The Product-Company Gap

▶ Watch the original lecture

You have heard the term minimum viable product, and you have certainly heard the term product-market fit. As seed investors, one of the first markers we look for when a company is getting ready to raise is: do they have product-market fit? So I do not mean to imply that product-market fit is bad. It is just not enough to build a big company. It is one step, one step along the way.

When you go out to raise a Series A, we want to see a little revenue, and a class of customer, at least a minimum viable segment, for which the product works. You have repeatedly sold that product a couple of times. But there is a lot more that has to happen to get to the next round beyond just product-market fit. The challenge is something we call the product-company gap, and I will tell you, at least two-thirds of the companies we see in the venture world never get across it. This is not just a “crossing the chasm” thing. It is a significant gap in the way people even approach how they start building their product.

A company that failed to cross the gap: Padiant. I started a company called Padiant, a mobile payments company doing QR code payments before it was happening in China and Korea. We actually had to code our own QR code reader from scratch; we hired a satellite imagery engineer to build it. The idea was, hey, we could use QR codes instead of credit cards to pay for stuff. We had started companies before, and we knew it would be hard to build a direct-to-consumer company. How are you going to get millions of people to stop using their cards and start using QR codes? So we took a different strategy to address the gap: we went after huge retailers. We had success signing Best Buy and Walmart, big retailers who would use our technology but build it into their own app. We thought, “Great, we’ll use a partner and they’ll deploy it for us, so we don’t have to do the whole direct-to-consumer thing.” The problem was we had no idea how hard it was to get an IT department at Walmart or Best Buy or Target to actually deploy this stuff. See that little payment terminal on the counter? That is a ten-year replacement cycle. Doing anything on that payment terminal is like brain surgery for a retailer. We ended up selling the company to PayPal because we had good core technology, but it never turned into a huge company. We got acquired doing about $10 million in revenue, and even at PayPal, with all those resources, we could not get across the gap into being a scalable company with millions of users.

A company that failed on the technology side: Actuality Systems. [Told by Greg Favalora] I am going to tell you a five-minute version of an hour-long story about inventing and almost commercializing 3D displays. The company was Actuality Systems. We created a device that looked like a crystal ball, about the size of a human head, that would create floating 3D images you could see without glasses, from any point of view all the way around, like a hologram. It worked by shining thousands of patterns of light onto a rotating screen 10,000 times a second. It was a 100-million-pixel display, the highest-resolution display ever built, and it took technical miracles: we had to beg Texas Instruments for years to give us the DLP chip, and there was no data sheet, so we had to reverse-engineer it.

The overarching lesson: even if you have an awesome product, you really do need to deeply understand at least one market where warm-blooded human customers will write you actual checks you can cash in a bank. And you need more money than you think, especially with hardware. In phase one it took me two and a half years to raise a measly $1.5 million; I talked to 40 venture firms and a ton of angels. In phase two, my board lovingly called it a many-year fishing expedition looking for a market. We would go to a whiteboard, write market segments across the top (mechanical CAD, pharmaceutical design, luggage scanning, military visualization, video games, oil and gas), write the ten biggest accounts we could reach down the side, and work our tails off to get into them. And here is the seduction that will trip you up: in every case they said “This is the best thing ever, your parents must be so proud, I want ten of these.” You say “Great, write me a check, it’s $100K.” They say, “Oh, I don’t have $100,000, can you show me a paper explaining why it’s worth $500,000 so my check earns me $400K in profit?” And you say no, and you move on to the next market.

The lesson: if you are going to spend a ton of money on engineering, do not just know the names of markets or their workflows. Understand who would write you a check, why they would make money from it, and map the living heck out of the flow of money. As an engineer, I had an anti-salesperson, anti-marketing bias: if you are not using diodes or writing code, you are wasting my time. I was so wrong. It would have been much wiser to budget for a really good product marketing person early, someone who could deeply understand a customer, define the market, and write the marketing requirements document that tells the engineers what on earth to build. We raised $15 million over 12 years, and all it was enough to do was make a product that just barely made a 3D image, then put lipstick on the pig year after year. We never figured out the market. We pivoted the company into machine-vision software for planning prostate cancer brachytherapy treatment, and almost got there, but in 2009 the market tanked. In the end we held onto the patents (100 filings became 30 applications became 20 patents), and after a year and a half of everyone saying no, we finally sold the portfolio, wrote small checks back to our 70 angels, and had a small exit. So: raise more money than you need, especially if you are selling atoms and photons rather than bits, and really understand the market.

A company that got across the gap: YouTube. YouTube, prior to Google, was founded around 2005. In one year it was one of the fastest-growing sites on the internet, 20 million users a month, absolutely crazy. In less than two years it was acquired by Google for $1.65 billion, which in 2006 was an astronomical amount. But before they were acquired, the platform was falling down. It was getting way too expensive to run. There was no economic model whatsoever, and no prayer of getting across the gap without someone like Google. What changed? They figured out how to monetize with advertising. YouTube alone inside Google is now something like a $30 billion business. So it is possible to get across, and it does not have to be Google. But having the best product is not enough. YouTube was one of the best products in the world and one of the fastest-growing internet sites, and it still could not get across the gap on product alone.

It is not just the product: the iPhone. Steve Jobs is probably the greatest product manager who ever lived, all about design and the aesthetics of the device. What was the big innovation of the iPhone? People say the touchscreen, and yes, the full-screen hardware form factor was innovative and gorgeous, a luxury item coming out of the gate. But the real innovation was the App Store: the ability to have tons and tons of apps so the device could be completely customized to you. And not just the App Store. One year after it launched, they introduced in-app purchases: download an app totally free, then upgrade with an in-app purchase, and Apple takes 30% of all of it. So it is not just Steve Jobs thinking up a beautiful device. It is the go to market and the pricing that turned it into a blockbuster business.

The expense flip. Here is something you need to know in advance. In the early days of a software company you spend almost all your money on development. When we invest in seed companies, it is often one maybe-non-technical founder and the rest engineers: no marketing people, no director of sales, no customer success. But as you scale, your expenses flip. Once you get past MVP and past founder sales (where the founder is the one salesperson closing all the deals), you start spending far more on sales and marketing than on product. Marketing is your pipeline; it is about pull. You create awareness to have a pipeline, and without a pipeline you do not have a business.

In the SaaS world (software as a service, where the software lives in the cloud on a subscription model, now the most popular software pricing model), benchmarks have emerged. There is a rule where, for a mature company in a mature line of business, roughly 40% of revenue goes to sales and marketing and about 20% to product and R&D, so SG&A runs about triple R&D. (GNA, by the way, is general and administrative, the back-office stuff.) These are literally benchmarks investors use to value companies and decide how much to invest, and they matter more and more now that valuations are compressing and everyone cares about metrics.

The data backs this up. Crunchbase did a survey across software segments looking at how R&D spend as a percentage of revenue changed as companies approached IPO. Salesforce (which invented the SaaS model) and LogMeIn, both profitable, show product spend trending down toward IPO. MongoDB, a more recent IPO, is not profitable and is still trending toward that 20%, partly because they are still spending a ton on R&D. Twitter’s R&D expenses generally trended down. Meta was fascinating: heading into its IPO it was a super-efficient product organization spending 7-10% of revenue on R&D, but now, investing in the metaverse, they are likely spending 30% of revenue on just that product. If you look at Apple’s financial statements on EDGAR, they spend around 2% on R&D, roughly triple that in SG&A. It is not a low figure in absolute terms because they are a big company, but 2% is striking. So it is not a hard-and-fast rule, but on a per-product-line basis, and certainly when you are just starting up, be prepared for the fact that it is going to flip.

Architect the company on purpose. So what if you thought about this like an architect and planned up front? What if you said, “We want to end up a very profitable company where we make 20% and drop it straight to the bottom line, and we want a product that flies off the shelves”? How would you build the company, and how would you build the product, to make that the case? The more effectively you build the product, the more specifically you intersect the market, and the more friction you take out of the go-to-market process, the more likely you are to have a successful business model at the end. That is what the rest of this session is about.