Almost every company spends over 95% of its time doing what every other company does. And it spends less than 5% of its time on things that are unique to the company.
That makes no sense.
Ideally it should be flipped. Your company should spend the vast majority of its time focusing on things that are unique to your company.
Crazy thought experiment: Imagine a new type of company that decided to only do what it was really good at and essentially outsourced everything else.
Recruiting as a service?
Pretty much every company spends lots of time recruiting. Recruiting is a massive time suck. Yes, some companies do it better than others … but there are a set of best practices that all the good companies eventually master.
What if your company could completely outsource its recruiting (including interviewing) to another service? I know this seems completely crazy … but imagine if that could happen. Imagine the time savings.
Office space as a service?
Companies think a lot about their office space. That’s starting to change with flex space (like WeWork, Knotel, etc). but it is still really important. What if you never thought about it? What if you did not worry about the conference rooms, the food, the music, etc. What if you outsourced all those decisions (including how much space, where it is, and how much it costs) to an API? or an outsource service?
Seems completely crazy … but imagine the time savings.
Marketing as a service?
Marketing is another area that every company has to do. Some companies are really good at marketing. Some are pretty mediocre. All could be a lot better. What if no one at your company did marketing?
Finance as a service?
Finance too. Couldn’t finance be a shared service across tons of companies? Does every company really need to stand up a finance department? Unless you are a bank, finance is likely not core to what you do.
Remember compute?
While all this seems crazy … not that long ago most companies were managing their own compute. I was at LiveRamp — we owned metal and actually had to deal with servers. It was a LOT of work. Now LiveRamp is in the cloud (they use Google Compute) and a lot more time can be spent focusing on the business. Thousands of other companies went through this exact same process.
When we started SafeGraph (in 2016) we, of course, decided to have compute-as-a-service (we use Amazon Web Services). Pretty much every company started after 2013 has made the same decision.
Part of the reason we don’t outsource recruiting, marketing, finance, office space, HR, etc. is that there are no great services that exist to do it. That could change in the future. Certainly there are a ton of software tools that help (and where a companies can get massive leverage) … but employees in the company still need a lot of knowledge in the area to put the tools together (and take the time to select the tools).
If you start stripping out everything that is not unique to your company, you’re left with just a few people who make the unique parts of the company. And then add a few people who need to explain its unique benefits to the market.
Imagine your 100 person company going to 6 people. Imagine your 1000 person company going to 20. How much faster could you move?
Very interesting read Auren. Thanks for sharing.
1. I think that for services that can potentially be 100% outsourced, it works well when these services are not a sum-zero game. Hiring for example would be a challenge because you are completing with other companies for the BEST talent. It’s hard to trust someone else do everything for you. Compute, on the other hand, is not a sum-zero game service. I am not competing with anyone else when I need resources in AWS. I can just get more and more with the same quality.
2. Your model is actually put in action nowadays in new, special tech organizations like Team8. There is a common “platform” providing hiring, space, bizDev, marketing and finance services, and a set of startups (each with its own founders, CEO, CTO, etc.) who use this joint platform and only invest in what is specific to their problem domain. I think that the mutual trust between the startups and Team8 is what makes this structure work. I expect that we will start to see more of these initiatives for other domains beyond cyber security.
Thanks for your information.
I like your point. In fact the economy is working that way under the idea of division of labor. Not only do companies use AWS for compute, they also use Google, Facebook for marketing, use LinkedIn for recruiting, WeWork for office space as you mentioned. There are other tools/service for technology infrastructure, for finance, etc.
However, at the end of the day you still need to have human beings in a company to customize and use these tools. When the company is small, you need less people to use these tools. When the company grows out of the startup mode and reaches certain point in terms of size, the challenges are more unique that you need more customized solution for further improvement, and there won’t be another organization which you can outsource the problem to.
Auren, you make an interesting point. In fact,
I often cite a similar perspective to highlight the value and role of innovation in the smartphone industry.
If one were to review the ‘bill of materials’ for several such devices, they would likely overlap 90%+, given the common suppliers for building blocks like compute/ comms and storage HW, and mobile software stacks.
Likewise, if one reviewed the ‘bill of development effort’ invested by manufacturers to deliver their devices, that too would likely overlap to a significant extent, given the highly similar, fungible developer talent available to most players in this industry. So, the key differentiation (or lack thereof) appears to be rooted in that 5-10% of non overlapping effort, which I liken to the starter for a delicious bowl of yogurt or an amazing loaf of bread; and this is where innovation resides. The value of this differential ‘innovation effort’, driven not solely by the quantum of such effort but also its timing and discipline of pursuit, is reflected in the huge disparity of market shares and ASPs among the plethora of smartphone makers globally, led by Apple and Android.
Clearly, anytime a player increases their innovation effort relative to the total development effort, they have the opportunity to grow market differentiation/share.
This works when you are outsourcing to a 100% machine service (AWS example). Machines treat all problems equally. Outside of that, you’ll fall into the principle/agent problem.
The thought experiment is good. And a direction every company should be moving in.
Pingback: The New Status Game for Companies: Fewer Employees - Summation by Auren Hoffman