Lean Startups with Few People Make no Sense

I keep hearing people brag about how their startups aren’t going to need any humans to run them.  Why are people bragging about this?  If you truly have an idea that doesn’t require people (everything can be done with machines), then you have no sustainable advantage.  Other people can buy the same machines you have and you won’t have anything unique to offer, anymore.  If you have an idea that does require people, and not just any people, but smart people who are going to keep innovating for you, then you can create a business that can sustain itself in the face of competition.

Let’s take a more concrete example.  Admittedly, this is a gross over-simplification, but it’s still a good example.  Which is a better way to grow a startup?

1) Pursue an idea that takes 10 racks of servers and 1 sysadmin to run them?


2) Pursue an idea that takes 8 racks of servers and 3 sysadmins to run them?

I would argue that it’s a no-brainer.  If you’re starting a business, and for simplicity, you’re buying the servers, then a 48U rack of servers probably has the same price tag ($100K-$200K) as a sysadmin for a year.  If you have a business that takes 3 sysadmins, and you get good ones, they will be developing technologies, systems, and experience that are worth something in year 2 of the business.  If you start a business that has 2 extra racks of servers, how much better will that make your company look in year 2, as those assets depreciate?  You may have been able to serve a few more customers, but are you really safe from someone else coming along and blowing you away?  I doubt it.

4 thoughts on “Lean Startups with Few People Make no Sense

  1. While I don’t think it makes sense to brag about not needing humans, I think automation can be important while scaling.

    I think comparing 1 to 3 people might be the wrong metric – what do you think of it with the ratios of servers:users and productive employees:users vs just constant numbers?

    However, I really do agree that humans (users or employees) provide appreciating value while the value of computer software can only depreciate as competition increases.

  2. I share in your conviction in the sense that a believe in people and what they have to offer really makes start-ups grow. Machines and computers most definitely need people to give them instructions to operate. Machines don’t bring ideas to fruition but rather smart people.

  3. However,Sophia do you think a time will come when computers and machines will dominate in the business world? Robots are doing what people do and i have fears that we are moving toward the “artificial business world”. So maybe your assertion may lose its value in future

  4. To the commenters on this posting, from its author, I believe you guys are right on. My main point was to show that we irrationally ignore the relationship of human costs to IT equipment costs. A rack of servers costs as much as a man-year. Will you really gain as much from one man-year’s worth of an additional rack of servers as from one man-year’s worth of personal development?

    Of course, for your first rack, you will, because without it the people will have nothing to work on. At some point, though, as the value of each subsequent rack diminishes, the value of another person will be greater. I used a very artificial example to try to show that “having just 1 sysadmin to cover 100,000 users” or something like that, while it may be possible (and reflect a very efficient operation), may not be a good investment.

    To Bayuo’s point, if robots can do the job, it has essentially been comoditized. There will always be a role for humans in developing whatever business model, process, or software is new and innovative. Even in a world where AIs are designing software themselves, people will still have to design better software-developing-AIs. Whatever part of the work has been comoditized is not where you want your startup to be… it’s something that could more efficiently be done at large scale by a large company.

Leave a Reply

Your email address will not be published.