The Innovation Stack: Building an Unbeatable Business One Crazy Idea at a Time

by Jim McKelvey

Amazon, the scariest monster on the planet, had copied our product, undercut our price, and was going to eat our brains. Then, without warning, on Halloween in 2015, the monster stopped the attack and handed us a treat. This treat was better than any bag of candy. Not only did Amazon discontinue its competing product, it also mailed the product’s existing customers a little white Square card reader in a smiling cardboard box.

One of these patterns often appears in businesses whose aim is to square up—bring fairness to a previously unfair system.

You’ll find more entrepreneurs in twelve-step programs than you will in business school.

Marcia Dorsey ran a coffee shop in the neighborhood and was our supplier of this chocolate stimulant. Marcia was friendly and funny and curious about why we were such massive consumers of this particular product. John told her about our little company, about our recent mistake, and that we did things with computers. “My son likes computers,” she mused. John didn’t miss a beat. “Hey, would your kid like to make $50 the hard way?”

Jack hated the auto industry for the role it had played in segregating our cities and what it had done to streetcars. He swore that he would never own a car.

A couple of ratios help illuminate the crime scene. Credit card vendors were making 0.04¢ on every dollar ($0.3 billion / $788 billion) they processed from their large merchants. Now compare this to 1.8¢ on the dollar, the profit they were making on small merchants ($2.4 billion / $130 billion). Their profit margin from small businesses was forty-five times higher than from billion-dollar corporations. I rechecked my math three times before that number sunk in. Small businesses pay forty-five times more than the giants do. We had identified a big problem and a good reason to start a company.

Jack and I started the presentation by taking the venture capitalists’ money in more ways than they were expecting. After our brief introduction to how messy and ugly the current credit card ecosystem was, we asked each prospective funder in the room to take out a credit card. We had developed a crude but usable card reader and, with this prototype plugged into the headset jack of an iPhone, we read and charged their credit cards, an act that by itself broke more than a dozen regulations. The amount we took varied from $1 to $40, depending on how much we liked the VC in question. Nobody had ever seen anything like this before. Some didn’t believe that it would actually work. It did.

In Square’s case, this border was defined by about $10,000 of annual sales. Merchants who sold less than this amount were all but excluded from accepting credit cards. If you try to enter a market at the top or middle, there are very powerful companies already in that space that will thwart you. But down at the bottom of a market lies an unguarded economic border.

Square’s initial team consisted of three guys and a cat named Zoë in a studio apartment in San Francisco. We divided up the work. Jack coded the server software. Our only employee, Tristan, wrote the iPhone client software. Zoë sat in Tristan’s lap to compensate for our lack of health insurance. Since I was the worst programmer of the group (with the exception of the cat), I did everything else.* My primary responsibility on that first day was figuring out the credit card business.

Jack is very quiet and comfortable with long silences; I’m basically the opposite but have trained myself to never interrupt anyone. Sometimes the key to an explanation is being quiet long enough for the audience to catch up.

We might never have been able to convince the Visa executives to change their rules by showing them how Square would help them. But since we’d already gotten American Express, Mastercard, and Discover* to accept Square, Visa now had to point its nose back toward the herd. Mastercard and Visa are separate organizations, but they move with the synchronization of samba dancers when it comes to innovation. Visa just does it backwards and in heels.

Two weeks after Square launched its product to the country, I got a call from the CEO of one of the largest payment companies in the world. He took me to dinner at some restaurant in New York City where every entree contained at least one word that I could not pronounce. He downed a martini, ordered another, and then spent the next hour telling me what an idiot I was for trying to allow small merchants to accept credit cards. He explained how people like my friend Bob had bad credit, were unreliable, were difficult to support, and were too small to sue if things went wrong.

I ran into a convenience store to look for chocolate-covered espresso beans. When I returned to the car, Jack and Anna had named our new company Squirrel. Not bad.

Jack had had similar troubles naming Twitter, which was originally called Twitch in homage to the motion people made when their phones buzzed with a new message. Twitch sounded too much like a neurological disorder, despite the fact that another company successfully claimed that name several years later. His team eventually solved the problem by consulting the dictionary and finding twitter on the same page. Jack did the same thing for us, starting with squirrel and eventually landing on square.

This problem-solution-problem chain continues until eventually one of two things happens: either you fail to solve a problem and die, or you succeed in solving all the problems with a collection of both interlocking and independent innovation. This successful collection is what I call an Innovation Stack.

And so we began with the simplest invention we could imagine: a known price. One price, a percentage of the transaction, for everyone, at all times. No hidden fees. This was exactly the opposite of what everyone else in the credit card industry did. In addition to creating a new level of trust and transparency, our simple price was also simple to explain. Customers knew what we charged, and they could tell others.

This decision had a painful consequence, because we were still paying the per-transaction fee to the card networks. Which meant that on small transactions, we actually lost money. We were forced to recoup these losses by having a massive volume of other transactions. We needed to scale up. Fast.

In 2009, the cheapest portable credit card reader on the market looked like an orthopedic shoe and cost $950. The original Square reader cost 97¢ to build. Our reader wasn’t just cheap. Fifty dollars would have been cheap. It was ridiculously cheap. With our costs 979 times cheaper than the alternatives, we were able to just give them away.

We had very minimal customer support options at the beginning. We had no phone number, just an email address where questions were answered by a small team of fast typists. It might sound crazy that people would trust their financial transactions to a company they couldn’t call, but few people objected. Then again, if they had objected, they would still have had to let us know by email. Customers who preferred to have a number to call presumably signed up with a traditional credit card processor and then had plenty to complain about.

We took our lack of live customer service very seriously. It was not just a way to keep our costs down, it forced us to develop more innovation to further reduce the need for customers to contact us.

Traditional credit card processors took several days to pay, which was absurd. This was another holdover from the early 1980s.

Speed was critical for several reasons. It delighted customers and kept our growth humming, but more important, it eliminated all those “Where’s my money?” support calls.

Our simple pricing allowed us to know what amount to send to the merchant, which we did daily. The rest of the industry had to wait days to know the cost of a charge, and then those charges would be debited monthly.

When Square started, most small merchants were paying over 4 percent for their credit card services. News about our price of 2.75 percent spread through the small business community like a cold in a kindergarten class. We never paid a penny for all this promotion. In fact, we didn’t pay for any promotion at all. Low

By taking that risk on our own balance sheet, Square was able to massively simplify the sign-up. Putting our balance sheet at risk before our customers’ also gave us the freedom to bet on millions of small merchants that the banks would not otherwise trust.

We want to allow millions of small businesses to accept credit cards for the first time, so we have to make it easy to sign up. We need easy sign-up, so we have to design simple software and eliminate paper contracts. We have millions of people signing up, so we have to keep our customer service costs down. We need to keep customer service costs down, so we have to have simple pricing, and net settlements, and no hidden fees, and no paper contracts. We need to have a low price, so we have to save money on advertising, so we have to have an amazing product, and hardware so cool that people talk about it, and a product that they can explain without our help.

We discovered that Amazon had copied our hardware (albeit as a black rectangle), had undercut our price by 30 percent, and was offering live customer support.

By 2014, however, Square’s product line was becoming more complicated, and we had already planned to add live customer support as an option. But customer support is not something we could implement overnight. To provide a good live customer service experience takes months of planning, hiring, and training, not to mention finding a place for everyone to sit. We weren’t going to rush, or rather, we already were rushing but could not increase the pace just because the world’s most dangerous company was advertising that it had phone support and we didn’t.

but I have never seen a nonreaction like we had at Square. What was remarkable about when Amazon attacked Square was that our energy level didn’t change at all. Everyone knew what was happening, and nobody did anything differently.

Copying almost always feels comfortable, but it will never produce the thrill of invention. Copying is almost always the right answer, but it will never produce transformative change.

So far, we have examined fourteen elements of Square’s Innovation Stack. Let’s say that a company has a 75 percent chance of copying any one element successfully. Since the company in our example is Amazon, let’s give it an 80 percent shot at getting each one right. So, with one element, it’s at 80 percent. To get two elements right, it’s got a 64 percent chance. And only a 51 percent chance of copying three correctly, a 41 percent chance of copying four, and a 33 percent chance at successfully copying five of the elements we were utilizing every day. You see where this is going. Even a place with all the talent and resources of Amazon can’t escape math. Its chances of copying all fourteen of our elements (0.814) were about 4 percent.

At Square we resisted hiring anyone from the payments industry for years. Actually, during our first week in business I found a payments expert who offered to consult with us, but we terminated that relationship almost immediately. His advice was straight from the industry we were trying to avoid copying, and it seemed foolish to pay someone to explain how to do everything we didn’t want to do. For the next several years, we avoided getting anyone with payment DNA into our company. This allowed us to think freely about what new things we wanted to build and not about how it had always been done.

On November 19, 2015, Square listed its stock on the NYSE,