Why Do the Biggest Companies Keep Getting Bigger? It’s How They Spend on Tech
wsj.com
wsj.com
People have limited energy to store and rank objects, so they settle on one or two per field/subject, and all future energy goes into maintaining that categorization.
An example would be "Hulk Hogan", even though I am not interested in wrestling, I do know that name. If I am similar to others, such as journalists, this would explain why that name is getting bigger and bigger: people just add on energy on the current most popular choice.
[1] https://www.amazon.com/Scale-Universal-Innovation-Sustainabi...
After hitting 201-500 employees, it becomes almost impossible to avoid politics, bureaucracy, wasted time communicating, oversight, management, etc.
Based on a graph that showed peak productivity with a team size of 6 (https://www.getflow.com/blog/optimal-team-size-workplace-pro...), one layer of management results in 43 employees (1 CEO + 6 managers/executives + 6 * 6 individual contributors). Two layers results in 259 employees (1 CEO + 6 executives + 6 * 6 managers + 6 * 6 * 6 individual contributors). More than two layers adds more inefficiency/employee than is taken away and such a downward trend continues.
If a fictional 4.6 team members is used, those numbers become 26.76 and 124.096.
Example article: http://www.erpsoftwareblog.com/2012/11/why-does-microsoft-hq...
(not sure if still accurate in 2017, but I think it is)
The amount of resources that go in to SAP, while large, are a rounding error compared to the resources spent on building a compute platform.
I’m only half-joking. The companies are big to begin with because they did something better than their competitors. Also they can attract and afford to hire the best in key roles.
Shrug.
But yes, it does seem rather obvious.
However, I do think it's illuminating in that this business practice is really only relevant to big companies. It's more effective the bigger you are. If you can have a team of a dozen programmers, testers, and sysops engineers employed full-time to make 10,000 other people more effective, your company will be more efficient than mine. A small business will be much better off spending a couple hours creating a website on Squarespace and calling it good, because the ROI just is not there. Software scales, and this makes big companies bigger.
This seems less obvious the more I think about it. You are saying proprietary home grown software scales. You can just as well say software provided by external sources scales, which is why you don't need a big company with a lot of employees. If you can argue it both ways, then the real question is what tips the balance? Maybe we just have a bimodal ecosystem where the big get bigger and the small get more numerous.
If you're using a vendor solution w/o source rights, and you need a new feature to pursue a new opportunity... you can't.
Best case, the vendor is persuaded they can make money from it, develops it, and offers it to all their other customers. Middle case, they charge you an arm and a leg for custom functionality (more than it actually costs to develop), and now you're off the main tree. Worst case, they ignore you or refuse.
So the primary benefit to owning your own software development is the ability to pivot to new opportunities.
E.g. You think AWS would exist if Amazon had outsourced their server management software?
From a tech worker's perspective, this seems super obvious, because that's how they generate their value in the market.
But this actually isn't "News at 11" material, because in the majority of non-engineer dense industries, companies purchase tech solutions when they become available and aware of them on the market rather than build out what they need when they need it.
Those are tremendously different modes of execution. When Goldman Sachs ramped up their spending on developers to produce proprietary solutions, for instance, it was a newsworthy development.
It's very different to say "We build things that make X product" or "We build things that provide you with service Y". Then to say, "We make product X" or "We provide you service Y".
We clean floors...
We build floor cleaning technology that we then use to clean your floor more quickly and for less money.
It's very different. It means you'll need to really invest in tech, hire engineers and scientists, put money in R&D, understand how technology is built, the cost of it, how to iterate on it, and leverage it. While at the same time integrating your business in actively using it, driving the tech based on the exact concrete needs of the service or product you target, etc.
This is hard to not half ass, to really commit to this mentality, and have everything designed around it is hard.
It's also higher risk, higher reward. If you fail to build useful tech that does end up delivering a better product or service to more people for less money, you might lose out and have ended just spending all your money on trying. If you succeed though, your growth will be massive, and your tech an unprecedented competitive advantage, you'll get bigger and bigger and eat away at your competition. And then an article will say, why do big companies get bigger, because they invest in tech.
There's a lot more to it. These just get ignored by the media and studies the most. The corporate media is doing similar stuff. No surprise they'd downplay it. Author covered talent snatching but misses most of the rest. Even the too hard to clone requirement comes from IP law (trade secrets). And a lot gets cloned anyway.