”. The concept originated with Balaji S. Srinivasan, former chief technology officer of Coinbase, to help visualize potential product strategy decisions and outcomes.

The maze is relevant at any stage of product development. From the standpoint of startups, you can use this before building your MVP to help you get started. It is also useful to revisit periodically as you grow to anticipate, as Balaji notes, “which turns lead to treasure and which lead to certain death”.

Chris Dixon, General Partner at Andreessen Horowitz, had a useful post to help founders develop their own idea maze that breaks down the process into four steps that we summarize below:

1. History – Ideas have been tried before, understand what worked and did not work

2. Analogy – Compare ideas from similar businesses and industries to understand parallels

3. Theories – Mine the historical data and research already developer around your idea

4. Direct Experience – The best ideas come from being hands-on with customers

You can find the notes for the excellent Balaji lecture here as he goes into the entire process of ideas, mockups, prototypes, etc.

The AWS Global Head of Startups, Howard Wright, traveled recently to Korea and Japan to speak with startup founders and ecosystem builders, reiterating our strong support for startups in the region such as RareJob (レアジョブ) in Japan and To Be Unicorn (투비유니콘) in Korea that are leveraging AWS to deliver innovative and scalable solutions in the EdTech market! Read more about the impact AWS is having in Howard's LinkedIn post.

While Howard was traveling overseas, Mark was traveling overland to sunny Southern California to meet up with startups and founders. He will have more to share on the Los Angeles / Orange County / San Diego startup ecosystem, but here is a fun pic of Mark underneath the famous Hollywood sign!

With AWS re:Invent coming up in just eight days (?!?!?!?!), we will be teeing up product announcements relevant to startups here and keeping you updated on highlights for those not attending in person. You can absolutely watch all the keynotes and other sessions online however from the comfort of your home or office. All you have to do is register here to gain access to the livestream. You can see who are our keynote speakers and when they are speaking on the AWS re:Invent keynotes page.