Pros: Quarterly company parties (if you're into that sort of thing), the technology stack is pretty solid, the office location is great, and the offices are attractive.

Cons: Non-mandatory gatherings are secretly mandatory. Management is a group of friends who will ignore numerous complaints about any single person within the group. The internal "restructure" was just for show. Everyone is hiding fear of being let go. Their junior developer turnover rate is enormous. The vacation package is a joke. The company politics are unavoidable and ridiculous. Everyone in management is two-faced.

Pros: A cool place to get a head start in your career.

Cons: Management can be too confusing.

Pros: Good pay Good benefits Impressive offices (pre-COVID, of course) Engineering "platform teams" dedicated to specific technologies are subject matter experts at your disposal. They're massively understaffed, though, so you usually just hear from this guy named Toby (not making this up). Stock prices shot up at the start of the pandemic, making the RSU grants far more appealing. Lots of modern tech (Kafka, Python3, Docker, Kubernetes, ELK stack, timeseries databases, etc.) Large data sets present interesting problems that you don't get at every company. People seem blissfully unaware of or apathetic towards the underlying technical problems the company has, so the overall atmosphere is pretty positive. Beer on tap in the office. Lots of communication from the high-level execs. Not astonishingly good communication, but certainly better than a lot of companies. An incredibly low bar for what the company considers technical excellence means it's pretty easy to coast under the radar and do the bare minimum required.

Cons: * Horrendous technical documentation is evenly distributed across READMEs, Confluence articles, Google Drive, and Slack threads. Most of it is incomplete. * No architectural oversight, at least in my org. Middle managers actively oppose seeking insight from the architectural teams. * "Ship fast and break things" and "don't let perfect be the enemy of done" are the common excuses used to dismiss any kind of proper automated testing or software design discussion which extends longer than 15 minutes. * Lots of the engineering division consists of level 1 engineers who just graduated from the Labs program and have no prior experience besides a coding bootcamp. They're at a huge disadvantage and generally aren't given the guidance they need to be successful because the senior members of the team are constantly putting out fires. * Absolutely nothing works. The VPN fails to connect; the laptop you're issued locks up because of Wayfair's proprietary update system; Kibana constantly times out; static analysis tools bundled with auto-generated application templates from language-specific platform teams have conflicting rulesets which cause builds to arbitrarily fail until you disable them; image publishing services used by CI systems constantly timeout and force you to run a rebuild (really fun when you're in a pinch); every application relies on undocumented APIs and hidden conditional logic that you have to force yourself to remember if you're ever going to get anything done; error logs are usually unhelpful and it's generally best to just debug an application; proprietary libraries are often buggy and annoying to use; upstream services arbitrarily send you bad data, etc. * Nobody owns some of the core, critical infrastructure. Database tables and internal applications are dumping grounds for dozens of teams who don't communicate, document what they add, or take ownership of cleanup and refactoring. There are tables with hundreds of columns, and nobody knows what they are, should contain, or are duplicates of. * There's a MASSIVE problem with a copy-paste mentality at Wayfair. This is almost always favored over a properly designed solution because it "unblocks" initiatives of other teams sooner and enables "parallelization" in development. * Heavy-handed managerial practices are used to force engineers to ship unmaintainable garbage so Jira issues can be closed, managers and their favorites can get promotions, and the rest of the team can be left to deal with the mess. * Nobody in software seems to think beyond the current quarter when it comes to their designs. There's a fanatical "solve for your current problem" mindset that I'd normally consider healthy but is taken to an extreme here and ends up being a massive liability whenever the inevitable scaling happens. Trying to go against the grain with this will get you in trouble. * Project owners understandably don't trust anything the engineering teams say.