Chasing the Wrong Rabbit: Vertical Integration

“If you chase two rabbits, you will lose them both”

Although I made almost all my points in this past article about the infamous Make vs Buy dilemma, after observing some current silly “pissing contest” of sorts in the space industry to see who is the most “vertically integrated”—and attached to that, who has the cheapest spacecraft cost, although that’s an opportunistically obscure metric— I’ll add a few more things which were left unsaid in that post from last year, or at least not said with enough clarity.

First of all, alas, engineers. What a particular bunch of fellas. So prone to imitation, and always observing everything through a deformed lens which distorts all that comes into its field of view. Let’s face it: the vertical integration fad has an undeniable root in SpaceX’s known approach of making everything themselves and the subsequent “monkey see, monkey do” which strongly prevails in the dysfunctional NewSpace family. Because SpaceX is currently pouring the skies with satellites at a pace no one else can match, many decision-makers in space are in absolute awe, mentally colonized to believe doing whatever SpaceX is doing is the only possible way to become yet another SpaceX. Perhaps somewhat depressingly, SpaceX—and more precisely Starlink—are frequently used by investors to benchmark new space startups, even if doing so would mean comparing apples and oranges. All in all, SpaceX might have good reasons to pursue a more vertical approach considering its scale, its organizational life cycle stage and the market it is chasing. But other companies might not need exactly the same.

It is a rule-of-thumb that manufacturers should vertically integrate only when products and markets are mature, so they can confidently shave off the last dollars of cost. The reasoning is rather straightforward: capture the market, encroach yourself there, and then work on your costs to improve your profits. Now, why is everyone so eager to shave when they haven’t really captured any market yet and, what is more, they don’t have a product/market fit or not even know what their product is? That remains a mystery. Even more intriguing: why can profitability only be achieved “by removing” (the cost of tangible things) and not just by adding more value with better, creative intangible products such as software, analytics, automation?

Most self-proclaimed “vertically integrated” startups (or more accurately, scale ups) out there have been around for almost a decade already. They’re not necessarily newcomers. Ten years chasing the R&D rabbit. And most of them are still “hybrids”, which is the worst of all worlds. This is, believing they are vertically integrated while still having several critical elements in their architectures coming from third-parties and so deeply embedded in their designs that changing sources would require months, if not years of rework.
Now picture an established manufacturer who makes, say, high-data rate downlink radios, who probably puts lots of effort and attention to that because, well, that’s their business. They characterize it, test it, and improve it. Now, picture yourself—a satellite manufacturer—deciding to make a downlink radio in-house (probably along with other things as well, because you’re embracing verticality full-on). You will probably spread your team’s attention accordingly, like the guy in circus spinning plates. You must keep all plates spinning at all times. Hence your quality, your processes and your thoroughness can never match the said supplier’s because your engineering attention needs to be spread accordingly, as opposed to this supplier who can afford to only think of radios all day long. 

Throughout the years, I’ve been part of several companies run by engineers. This means, with engineers also at the helm on the business strategy and commercial side. They all have something in common: they all were, and are, commercial duds. Why? Basically because we—generalization alert—don’t have a sound sense of business: our lens is aberrated to observe things mostly from technological perspectives, so we always prioritize technical prowess versus perhaps the single most important thing in business ever: time to market. And a merciless sense of opportunity.
A golden, simple rule for every business on the face of the earth is: time is money. A boring platitude? Not really. Good timing means markets taken, first-mover advantage, flags captured. If time is money, then you better act quickly and gather the stuff from somewhere, ideally anywhere, and augment with intangibles, for example writing the right amount of software to make everything work together seamlessly and get fast to market. And then amplify profits by making the overall product better and not by just removing costs of tangibles, nuts and bolts. If after 8, 10 years in the game you still need to buy a drill so you can do your own holes to increase your profits, something is odd. Simple as that. But engineers insist on engaging in long, pointless R&D expeditions with uncertain ends, doing a lot of painful things they can’t reuse (the famous non-recurrent activities), increasing their sunk costs until the point they will be too proud to accept they made a terrible mistake. And finally, fabricating themselves an epic narrative where they will brag about the complexity of what they have crafted while opportunistically neglecting all the tears and blood spilled in the process. There’s a specific term for this: pyrrhic victory.