Dropbox

Dropbox is remote only, and is staying there for a while. There is zero risk of a return-to-office order in the next few years.

Dropbox has two health benefits from their glory days still hanging on that are worth a callout:

• Fertility coverage. This is self-insured, but covers IVF and other fertility procedures. I knew a few people who took advantage of this.
• Industry-leading coverage of trans procedures. They self-insure, but their program covers more procedures than the usual HMO (Anthem BC/Kaiser) riders typically exclude as purely 'cosmetic.' I know several people who took advantage, and there are few companies that rise to this level.

These benefits are already established, and relatively cheap for the company to maintain. They are likely to survive future austerity measures, and are top quality.

Dropbox is undergoing a state transition from rocket-launched unicorn to trading on past glory to find its new path to growth. You can see this transition play out in investor-call transcripts going back to 2023, as language has shifted from promoting growth in Dropbox's file sync and share (FSS) business, which brought Dropbox the brand recognition it currently has, to language promoting growth hopes in their AI/ML bet known as Dash. Dropbox led the market on FSS, in fact proving that there was a lot of money to be made in syncing files. The AI/ML-enabled workplace is a bet everyone else is making at the same time, so the odds are stacked against Dropbox finding its new path to growth.

Dropbox knows this, so it is funding its Dash development to get to market fast with a quality product. Both the April 2023 and October 2024 staffing reductions came with reorgs that focused more headcount on Dash development and go-to-market. This is classic "we're going back to startup mode" thinking, but Dropbox already has a multi-billion dollar revenue stream in its FSS product that comes with reliability and relevance expectations. The October 2024 reduction brought staffing below the threshold needed to keep the FSS product running without burning out engineers; reliability will still be there, but the scale of hidden firefighting is going to be significantly larger.

This is the big risk when choosing to work for Dropbox: If you're not working on Dash, your job is at risk. Everyone working there is tied to how Dash performs in the marketplace. If an AI winter falls and Dash doesn't bring Dropbox's growth past the Rule of 40, they will have to make up the growth deficit through profitability.

Like many companies over the last four years, Dropbox has introduced several plays from the Amazon school of leadership. Most important to workers is the stack-ranking that goes on as part of the annual performance review cycle. PeopleOps is quick to point out that this isn't a true stack rank; they're not rating all 98 people in a division from 1-98 and firing everyone below 90. The Dropbox system uses a 1-5 rating system, which is functionally --, -, 0, +, and ++. Most people will get a 0 for meets expectations. However, people in the -- and - buckets have gotten terminated or put on a Performance Improvement Plan (PIP) after both the 2023 and 2024 cycles.

Also, there are quotas put out by PeopleOps for what percentage of rankings need to be in each bucket. While PeopleOps makes it clear that you can trade + and ++ rankings to get people out of - and -- rankings, no Director does this. To make sure the buckets are enforced, bucket-filling is done at the Director, Sr. Director, and VP levels; meaning any coalition of Director-level people taking the option to reward fewer people to save people they don't want to yeet will be broken up at the Sr. Director level. Directors and above absolutely do keep track of how many + and ++ rankings their group got relative to others. The buckets are enforced due to how the system operates, not policy.

While it may not technically be a stack-rank, it's a bucket-rank that operates close enough to count for workers.

Related to the bucket-rank system are the wide shifts in strategy that land twice a year. Dropbox follows agile, data-driven management, meaning that they look at sales and other numbers to determine how well various market strategies are performing and then make changes in response. The effect of this is that any given internal policy statement has about a 9-month half-life. This affects the performance review cycle when you've been working on a project for 9 months and it gets canceled, leaving you with very little realized impact to put in your annual review, thus putting you at risk of earning a PIP. Even policies that were central to annual planning (typically running August through October) can get pared back or terminated mid-year.

Regarding commitment to diversity, Dropbox is strictly average; the 2023 affirmative action case at SCOTUS pruned any 'exceptional' aspirations for most US employers. Their trans-related benefits are a relic of earlier times that is cheap to keep, and I'm glad they are being kept. However, the trans and disabled communities have taken disproportionate hits in both the 2023 and 2024 reductions. This has less to do with transness and more to do with how many more trans folk are AuADHD and how AuADHD doesn't overlap well with expectations of consistent high-level performance.

The Employee Resource Groups are a source of community for workers, but the ERG process is fully owned by PeopleOps. Each ERG gets a heritage month and a budget for events and swag. ERG heads are also the go-to people for VPs and C-levels to consult with regarding policies and language usage. ERGs are also expected to contribute content for recruiting.

Finally, Dropbox is actively engineering its workforce to reduce labor costs. They're beginning to shift whole business lines to Poland from the US where an engineer costs 30% what they do in the US; the HelloSign acquisition being one of a few. They have already spent years selectively hiring in lower-cost US metros. AI/ML staff can come from anywhere, but most US positions will be offered outside of Tier 1 (NYC/SF). There is anecdotal evidence that RSU grant sizes are falling relative to earlier periods, and they already pared back the parental leave policy to give non-gestational parents less time off.

Dropbox listens to the market, not workers.