28 November 2022
The big difference between Twitter and Tesla
Tesla make great cars. I have had a Model 3 and recently upgraded to a Model Y. I even tried to upgrade to a Model X, but after waiting almost 2 years gave up on my reservation. I even bought Full Self-Driving (and then lost it when I upgraded from a 3 to a Y, infuriatingly.)
The 3 (Distinct) Benefits of Tesla Automobiles
People conflate three overlapping but independent benefits of Tesla automobiles:
They have an electric drivetrain
They have advanced autopilot and vision technology
They have delightful software operating systems.
Put it this way:
Tesla could make a gasoline automobile and if it used their Tesla software (the screens and such) it’d be a great car
Tesla could have totally botched their software but made an electric car, and it’d be a great car.
Hardware + Software Problems vs. Software Problems
Of these three benefits of Tesla automobiles, two are distinctly challenging problems — it’s really hard to build reliable electric drivetrains at massive scale, and it’s really really hard to build advanced full self-driving technologies.
The third, having delightful software systems, is not in and of itself a tremendously challenging problem — and if you watch Doug Demuro car reviews, you’ll increasingly see cars with equally as good, fast and robust software operating systems as Tesla.
There is a clear delineation between these problems — two of them are very hard problems that must be brute forced with relentless hard work for years and when the problems are cracked they represent massive trillion dollar markets, and the final one (delightful software) is a very different kind of hard problem, one of culture and thoughtful consistent execution, rather than solving a singular hard problem.
The problem with Twitter
Twitter does not have a singular hard problem — the problems with Twitter are that of nuance. If you could wave a magic wand and “fix” all of the bot problems on Twitter, would that fix Twitter? not in the slightest.
The problems with Twitter spread the entire gamut of product design and software engineering problems. You need ensure new users are onboarded effectively and see interesting content as quickly as possible. You must ensure those users quickly download mobile apps and stick around. You must ensure as they travel and are in low mobile data environments, everything works, etc etc.
Tesla Operating System .... Delightful?
This is a very different proposition to how Tesla has created their automobile operating system.
I would also argue that the Tesla operating system is delightful but not actually that stellar. For example 12 months ago Tesla introduced a major update that was widely panned:
I bought the car because tesla was known for having the best UI in the business. Now for zero good reason everything is buried in a menu. Seat heater? Menu. Tire pressure? Menu. Glovebox? Yes. Dash cam? Menu.
Now of course Tesla shipped updates and improved this design over the subsequent months, but with consumer software like Twitter you do not get that luxury. It isn’t like Tesla owners are going to instantly sell their six figure automobiles, whereas Twitter users could very easily churn after a handful of annoyances.
Finally — whilst the Tesla operating system is delightful, the company does not ship improvements at a terribly noteworthy pace compared to software companies (but do compared to other automobile companies) — I would say that Tesla ships a noteworthy improvement to the Tesla operating system once every four months, at best.
I struggle to reconcile all this with how people can look at Tesla and think that Twitter is similar.
And in fact, we’re already seeing this from the disastrous roll out of the Twitter verified program.
What Twitter needs is a steady patient hand to build tremendous software experiences for new and existing users alike. If Twitter thrashes around, changing interfaces left right and center, they’ll alienate their user base.
The big difference between Twitter and Tesla
Unlike Tesla and SpaceX, Twitter does not have a singular discrete problem to solve:
Tesla: Build a software powered electronic automobile
SpaceX: Build a reusable rocket
These two problems, while vague, are quite actionable for large teams of engineers. Whereas the Twitter product mandate is far murkier — because Twitter is obviously a software company rather than a 'hard problem we're going to solve with engineering' company.
Despite this, is Elon Musk the right person for Twitter? Probably! Despite SpaceX and Tesla having even more in common than most people appreciate — since both solve terrifyingly difficult multi-disciplinary problems — you've got to believe that if nothing else, considering Elon is an avid user of the Twitter product that he'll ensure the company moves quickly and thoughtfully.
I imagine that after the first few romantic months of him being involved at Twitter, he'll work to ensure Twitter has an equivalent leadership team as SpaceX and Twitter, and at that point you'd hope that Twitter will be on the right track. It is, after all, hard to bet against someone as accomplished as Elon Musk.