Founder's Guide
10 min read

Don’t Pick a Product — Pick a Market and Iterate

Share this article!

In 8 months, my co-founder/best friend and I have iterated through four different blockchain products, raised money, and built a company with customers. It’s been a crazy journey — with many learning moments that can be applied to startups in all spaces.

We have moved so rapidly because (1) we focused on the market, not the product, (2) pitched our ideas before building them, and (3) continuously tested our riskiest assumptions.

In this post, we’ll discuss:

  • How we tested our product’s biggest assumptions (including one in Dolores Park)
  • Why focusing on a market over a product leads to great results
  • The psychological pitfall we were able to avoid by not committing


My co-founder Rahul and I discovered we worked well together while consulting on digital strategy at McKinsey. We were building an engineering team for a client and we immediately bonded over our shared interest in coding and blockchain technology.

In January of 2018, we quit our jobs and moved into a San Francisco apartment together. We’re both engineers — Rahul was an engineer at Amazon, Opendoor, and Bloomberg, while I picked it up as a hobby. I had been working on my previous startup, where I had the traumatic experience of being a solo founder (and vowed never to do it again). I’ve since come to believe that the best way to start a company is to move in with your co-founder and hack together all the time.

At that point, we had no income or funding and were subsisting entirely on savings. Around April, we were able to acquire financial support from amazing angel investors that helped us survive through the summer until we landed on the idea that ultimately worked.

Why we picked Blockchain

Last year, I built a failed social app — an undifferentiated product in a crowded market. I came away wanting to start a company in a market that’s small today but could be huge in ten years. That way, there’s less competition and it’s easier to build a novel product. Since blockchain is a transformational technology that enables a new class of user experiences, we decided it was a great space to invent something new.

After analyzing the crypto space, Rahul and I noticed one of its biggest barriers was a lack of compelling user experiences. That led us to work on an augmented reality game for three key reasons:

  1. Gaming is the largest category of apps by the number of downloads (more than the next four categories combined).
  2. Augmented reality (AR) is the fastest-growing segment in gaming.
  3. There’s a large overlap between gamers and early crypto adopters.

Idea #1: Pokémon GO, Crypto Style

Our first idea for a delightful UX was a crypto game in the style of Pokemon Go! Imagine AR overlayed on a map, coupled with partnerships with token issuers. Players could race each other in the real world to pick up digital items with real, monetary value. Like Mario Kart, players could affect each other through virtual spells or attacks.

Rahul and I were excited by the idea but weren’t convinced it would resonate with other consumers. We had some key tests we needed to run.

Test #1 – Validating the idea in Dolores Park

Our first question was whether people would click if they saw it in the app store. A customer only spends seconds evaluating your product, so we needed to know we could succeed in those initial seconds.

To find out, we created a flyer with a mockup of the app and a one-line description. We walked around Dolores Park in San Francisco, asking 100 people, “Would you play?”

dolores park 1

dolores park 2

The reactions were incredible. People of all ages showed huge enthusiasm for the idea of running across city blocks for digital prizes. One guy literally jumped up and down asking for early access! Even a month later, a family emailed us asking for a release update. We had no expectation going in and learned just how much people liked the idea with just a mockup.

That initial test gave us validation that users might actively download a game like this.

Test #2 – Simulating a game in Berkeley

It’s very easy for someone to say they’ll do something, but actually moving your feet to play a game requires real effort. People had said they would play, but we needed to be confident they actually would.

For our second experiment, we gathered phone numbers from the freshman dining hall at UC Berkeley.


We then bought a bunch of coupons from popular Berkeley establishments for ice cream, coffee, and pizza, and put them in a plastic treasure box. We hid the treasure box on campus and sent out the most epic mass text message of all time:


Within moments, students were sprinting across campus. Some showed up in pajamas and slippers, having jumped up from watching Netflix. We had a boyfriend sprinting with his girlfriend’s phone proxy-running on her behalf. It was pandemonium — and Rahul and I hiding in the bushed witnessed the chaos unfold. All-in-all, approximately 20 of the 60 people we texted actually showed up. We couldn’t believe it — a full third of all the people we invited!

It felt incredible to know we could have that impact. We now didn’t have to rely on the hypothesis that people would be interested; we knew because we saw them all actually running across campus.

Test #3 – Testing feasibility for ourselves

Coming away from those experiments, we knew our idea had legs, but we still needed to make sure we were up for the task. We had two open questions:

  1.   Would we be able to build the back-end blockchain piece?
  2.   Can we build the front-end AR game?

To test the first question, we spent a weekend hackathon building an Ethereum smart contract. It allowed us to receive tokens from different issuers and distribute them: success!

For the second, we didn’t know anything about building games, but luckily there was the Game Developers Conference in San Francisco that week. There, our previous successes came to a screeching halt. We learned that games are a homerun business — you either hit it out of the park or you strike out. To make matters worse, building high-quality games also requires an incredibly large amount of capital. Each of these was enough to turn us off. Both? No chance.

Despite all of our positive experiments, building a game was now off the table. We had run two successful experiments, but this last experiment had flopped.

It was time to move on to our next idea.

Idea #2: Tinder for Tokens

Since building a blockbuster game was off the table, we decided to build something simpler — something we would be able to make entirely ourselves. What’s a simple game? Tinder. That idea took us to “Tinder for Tokens.”

Many blockchain companies that create crypto tokens choose to give them away for free, in what’s known as an “airdrop.” It attracts users to the network and validates the market. While there are websites that list airdrops, they’re typically hard to discover. The sign-up process is also often buggy and annoying. Coming from our mindset of building a game, we asked:

What if we could make that process as seamless as Tinder—swipe right to get tokens and left to pass?

In under four weeks, we built an MVP and shipped product. We acquired 300+ users and a 5-star rating in the App Store.

Unfortunately, Tinder for Tokens was hard to scale. There weren’t enough quality token issuers launching airdrops, nor was there a consistent onboarding UI across issuers. While some airdrop signups were a Google form, others came through Telegram, and some were websites, many of which weren’t mobile-responsive. The market was so fractured that it killed our user experience — but helped us identify another problem we could solve.

Idea #3: App to launch token airdrops

Now that we had discovered a fragmented onboarding system with bad UI, we decided to solve those problems by helping issuers launch beautiful airdrops from a standardized platform.

Our initial solution was a web app where issuers drag-and-drop their airdrop requirements. This became a page like the Common App for college admissions: Once a user creates a profile on our website, they opt-in to join any airdrop on the platform.

From there, we just kept iterating to solve customer problems. Token-issuer customers kept asking: “How do I measure the ROI of this marketing channel? We just gave away 2000 tokens through your app, but how do I know whether it was worth it?”

To solve that, we wrote an on-chain analytics tool to track whether those users later bought more coins. For the first time, people running airdrops could measure the effectiveness of that distribution channel.

We then noticed that our analytics tool was used by crypto companies to develop long-term relationships with users. Previously, many of these relationships had been one-off: an airdrop or a token sale, then no back-and-forth. However, with this analytics tool, issuers started to see an airdrop as the beginning of the relationship. We knew that if token-based networks took off, the token issuers would need a tool to manage relationships with their token holders. This led us to start building Token Relationship Management (or TRM for short), the first CRM for crypto.

(Winning) Idea #4: CRM for token issuers

We built TRM over the summer and launched our beta in September 2018. TrustToken signed on as our first customer because they needed a tool to understand their token holders. Their token, TrueUSD, is the largest fiat-backed stablecoin (~$160 million market cap as of November 2018). They use TRM to analyze their organization’s on-chain and off-chain data so they can grow faster and stay compliant. TRM is now being used across the organization:

  • Marketing teams use TRM to launch and measure token airdrops.
  • Growth teams use TRM to understand which user segments, channels, or exchanges are driving growth.
  • Compliance teams will soon be able to monitor suspicious activity and file required reports to regulators, straight from their central TRM dashboard.

In 8 weeks of beta, enterprise logins have doubled week-over-week, nearly 20,000 consumers have created TRM accounts, and we are building partnerships with other leading token issuers to join the TRM platform.

We’re now on a mission to build a world where value flows freely. We do this by helping the best entrepreneurs launch and grow token-based networks and we’re just getting started.

Two common mistakes we avoided

I highly recommend that other founders focus on iteration over ideation. It helped us find a successful business and avoid two common mistakes.

Sunk cost fallacy

In our previous startups, both Rahul and I fell into the trap that behavioral economists call the “sunk cost fallacy.”

People naturally ascribe more value to things they’ve already invested time and energy into. Most coders know this feeling: As you code more, the weight of your codebase creates a gravitational field pulling you closer and closer. The more work you put in, the more you mentally commit.

The more you commit, the less you challenge assumptions. This leads to a counterintuitive conclusion: don’t work too much before shipping product. Click To Tweet

We were consciously careful not to put too much effort into a product before shipping. We wanted to remain open to more promising opportunities. We wanted to maintain our flexibility and have the option to drop the project.

Building a product before testing it with customers

Many founders put in a lot of effort before testing their product with the market — something I experienced firsthand.

My failed social startup was a marketplace for local experiences. I wrote 200,000 lines of code before acquiring my first user. This proved to be a terrible mistake.

When I talked with my first potential user, they didn’t even look at the app. They just asked for a description and, upon hearing, said: “Meh, so how is that different from Meetup?”

I incorrectly assumed people wouldn’t understand it until it was built. I would have saved a lot of time and money if I had started by testing people’s reactions.

Our lessons learned

The lessons we learned from our fast-paced experience in blockchain can apply to any industry. No matter what sort of company you want to build, you can pick a market, pitch your idea, test your riskiest assumptions, and follow the smoke.

Pick a market, not a product

When Rahul and I started working together, we had two motivations:

  1. Solving problems in blockchain
  2. Working with each other

When your core motivation is the market and your team, it’s much easier to iterate and pivot. Since we were motivated by the space as a whole, not a specific solution, we were comfortable dropping our current solution to tackle bigger and bigger problems.

Pitch your idea, don’t build it

Most entrepreneurs are so excited about solving a problem that they start building a solution right away. As my 200,000 wasted lines of code show, that’s a sure-fire way to fall into the sunk cost trap.

There’s a famous quote often attributed to Henry Ford: “If I had asked people what they wanted, they would have said a faster horse.” This line of thinking leads entrepreneurs to assume customers don’t know what they want. To continue the analogy, many entrepreneurs just start building the car.

An alternative approach — describe the car instead of building it. You might say:

  • Imagine a new type of horse that you don’t have to clean up after.
  • What if it could fit four people, not just one?
  • Instead of riding, it moves at the mere press of a pedal.

To some degree, of course, that quote is right: customers don’t always know the best solution to their problems. When you describe a solution, however, their level of excitement can be predictive of future demand.

Identify and test your riskiest assumption

One bad assumption can take down a company, making it extremely important to identify and test your riskiest ones. Once you identify an assumption, the fastest and cheapest test usually doesn’t involve writing code.

  1. At Dolores Park, we tested whether people would download our app from a one-line description.
  2. In Berkeley, we tested whether game players would be willing to participate in the physical world.
  3. With TokenDrop: Tinder for Tokens, we tested whether the market was big enough for token airdrops (turns out, it wasn’t).

As we demonstrated with those experiments, sometimes you can test really quickly by sending a lot of text messages, setting up a quick website mockup, or using photoshop and walking around a park.

Our first two experiments only took one day each. Our third took a weekend, and our fourth took two weeks. Our entire Tinder for Tokens idea took under a month from ideation to shipping.

Many people spend months, even years, building a startup only to discover the idea is fundamentally flawed. Since we tested our assumptions, we only spent weeks failing, which led us much quicker to a successful conclusion.

Conclusion: Iterate by following the smoke

In every version of the company, we’ve solved a bigger and bigger problem. With Pokemon Go, it was user acquisition. For TRM, now it’s the entirety of user relationship management for crypto networks. By noticing where there’s more intense customer demand, we’ve continually iterated to bigger, more important problems.

In any company, no matter the market, you can keep on iterating. You can follow the smoke until you find a problem that’s so important, with so much potential, that it supports building a legendary company.

Share this article!

startup straight talk

A collection of our most popular blogs in audio format.

  • I very much agree with the finding a big enough market, I totally disagree with the rest, again every founders finds his way to success. For example, I found all my fuel from wanting to personally build my product to solve my personal pain. & explore a alternative to the market, more importantly wanting to share that with the rest of the world. The second part is very essential, which is to figure out whether or not that corresponds to a general interests, and validating the assumptions from there and launching!
    I want to thank our menot Andrew from, in his own words tony why haven’t you launch yet 😂.
    So getting your Early users is quite critical, And then build something that surpasses their expectations.

    Again, in our arena there’s no doctrine. Just wanted to clarify, that there’s many ways to get there, but to me it comes down to mission and market and then launching quick to then iterate to PMF.


  • Esteban is co-founder and CEO of TRM, a CRM for token issuers. Before that, he worked at McKinsey & Company, where he built the engineering team for Generation, the world’s largest global youth employment program.

    Published In

    Founder's Guide

    Best practices for building your startup.

    Browse all 44 Articles

    Up Next