One of the (I thought) most useless courses I took in B-School was called Organizational Design - it seemed mostly to be about whether to put functional hierarchy or geographical hierarchy at the top level of the company. At some point during my startup experience I realized that OD is actually important, but the key insight is that it isn't a static thing - a single chart that you make and that's it. Instead it's dynamic, changing as the company grows, and as the people within the company grow.
How you manage the hierarchy of a growing company is crucial to managing people. It's nice to think that at a startup everyone is egoless, and will do what needs to get done, but its not true in the long run. People have careers and goals, and they will need to see their contributions reflected by the company contributing back to their career and goals.
1. When you give titles to people, give them based on how you see peoples' places in a company being in the future, not what they've done in the past. It's incredibly common for people to be really spectacular at the earlier stages of a startup and get put in a position of authority where they start causing problems when the company scales.
2. I don't believe in people having a "level of incompetence". More common is when you give someone a promotion based on their performance in a non-management/leadership job and discover that although they are good at doing who the jobs of the people they're managing, they aren't good managers.
3. Be really careful with giving people titles that imply some kind of authority over their peers. Even the smallest bit of authority can make otherwise pleasant people turn evil.
Titles matter in a lot of companies because if you want to find a new job, the title is used to 'level' you. Do you want to be a VP at <well-known-tech-company>? You probably need to have a strong title at your previous role too.
On a tangential topic, somebody once pointed out to me that reading the exec bio pages at tech companies makes for interesting reading. If you skip past the famous founders, you'd find a very large number of people from a background which goes something like this
- Go to B-school (preferably Harvard)
- Join one of the consulting firms. McKinsey is great on your resume
- Leave McKinsey and join into a management role at a some large company
- Switch jobs every 2 years - each title being larger than the previous one, wind up as SVP at famous startup/large tech company.
If you're someone who wants such a career trajectory, a title 'deflation' is going to hurt you.
Having said all that, if I do a startup, I wouldn't want to hire someone who cares so much about their title. It just doesn't...feel right.
Who would not want to hire people who don't care about money, titles, stock options and works 24/7? There are very few people like that.
Even people who are joining Facebook at a step down are not doing this because they are above materialistic desires. Having Facebook on your resume today is like having Google on your resume 5 years back.
We should stop kidding ourselves about trying to find the truly selfless smart person.
The only solution that I have seen working to solve the "Peter Principle" problem, is re-org.
Once in a while, parts of the org chart, get re-organized. People moved vertically and horizontally. Carefully moving the "newly incompetent" to a horizontal position that suits their competence level (ie remove the responsibilities that they handle badly and increase those they handle well).
I realize that anecdotes make for very poor data, but I have never seen a re-org work. I've survived dozens of reorgs at my own employer and at client firms, and I have never, ever seen a successful one. The best you can hope for is to see incompetent management pushed into do-nothing positions, such that the damage they do is contained.
More often, the shuffle creates new fiefdoms, new dotted line bosses, and not one substantive improvement in function. (It's usually too painful to make the necessary horizontal changes to push "peter principle" types into harmless positions. These people are the re-organizers' friends, vassals, or patrons. What inhuman manager could bear to throw his friends under the bus to no immediate advantage for himself?)
For example: when you have turf war between two managers, a re-org creates the opportunity for turf war between three or more: the old empire-builder who was ejected, the new empire-builder brought in, and the manager who was already encroaching on the old fief.
I find this whole vocabulary for describing your office environment (empire-builder, fiefdom, turf war, etc) to be incredibly negative. Thinking or your peers as bumbling warlords has got to make your time at the office a pain.
That said, I have not worked at a large company before, so I might think the same way in your shoes.
The problem is that in large companies, that is the way it works. Politics are an unfortunate fact of life just due to human nature.
In my opinion, the biggest reason why those upper management folks in big companies make those mistakes when they reorg is because they don't have a view of what happens in the trenches. The details are critical when determining whether a specific service can be provided by a specific group and be tracked by a specific system.
They don't get that a system's order database has a data structure that can't accommodate a new category of product types and a new way of doing work. They crunch the numbers, but the inefficiency lost from poor processes due to lack of understanding at the ground level makes those numbers meaningless in the end.
The thing is, getting promoted is a full-time job in its own right. You can't do real work and get promoted, you won't have time, because you will be competing for a very limited number of positions with people who do make getting promoted their full-time jobs.
At my present employer, I have been very explicit: I do not want promotion, I do not want to be management (there is no tech career track). Which has the interesting side-effect of making me basically unmanageable by their rules; there's neither carrot nor stick that can be used on me. So I get left alone to get on with real work. I've never been more productive!
I agree with this and would like to add that a culture of frequent re-orgs just brings out a different form of the Peter Principle. Instead of regular people getting promoted to their level of incompetence, the organization now optimizes for politicians getting promoted to their level of incompetence. I've seen my current and former employers lose extremely good engineers because they felt they'd been screwed over by bosses who were optimizing their careers to take advantage of the next re-org.
You can't really prevent politics from rearing its ugly head in a large organization, but the best you can do is make sure that you hire extremely competent, honest and straightforward employees and treat them absolutely fairly. One of my bosses had a really great rule for making decisions on new hires. Besides being technically competent, committed and stuff like that, you need to ask yourself whether you'd like to have dinner with this person every single week. If the answer is no, you probably shouldn't be hiring that person.
How you manage the hierarchy of a growing company is crucial to managing people. It's nice to think that at a startup everyone is egoless, and will do what needs to get done, but its not true in the long run. People have careers and goals, and they will need to see their contributions reflected by the company contributing back to their career and goals.