It’s been quite a ride for OnePlus, the two-year-old Chinese smartphone maker. The startup made waves last year with the release of its first model, the OnePlus One, which dared to directly compete with the flagship models of industry leaders like Apple, Samsung, and HTC for half the price. It was a bold claim—bold enough to sell thousands of units, and bold enough to merit the company a mention in a 2014 Fortune magazine feature story. This year, the company doubled down on its hubris with the release of the $329 OnePlus 2, which it billed as a “2016 flagship killer”—that is, a phone so good that it would best next year’s top models.
Not everything has been so smooth for OnePlus. A small scandal erupted when one of its marketing campaigns asked women interested in its phones to upload photos of themselves and subject their appearance to a popular vote. And reviews for the OnePlus 2 have so far failed to match the company’s pronouncements, even as it sells ever-more handsets.
Fortune recently spoke with OnePlus co-founder Carl Pei about his company’s strategy and lessons learned over the last year. Here is that conversation, edited and condensed for clarity.
Fortune: What makes OnePlus different? What does it stand for?
Pei: There is some confusion about the vision of OnePlus. I see a lot of people say, “It’s a great product with high specs at a low price.” That’s kind of a little bit different from what we set out to do.
The entire thing started as a coffee-table conversation in August 2013 between a bunch of early OnePlus employees—at the time, just friends. We had our phones on the table and we noticed that everyone was using iPhone. This was surprising—some of us were working at other smartphone manufacturers making Android phones.
So we asked ourselves, “Why do people around this table all use iPhones?” Our conclusion was that no one else cares about the product as much as Apple does. It puts the most effort into making a great product.
So then the question became, who is number two? What is the second-best product focused smartphone company? And that’s where things got pretty interesting. There was no clear second. Some people said Samsung. Others said HTC. Some said Sony. That’s where we saw opportunity. In the end we decided that no Android manufacturer cares about, or focuses on, the product as much as Apple. That’s the spot we thought we could fill. We were very confident in our product management abilities.
You learned many lessons with the introduction of your first product, the OnePlus One. Tell us about them.
From a marketing standpoint I think the biggest lesson learned is that there’s a huge gap between and idea and execution. I think our biggest blunder in terms of marketing was the Ladies First campaign. The initial idea was not a bad one. We knew it was very hard to get an invite to buy the phone and we saw that our community was very male-dominated. We wanted more women to be involved in tech. The idea was, let’s find a way to make it easier for women to buy our phone.
But the way we executed it and the way we communicated it was not very tastefully done. Once we saw that it was a problem, we removed the content immediately. It ran for about four hours, but after only an hour it had already spun out of control. Going forward, we’re focusing a lot on the execution piece.
We’ve seen the biggest improvements on the product side. We owe that partially to feedback from our community but also internal product vision and direction. For example last year we introduced a bunch of different back covers—you could replace your OnePlus One back cover with different materials like bamboo or wood. They were really well received—but we didn’t execute on this as well as we wanted to. We placed the antenna inside the cover, making it difficult to manufacture to our standards. The factory that made the bamboo backs for us had yield rate of 30%. That meant for every 10 covers they made for us, we would reject seven of them.
This time around, for the OnePlus 2, one of our hard requirements during the product development phase was making sure those covers are easy to make, swap, and be available at launch.
As a product-focused company, you have to have your own vision. You cannot just build everything your users want. Some of our decisions were made for ourselves. One is the metal build. Last year we were a bit disappointed in ourselves because we didn’t manage to make our bezels metal. This year, we made that addition.
We’re trying to push the envelope a little bit in the commoditized, boring smartphone market. For example, our new phone is the first flagship smartphone [to ship] with USB Type-C—a new standard that’s coming to all types of electronics. It’s in the new MacBook and Google’s Chromebook Pixel. People were asking us, “Why did you put it in a smartphone now, instead of waiting two years when it’s a standard?” From our perspective, someone has to push the envelope a little bit.
Let’s talk about features and the future. One feature left out of the OnePlus 2 was NFC—near-field communications technology. It’s something that has been present in phones for a few years now. With contactless mobile payment services like Android Pay and Samsung Pay upon us, why leave NFC out?
NFC was in our previous generation phone, the OnePlus One. We found that not a lot of users were using it. So we decided not to include it in the OnePlus 2. We know that Android Marshmallow has Android Pay and that will partially rely on fingerprint scanners but also on the NFC chip. But we think mass adoption for that, in terms of in-store retail, is not going to happen in the next 12 to 18 months. So if a person upgrades their phone every 12 to 18 months, and we put it in our next-generation phone, it’s not going to be a big issue.
I’m sure there’s a small subset of users that use [NFC] frequently, but for the vast majority of people they’re worrying without the need of worrying.
Who is your ideal customer? You have a big fan-base in Internet forums. How do you take that sentiment and broaden it to include the average consumer?
We are asking ourselves that question: How do we reach out to a larger, more mainstream user base at the same time as keeping our voice and not becoming like everyone else? And not losing what originally made our fan base like us.
We are not necessarily after the users who are out for a bargain because they are not going to be loyal. Today, they are going to buy your product. Tomorrow, they’re going to buy whatever’s cheapest. Instead we hope to find users that really appreciate the thought we put into the product and our vision for it.
I had a gut feeling that people would actually want to buy our phone even if it was more expensive. That’s why we had our campaign last year called “Smash the Past,” where people had to submit their flagship smartphone that was more expensive than the OnePlus One in order to get the OnePlus One for free.
It was basically a trade-in program. They had to smash their phones on camera and upload it to YouTube. The response was overwhelming. Within 24 hours 100,000 people had applied. Within six days there were 140,000 applications. A lot of people wanted to give up their more costly device to get our device.
The OnePlus 2 runs on Google’s Android mobile operating system, with what you call OxygenOS. What is OxygenOS?
Google has been making a lot of changes to the Android user experience over the past two years. When Android first came out it was not a very consumer-friendly operating system. The OEMs put their own skins on top of it. But since [Android version 4.4] KitKat, and now [Android 5.0] Lollipop there have been big improvements in the user experience.
When it came time for us to build an OS, we didn’t want to reinvent the wheel. We just wanted to use what our users already like—basically the pure Android experience—and just build on top of it. For example, the status bar on Android. If you have a big phone, it’s hard to reach the status bar with one finger. On OxygenOS, you can swipe down from anywhere and have the status bar appear.
What is it like to be the underdog in a very competitive smartphone market—and a company based in China, the world’s largest smartphone market, for that matter?
It’s great to be an underdog, because throughout history the underdogs are the ones who have really disrupted other people. Especially when they are under the radar and suddenly they come up and change the game. Since we launched the OnePlus One and it started getting traction in the United States, we’ve seen a lot of brands start to follow this model.
Ultimately igniting this unlocked phone market in the U.S. is going to help consumers because carriers are going to have less influence on what consumers can buy.
But it will also help OnePlus. When things become more transparent and more fair and fewer companies are meddling in the consumer market, the quality of the product will shine through on its own, through word of mouth or Internet reviews.
Regarding the China aspect, every time someone has mentioned the China thing it’s always been kind of a weird feeling in my gut. A lot of times we’re labeled as a Chinese upstart. I don’t really understand. I guess a lot of people write it because we started in China and that’s normally how it’s described. But I think we’re a lot different than that. Unlike a normal Chinese company, when we launched we launched worldwide, in 17 markets. And our staff is from 14 or 15 different countries. We now have offices in Bangalore, Singapore, Taipei. We’re opening an office in London. It has become a very distributed team. I think we’re past the point where companies need to be labeled by their origin.
How many people work at OnePlus?.
Just under 900 people.
That’s a lot of growth in less than two years.
It’s also a big burden in terms of cash flow. Last year we eventually broke even, but this year it’s looking like we are going to lose some money because of the increased human resources.
That was going to be my next question. With such aggressive pricing, how are you making money?
I should add that it actually costs us more to manufacturer our phones because of our smaller scale. We have nowhere near the scale of a bigger manufacturer. We cannot get the same price on components.
How do we make money? Obviously because we are using the Google suite of services, and we haven’t built any of our own, we’re not making money on software. We have to make all of our money on hardware. The difference is we don’t spend a lot of money on marketing. There, we save a bunch of money.
And we sell directly to the consumer. So instead of giving someone 25% to 40% for selling our phone for us, we’re removing all of that. When a person buys a phone, maybe 30% of that price goes to the store and the various distributors in-between. I was surprised to read recently that, on average, LG made $0.012 cents per device. I read it on Twitter and retweeted it and wrote, “Hey, that’s less than we make.” It’s much less than we make. And OnePlus is a company that’s doing it for the user and not all about making a bunch of money.
We have a margin, it’s just pretty low. But our costs are low, and most of our margin comes from the fact that hardware prices fall with time and with scale.
In the first few months of the OnePlus 2, we’re not going to make money. But as the prices drop for hardware and we achieve more scale, our bill of materials will drop. Selling for the same price, our margins will increase.
During your OnePlus 2 launch event, you mentioned you sold 1 million OnePlus One units. Is that right?
Yeah. In the first year, which is actually a half-year because we started selling it in June, we sold a million devices. And since then we’ve shipped 1.5 million OnePlus Ones.
What number would make OnePlus 2 successful to you?
This is a super risky metric. It’s not a metric we’re ever going to chase. We think that success will come from focusing on the more important things. If you’re doing the important things right, everything else will fall into place.
On the contrary, if you focus on a big IPO or a big fundraise—or valuation, or growth, or sales—I think that you would be chasing the wrong metric, causing you to make more short-term benefit decisions that will help build the company in the long run. So for us, we’ve never made sales a priority. We’re just happy if our users really like our product and are happy with our company. That’s what important to us: the product and how we treat our users.
Do you ever see getting rid of the invite system to buy phones?
Yes. The invite system is necessary at this point because of our small scale. We cannot afford to take any inventory risks. We have to always make sure that we have a little bit more demand than we have supply. Otherwise we end up with inventory and we’re already operating with a very low margin. We’d have to dump that inventory to release cash flow. That could make us lose money and go into the red immediately.
We were a little bit over-confident last year. In the beginning we had very few units available. We weren’t sure how many people would want to buy our product. Toward the middle of its life-cycle, we thought, “Wow, this is a really popular product. Let’s buy a lot more components.” But toward the end of the life-cycle we had to discount some of those components—not the final assembled phone—on the second-hand market. We lost a few million dollars there.
So this year we’re still doing an invite system to buy a phone, but launching with a lot more inventory. So it’s going to be a much better experience. Once we feel secure about our demand and feel we have a grasp on how many people want to buy this phone, then we will loosen up the system. We could start with open sales maybe one day a week or move away from invites altogether.
If someone signs up for an invite today, how long will he or she be waiting?
It’s hard to say. We have a lot of reservations now. Within 24 hours we had half a million. 72 hours after launch, one million.
But it’s really hard to say how many of these people actually make a purchase. Last year, initial conversions were around 80% but they dropped to 10% towards the end of the product life-cycle. This time it’s really hard to foresee. I’d say within two months we can start really fulfilling all the demand we have.
Subscribe to Data Sheet, Fortune’s daily newsletter on the business of technology.