Going Full Stack
Schemes of classification and taxonomy help a person comprehend and analyze what’s going on in the world around him or her. That is why I consistently come back to Chris Dixon’s post from 2014 — Full stack startups. Dixon writes:
“The new approach is to build a complete, end-to-end product or service that bypasses existing companies.”
The thread and belief in full stack companies ties into Andreessen Horowitz’s mantra, “software is eating the world.” I like the post because it does a wonderful job of concisely explaining how some of the most interesting startups of this generation will be built. There’s a wide swath of old-school industries such as insurance, construction, and wholesale distribution that software has yet to fully penetrate. The full stack phraseology helps one grasp how an upstart can succeed in those spaces.
More current examples to tack on to Dixon’s include Oscar, Hims, Flexport, and Bowery Farming. I want to note that building an end-to-end disruptor is not always the solution, as some large players in industries have built incredibly deep regulatory and distribution moats. In those cases, you could build a vertical software solution that could end up turning into a really big business. There’s been a good amount written about vertical marketplaces and vertical SaaS, but I’ll put a plug in for my friend Doug Nelson’s piece on vertical payments as something to think about:
But how do you know whether you should build a disruptor or an enabler? Here are some specific characteristics and features of full stack companies if you choose to go that route.
The industry is structurally flawed
Value chains in the space you’re looking at are so bloated that giant companies can’t radically alter their corporate structure without collapsing. A prime example of this would be in the office furniture industry where a network of vendors and dealers (middlemen) passes on additional expenses to the end customers.
Since there is so much structural degradation in the market, layering on a marketplace or SaaS tool does not provide truly differentiated value to users.
Bet on net margins in the long run
Nothing beats gross margins for software companies. But with full stack businesses, you’re betting on the viability of your net margins. The COGS are going to scare some traditional venture investors, but the idea is that you can scale operations and customer acquisition to be significantly lighter than rival legacy players (who are your real comps) over a longer period of time.
Obviously, you’re going to need healthy gross margins in order to make the equation work.
Software isn’t just for your end users
A lot of the software you’ll end up writing will be for internal teams to streamline functions like industry-specific operations, customer success, and sales. Build a product org that values this and can move fast thanks to close proximity to their users. This characteristic gets even more interesting as companies like Retool and Forest are popping up, which allow organizations to plug in internal tools as micro-services.
Financial engineering and asset management expertise matters
Full stack businesses are often more logistically and financially complex than pure software startups, so they require early hires and talent that can fill needs in those arenas. This differs from a traditional venture-backed company, where most early hires are engineers or product people. For example, all the following concerns Bill Gurley has about Bird’s unit economics get wrapped up in the world of asset management.
If you're thinking about building a full stack startup, please get in touch! I can be reached at aashaysanghvi[at]gmail.com
