Groupon

Free Caltrain pass.

Free lunch in SF and PA.

The biggest cons is the company culture, and the fact that it's not a tech company.

Very, very poor engineering culture. One of the best examples I can think of is the way the company provisions VMs. There is this internal software used for people to set up VMs. It was implemented by using Git as a storage (??!??). So, for even a small change for any team in the company, for any VM in any environment, there is this centralized Git repository used as a storage. In order to make a change, you have to, in the end, commit there. So, can you imagine when all those hundreds of developers needing to make even the smallest change on any of the VMs start working? It is literally impossible to commit.

It sometimes takes hours because, in between you pulling, and rebasing, and pushing, there have been multiple commits. If you do it at 3 o'clock in the night when nobody is around, it's smooth. The fact that the company, which has been around for so long, didn't figure out a better way to do this tells a story. And what is worse, nobody sees this as a problem worth solving. Why? Because in their mind, it doesn't bring money (they only see direct money, revenue lift in that quarter).

It's very product-driven and marketing-driven. Hacks are preferred to stable solutions if they are quicker to implement. The credo is: "Yes, we are aware it is wrong, but do it anyway because it's faster to market that way."

There is barely any recognition for infrastructure work. No, if your code change, or your app doesn't make money, it's useless. The people will be moved to something else.

The super interesting thing is there is this belief system that software written like that, in SOA architecture, can be left to run staff-less. For instance, a bunch of interns write a software that does something. It shows some initial promise (money-wise), so they invest a bit more time. If it cannot be made to make more money, the management will move the people back into the 'pool' to wait for another task. While the software written by people that didn't really know what they were doing will be left there to run. Even more interesting, the whole project like that (shady software, no people) will be lumped over to some other team.

So, in meetings, you will often have multiple managers introducing themselves as "I am such-and-such, I am a manager of this app, this other app, this third app, this fourth app, and other things." And to be even more interesting, everyone in the meeting will have a similar bunch of applications on their plate, without any people to work on them.

Just the burden of having all these apps that you don't care about and don't own be on your pager duty is mind-blowing.

The amount of pages you can get sometimes at night is amazing. But when you offer the management to fix it, they tell you that they are not a priority right now. Sure, they are not; you don't get awakened at night for it.

Recently, it has gotten significantly worse. While Wall Street sees this as positive movement (cost reduction), the upper management has decided that headcount is not given to almost anyone anymore. People won't be replaced. There are literally whole teams that are leaving. Great talent too. Some people leave to Google, some to other big companies. Management rarely even talks to them why they are leaving.

I cannot even begin to imagine what kind of strategy that could be. Let the talent leave, the software will take care of itself. If not, lump it on some poor team with an overly ambitious manager.

My team, for example, lost two people last December, both very senior. They didn't get replaced. And as we were crying that we were understaffed, the management decided to not only not give us replacements, they decided to give us two apps that were written by two teams that were staffed with 10 people combined.

So, you lose two, you get new software worth for 10.

It's literally unbearable.

"Do more with less" taken to the extreme. I am also leaving now, after four years. I have to admit I learned a lot in the process, about how big companies work. I learned a lot as an engineer too. I got to live in two beautiful countries thanks to Groupon. But what is enough is enough.

I would seriously not recommend working for Groupon to anyone who is an engineer right now. Wait a little bit, either find something else, or wait six months and check again, maybe things have changed. It's literally super toxic now.

But once it gets this downhill, I really don't believe it will recover.

Even if you are just starting, it's better to go somewhere else and learn proper engineering from the get-go.

If you, by any chance, are interviewing at Groupon and you get to ask questions, ask them what kind of tool is used for provisioning VMs, ask them about the storage. If it is still what I describe, stop the interview and go somewhere else.

There exist other things, equally if not more important, than the sheer quarterly lift.