Imagine we invented a new technology that let semis draw a small amount of power from human motion. This is impressive but does not bend the laws of physics, it has a set of obvious applications today, and our team has a few ideas of bigger things down the road. We found a company to commercialize the technology, David Semi and set out to commercialize it. How big could this company grow? Would it be able to withstand competition from the semis giants and eventually go public? And beyond that, would it ever be able to grow into a large, multi-product company that stands with the industry majors?

To start with, the company would need to first prove its product and technology. The world is awash with good ideas, execution and delivery are much harder. This company sells into consumer electronics so it first needs to prove itself to customers, all of whom are much larger and accustomed to pushing around semis vendors much bigger than David. We have seen many companies get stuck at this phase. Big customers can become a massive resource drain, requiring extensive customization work. There will always be one or two willing to kick the tires and indulge a start-up, only to walk away after learning all they could. We have suggestions on this, but will save them for a future post.

Let’s say David manages to get a handful of customers engaged to the point that they are willing to pay for serious demos and include our chip in some of their upcoming products. This is a major hurdle, and companies that achieve this find themselves much better positioned for raising funding. One or two demo customers does not say much, but five or six paying customers effectively validates the technology, marking a considerable de-risking of their prospects.

At this point the company is well positioned. They can start to raise money, moving beyond survival to growth. We can start to build out a product portfolio – add a high-end and low-end product, add a module, maybe design a custom implementation for a particularly large customer. At this point, the company can even start to think about approaching Apple. However, the limits of growth also start to emerge.

The company is likely selling into someone else’s systems. The consumer device OEMs, the end-customer, is requiring our product, but we have to put our chip into a board that is likely designed by a much larger chip company. As soon as the OEM picks our product, those big companies start to view us a competitor. Chances are they have already dabbled in our technology, but a small, niche, unproven technology got lost inside their behemoth organization and was abandoned along the way. The big companies will now restart those efforts. David Semi has put in a lot of research work into our systems, we are confident that we have a considerable head start, but the clock is ticking. The big companies do not have to match our performance, they just have to get to close enough and then throw their product into a bundle.

Now our technology is fairly disruptive, it opens up new performance frontiers and makes old ideas feasible suddenly. End customers’ engineering teams are delighted to have new options. This leads to new opportunities for us. In addition, customers get trained on our tools, and they appreciate our nimble customer service with quick responses to their questions. This entrenches us with their engineering department. And once their competitors launch products with our technology, top management starts to pay more attention as well. This opens more doors for us, but it also sets an alarm bell ringing in the purchasing department which hates reliance on a single vendor for any component. This opens those same doors to our competitors as well.

As the years pass, our growth surges, passing $10 million in sales, then $50 million and then $100 million. We have visibility to getting our technology installed into every smartphone, AR/VR headset, and whatever other new devices emerge over the horizon. This is a massive opportunity. On cue, the investment bankers start to arrive. (In fairness, the good bankers have already been here for a while.) They start to talk about an IPO. Our finance intern drew up a long-term model last summer in Google Sheets. The bankers take those numbers and paint a fairy tale ending. Our $10 chip is going to end up in 50% of smartphones – $5 billion in revenue in five years! We have been heads down fighting for every socket for years, managing HR headaches and operational hiccups. The bankers take us to nice dinners and their siren song starts to sound very appealing.

Do we take the plunge?

On the one hand, we know our technology deeply and believe in our heart of hearts that we have sparked a massive opportunity. We have barely scratched the surface of the places our technology will work. That crazy revenue figure is maybe not that crazy.

On the other hand, there are some messy realities. We still have to run the company every day, and every minute we spend thinking about the IPO will be one less minute we can spend on the business – launching the next product, honing the roadmap, ironing out conflicts. It will be a distraction. Moreover, our revenue is still small today. If we can do $150 million next year, we can probably go public at around $1 billion. The markets are mixed, the IPO window is open, but will market conditions persist? Also, if we are being honest we have three customers that consume 80% of our sales (and if we actually won Apple it would be even more concentrated). Last week, we talked to one of them and they sounded a bit less enthusiastic about our roadmap and volume commitments for next year. Our EDA sales rep let slip a rumor that another customer is looking to design their own part, they will never get it to work but it would be an ugly headline. Maybe the IPO will not be quite so easy.

But let’s say we take that path. We read this book, so the IPO goes well, but now we are a small cap semi stock with hordes of investors crawling all over our numbers. They are new to our story and we have to spend a lot of time explaining the big vision to them, many of them do not seem to see it all, and just want to ask about the next quarter’s gross margins. We gain a bunch of new customers, lose a big one. The business is doing well enough, but the Street seems yet to make up its mind. That $5 billion revenue figure still looks achievable, but may take a bit longer, glad we did not share it with the Street. Our competitors are catching up. Their products are not fully performative, they do not have many features, but they still rack up the occasional design win. There also seem to be a dozen companies in China clawing away at the market; they are not getting far but they do not help us with price negotiations.

Our biggest problems is that the Board is split. Half think we should sell the company, half think we should bulk up and build a bigger platform. Bulking up is not going to be easy. There are no obvious targets. We have grown by adding new customers and expanding our penetration across their products. There are a couple products we see sitting on boards next to us, but these are all supplied by the big companies, so there are no easy targets to buy there. We could buy someone in another field. Again, we are not spoiled for choice, more importantly these are all far from our core competency, technologies we do not fully understand, integrating them will not be easy, and there is no obvious reason why a customer would buy both our products in some kind of bundle.

We have seen all these stages play out at companies over the years, and it has been at least a decade since we have seen a company really emerge from this chasm. There are many examples of companies that tried to bridge the gap – Teligent, SiRF, Cambridge Silicon Radio – all of which ended up getting acquired for less than than the executive once hoped. Image sensor maker Ambarella is a case in point. They are trading well above their IPO price, but other than a few short-lived surges, have seen their stock range bound for close to a decade.

Admittedly, there are differences. For instance, those companies all went public with very high levels of customer concentration. That being said, they all fell victim to the same pattern – the competition caught up. All of them thought they had a multi-year advantage over the competitors. In many cases that was actually the case, but after a while, the differentiation between the nimble, stand-alone company and the large, multi-headed giants dwindled to the point that customers no longer preferred one over the other. And then scale economies started to matter a lot. Competitors drive much more volume through the foundries and thus get much better pricing. So customers chose the company with 80% performance at 70% of the price, and the stand-alone companies had nothing to fall back on.

We do not want to play the role of pessimist. There is still ample room left for innovation in microelectronics. Someone with a hot, new technology will spark a shake-up, grow a product, then a category and then maybe a way of life. That is still possible, but not many companies will make it all the way there.

Photo by Steve Barker on Unsplash