(Like this article? Read more Wednesday Wisdom! No time to read? No worries! This article is also available as a podcast).
Note: This Wednesday Wisdom mini-series is the article version of a talk I gave at Google Zürich in November 2024. If you are inside Google, you should be able to find the video. It is a bit of a lengthy exposé, so for easy digestion it has been cut in three parts. Last week (in part 2), I discussed how companies change because hiring more people means hiring different people and these different people have different behaviors, different motivations, and different communication styles. This week: The conclusion: How companies change because, as they grow, the nature of the work changes and that changes people’s behaviors.
We are almost at the end of this Wednesday Wisdom mini-series, but I promise you: I saved the juiciest bits for this third and final episode.
However, before we get going, here’s a trigger warning: This episode is not for the timid and I am going to say things that might offend some people. If you are easily upset or unsure of your own performance, you might want to skip this episode.
Another warning: This week’s article might make me seem to come off a tad arrogant, but, to make it up to all of you, for next week I am planning an article on how I became humble by working with people who are much better than I am (preliminary title: “Pressure makes diamonds”), which I hope will restore some balance to the Wednesday Wisdom universe.
Ok, so here we are: The company is growing and went through various changes because external pressures demanded it and because it hired many people who are fundamentally different from the first batch. We now get to the third vector for change: The work changes. This prompts people to start competing and it inevitably leads to level inflation and the hiring bar dropping.
When the company starts out, it is a greenfield: There is no existing infrastructure and no technical debt yet, meaning there are no constraints other than how fast you can build stuff. In this phase, the company needs people who can homestead: Pioneers, trappers, and explorers. While the company is building new products and infrastructure, there is an abundance of work, almost all of it interesting. Consequently, there is little or no competition between employees as there is enough great work to go around for everyone. It is also relatively easy to get promoted on the back of that great work (assuming you do a decent job, of course).
As the hitherto empty land gets built up, the work changes in very fundamental ways. Whereas in the “wild cat” phase, the focus is on building large new products and infrastructure (at high cost with no or low margins), during the “cash cow” phase it is all about keeping the lights on and reaping profits (low cost but with high margins) .
The terminology in the previous paragraphs comes from the Growth-share Matrix, which is a model that can be used to figure out where your company or business unit is at in its lifecycle.
In this phase, there is less need for new things and more and more of the work comes down to tending to the stuff that has already been built and that is now in need of maintenance and improvement. There might be the occasional project for a new skyscraper, but mostly the work becomes about mowing the lawn, keeping the elevator running, repairing the plumbing, and dealing with the small house fires that happen every now and then. The company has pretty much everything it needs. They might need more of it, but in the interest of efficiency, the company typically invests in mechanisms for easy replication of the work that has already been done rather than building things from scratch.
In a somewhat tongue-in-cheek way, employees who are supporting the cash cows work in the “JSON mines”, as their efforts are more about configuring existing systems than it is about building new ones. Even the work of building new systems or components is fundamentally different because there is by now a solid infrastructure to build upon.
These changes in the work drive changes in the employee base. Without putting too fine a point on it: The employees you need to hack around in the JSON mines are of a different calibre than the ones that built the mines from scratch. This leads to the hiring bar dropping because really, not every new hire needs to be a genius anymore. This impacts hiring.
In part 2 of this series I wrote about changes to hiring caused by an increased appetite for riskier hires. That increased risk tolerance impacts the hiring of candidates that are in the nebulous grey zone surrounding the hiring bar, but, as the company continues to grow, we see an actual dropping of the bar, as candidates are being hired that really had no hope of getting through the door in the early days of the company.
As the company grows, they need to hire in order to keep up and, unfortunately, in every market you operate in, there are only so many really great people you can hire. Eventually you will be faced with a “this candidate or nobody” situation and all the organizational incentives are pointing towards “this candidate”. Recruiters have targets and they don’t care who they hire as long as it counts against their targets. Managers with hiring targets hope there are corners of the JSON mine that this candidate can be applied to and of course there is always the possibility that the candidate can grow in their role.
Additionally, in growing companies, engineers play a much smaller role in the hiring process, with power shifting to recruiters and managers. That is not optimal because recruiters and managers have fundamentally different incentives than the engineers. Additionally, it is often extremely hard for them to see why you might want to hire one candidate but not another, especially if these candidates look approximately the same to a non-engineer or to a no-longer-handson former engineer
Once the hiring bar starts dropping a bit, the situation is exacerbated by the fact that A’s hire A’s, whereas B’s unfortunately hire C’s. Once the B’s get in, you are effed. Of course I know that all of the people reading this article are A’s (as evidenced by their excellent choice of literature to waste their valuable time with), so it does not apply to this readership, but I know that the readers have seen this happen. I know I have.
However, the company is still hiring people with a degree and N years of experience as a level L, so the dropping of the hiring bar also starts the inevitable level inflation.
However, hiring is not the only factors that plays into level inflation. Apart from the in itself correct realization that the company can make do with people that are not as accomplished as the ones they hired before, competition for candidates can also play a role. Until recently, there was a battle for scarce talent at all levels and so people negotiated higher levels on hiring. I vividly remember talking to candidates with ~6-8 years of experience who “demanded” being hired as a staff engineer. Companies that feel a need to hire will accede to such demands, but really the only thing that changes is the meaning of “staff”. I used to joke that I expected to become senior principal by the simple passing of time.
To be honest, at this point in time, the battle for talent is no longer a big factor in level inflation. For the lower levels of experience and skill there are an abundance of candidates available and we now see credible evidence of the opposite effect, namely downleveling on hire.
Of course, hiring new-grads as senior software engineers was never a good idea to begin with…
Another force that is driving level inflation is the promotion process. Once all of these candidates are through the door, they want to get promoted because, if you don’t get promoted, what are you doing? (The answer is probably: You are just being very good at your job) But in a world where vapid titles seem to have a direct link to people’s sense of confidence and self-importance, everyone is always chasing the next promotion.
Here, the fact that the company is no longer a greenfield gets in the way. In tech, promotion usually means that you need to have been performing at that next level for a while. For this, there need to be opportunities to showcase these amazing next level skills. And what is in less supply as the company transforms from a greenfield to a post-industrial wasteland of JSON mines? Right: Opportunities! In the company’s earlier days, there used to be an abundance of opportunities because there was so much still to do. But as the holes get filled in, these opportunities become rarer and rarer, especially at the senior→principal levels.
This sad fact has three consequences.
First: Because of the buildup of pressure from below, the promotion barrier drops. People still want to get promoted though so they keep trying, whether the work really allows it or not. So, promotion packages are still written. They might not be as amazing as before, but they are invariably much better written and definitely more beautifully typeset. But, people will continue to be promoted, leading to more level inflation.
Second: There is more and more competition between employees for the few good opportunities that still exist. This leads to more competitive behavior and teams becoming territorial so as not to “give away” valuable opportunities to get promoted. This also leads to people storming into new product areas in the company because that is where careers are still being made.
Three: Projects that are technically suboptimal get commissioned anyway because they provide a pathway to promotion. Things that should not be built as a separate component are, because it allows for an end-to-end development process with a launch to production at the end. A ready-made promotion opportunity!
These processes, all a result of growth and success, work with the changes outlined in parts 1 and 2 to make the company a more normal company. Probably still an amazing company to work for, most likely still a good place to learn lots of things, surely a company where you get to work with some amazing people, but it is not as amazing anymore as it used to be.
This fact leads to considerable angst, especially in the employees who have been there longer. Pretty much all of my friends who have seen this happen to, or who are seeing it happen in, “their” company are upset about it. And I get it: It is upsetting if something you love and to which you contributed, regresses in the direction of the mean. There are fewer (or no) interesting projects to work on, not all of your colleagues are amazing anymore, some douchebag director or VP who started ten years after you is calling the shots or maybe they are just outright muppets. It is all very annoying.
That said, everyone reading this should still count themselves totally lucky because, most likely, you are in the top 0.001% of all people who ever lived in terms of wealth, health, safety, and general welfare. So, maybe the quality of the free sushi isn’t as amazing anymore and some of your colleagues use Visual Studio Code instead of vim or emacs (as Bram and Richard ordained it). But, as the Germans so nicely put it: “Das war einmal”: At some time in the past, it was like this. Just be happy you got to experience that first hand.
And for the people who missed this boat? Don’t worry, your time will come. It always does, as long as you are willing to take some risks.
Share this post