169 comments

[ 3.2 ms ] story [ 215 ms ] thread
This is great. I will say though that, unsurprisingly, Puerto Rico isn't part of the dataset. Sites like this could really help make the point to people that the US is bigger than the 50 states.

[edit] Alaska and Hawaii aren't on here either so it should have been 'bigger than the continuous 48 states'.

Including other countries like Canada, UK, France, and Germany would certainly be nice.
Helpful tip, the word you are looking for is "contiguous" when describing the US states all connected together. Can also use the Lower 48
I have no idea why I typed that but I will accept the public shame!
No shame intended! Just helping in case you didn't know!
I think the granularity of this map is far too coarse grained. Forks, WA and Seattle are in the same bucket. Same with San Jose, sharing the same bucket as Ukiah. The "Greater Portland Area" stretches all the way to the southern border of Oregon.

That said, there's still some surprising results. I would have expected NYC to be on par with Western Washington and the Bay Area, but it's significantly less, ~190K vs ~260K.

Agreed, the "Greater Boston Area" doesn't really reflect the reality, you are mixing Boston Metro (HCOL) with many others in Mass, Vermont, and New Hampshire that can't be compared.
> The "Greater Portland Area" stretches all the way to the southern border of Oregon.

There may be an off-the-grid remote developer in Steens Mountain somewhere, but there aren't any employers there.

Lots of techies out in Bend though
Not nearly as many as there are in portland by a factor of about 25
This is accurate, as a techie in Bend. But the map hives off Deschutes county all by itself, and at the same time lumps in Baker City with the "Greater Portland Area". That needs some work.
Given how many areas are marked as not having enough data, I'm going to guess that the dataset is pretty small, which is why some of the areas had to cover large spaces.
Including Pike County, PA as part of the New York City area is kind of wild, too, from someone who grew up there. If you're making $190K and living in Pike County, you're living like a Sultan of a country with a palace made entirely of gold.
Not all goods are nontradeable. Flights and lodging at a destination cost the same to everyone as do iPhones etc.
As an immigrant from the 3rd world -> Canada, I found a surprising number of things fit this bill.

Some things have a global market and everyone is paying ~ the same price everywhere.

- Meat, to an extent

- Any oil-derived product

- Electronics

- Software

- IP

- Cars

- Clothes

Of course there are always local taxes, regulations, and logistical considerations that skew the price this way or that way by 10-30%, but these markets can be pretty efficient.

While I do agree with you, that doesn't really impact salary all too much. Allow me to explain.

Taxes, Housing, Transportation, and Insurance is what eats up most of your expenses anyway. Housing varies wildly across the world. Many parts of the US now cost > $3,000 /mo to rent a median home. Most of those homes are probably selling for ~$500k putting your mortgage at current interest rates in the same ballpark of $3k /mo. For somoene making $120k that is 1/3 of their income. For someone making the national average (around $70k) that is 1/2 their total income. Just to mention, those areas (Seattle, San Fran, NYC) where you see $250k-300k salaries, those people are paying much higher than these figures. Probably $5-6k in rent or buying modest homes that just happen to cost $1.25M-1.5M

In the US you also can pay $1,000 (or more) /mo for health insurance. Most other parts of the world don't have this expense.

As of 2024, the average price of a new car is now $47,000 in the USA. Now not everyone is buying a new car but the used car market swells based on this figure.

Then you pay 15-30% in taxes on average for most people here.

So You add all that up and probably 2/3-3/4 of your money is gone.

I agree with you that things like clothing, consumer goods, software, meat, and oil are largely comparable globally now, but these goods only account for probably 10-20% of a person's monthly expenses.

Not to mention retirement. Because all these costs are so high, it means I need to put a larger percent of my paycheck to retirement so I can survive a few years of not working before I die. When basics like healthcare and housing are as high as I outlined, it means more money needs to go to retirement, which is money that can't be spent on these basic goods that you mention.

So yes, these goods you call out are in fact comparable around the world. But they only account for a relatively small portion of someone's expenses. The outstanding expenses are the most variable (housing, transportation, insurance, taxes).

Before I moved from my very cheap birth country to the entirely-not-cheap area known as Toronto, I thought along these lines and was concerned that my standard of living would be the same or lower. However, surprisingly, I found it to go up a lot.

- Real Estate did take up a larger % of my budget, but it was nicer

- The globalized products are only a small % of the budget but they're relatively much cheaper so I can have more/better/both

Of course this doesn't mean expensive cities are perfect or even good. I later moved from Toronto to Alberta and it's night-and-day better. Of course the lifestyle is different and some might not like that.

There's a big qualitative angle so you can't really compare these things on a spreadsheet, but more after-tax dollars is almost always better.

This is true but the number of people who commute from Pike County (and the whole tri-state/NY-NJ-PA corner) down to the crowded parts of North Jersey and even NYC is insane. I live just across the river from Pike County and experience the traffic though our small town.
The pay is for the office location, not home location. And as a tech corp employee in NYC I have multiple coworkers commuting 2h each way 2-3 times a week (including from the Pike County area). Rural PA has become pretty popular for the 4br SFH for sub-$1M crowd
This is a great point, and something we plan to address. We currently use Nielsen's DMA (Designated Market Area) mappings within the US to separate out regional areas which was used for TV / media market surveys. We happen to use DMA categories for our regional pages on Levels.fyi which is why it was easiest to start with since we already had this data captured. The features can sometimes be a bit off and seem like they're grouped very far and wide (you'll notice there's a bit of Denver within Nevada and its just a vestige of how it used to be categorized), but it still provides a bit of a broader level grouping than something like zip code. We've also been considering using Combined Statistical Areas using population instead, but the benefit with DMAs is that it offers full coverage of the entire US whereas some major tech hubs are still missing from CSAs if relying solely on population.

We're planning to create some of our own regional definitions and borders using our own submissions and that should offer some more tighter bounds. This was just a v1, and I think its already resonating with folks.

GeoJSON data for the map borders: https://github.com/PublicaMundi/MappingAPI/blob/master/data/...

Nielsen DMA regions: https://blocks.roadtolarissa.com/simzou/6459889

Is this also using the levels.fyi salary data?

If that data is submitted by individuals to a particular company, is it possible to see a lot more detailed heatmap, perhaps down to each address of each company?

> We've also been considering using Combined Statistical Areas using population instead, but the benefit with DMAs is that it offers full coverage of the entire US whereas some major tech hubs are still missing from CSAs if relying solely on population.

I think just using the 387 MSAs [1] instead of the 181 CSAs would get you far enough to cover all the major tech hubs.

[1] https://en.wikipedia.org/wiki/Metropolitan_statistical_area

You think that's shocking?

This is just pre-tax TC.

When you take into account taxes and cost of living, comparing NYC to a place like Austin makes NYC TCs look positively stingy.

This thing feels weird. In Southeast Michigan the popup is for "Ann Arbor Area", yet it stretches from VERY rural areas to Ann Arbor (a wealthy college town), across Detroit, across where all auto companies are, etc.

This makes it feel very not-representative nor accurate for the area as a whole.

EDIT: Ohh... clicking further, now I see. This is just an ad for a "salary negotiation" company. No thanks.

It may or may not be an ad, but I’ve found levels.fyi to be a valuable source of salary information, much more than a “salary negotiation company”. They’re far more accurate and detailed than, say, Glassdoor or Indeed.
It's not bad to acknowledge that Levels sells salary negotiation, so they are more inclined to skew perception of salaries as higher, so more people look and say "Oh, I'm underpaid!"

While Glassdoor does not sell this service and thus has no incentive to overhype current salary distributions.

I do think it's much more accurate though

Are these numbers base salary or TC?
There's a breakdown of the compensation categories with each area.
(comment deleted)
Is it possible to exclude FAANGs and other large corporations? Mixing smaller companies and soul-draining leviathans all in the same pot does not produce meaningful results.
Smaller companies are also soul-draining in many cases. Why to exclude only faangs? Let's exclude banks, insurance etc
You're getting caught up on "soul-draining," but FAANG's and adjacent are generally acknowledged to have the highest TC. It would be useful to know what the statistics are with these outlier companies removed.
The "outliers" are the companies paying these insultingly low salaries for technology development. That's why there's so much low-quality software in the world.

FAANG (and a few FAANG-adjacent) companies are the only ones paying close to decent wages, and even they've been making frankly egregious cuts to their protein-bar budgets lately.

Let's not sit around manufacturing skewed datasets that give people the wrong idea about what software engineers should get paid.

While I agree that FAANG data should be present/available, I will say that the only reason those companies are able to pay such amounts are due to their outsized valuations and the market shares they have captured. Vast majority of companies do not bring in per-employee revenues on par with the FAANGs, so it's not realistic to set one's definition of "decent wages" at those numbers and expect everyone else to pay them. Lots of high-flying startups made a play for fast growth and paid similarly high comp, with the end result of laying huge numbers of people off when the market demanded accountability.

TL;DR: I love being an engineer in the Bay Area, but we truly are a bubble.

Not everyone who works on faang codebase receive a decent wage. Don't forget an army of cheap external contractors sold to faang by bodyshops.
The "outliers" employ a quarter million software engineers. You are asking for reality to be warped and censored to suit your weird vibes.
No one called for reality to be censored - simply a toggle that shows the summary stats with and without outliers included. This is the kind of thing folks learn in Stats 101.
How are you defining "outlier"?

You can obviously exclude the top 25 percentile of salaries, since the app shows breakdowns by percentage, but I doubt levels.fyi has accurate data on the number of employees of each company in a particular region.

Why to exclude only faangs?

Because they skew the numbers upwards. Most of the market doesn't pay their wages.

(comment deleted)
(comment deleted)
(comment deleted)
Even in Silicon Valley, not all large employers pay FAANG rates. In general, older, “enterprisey” companies (think HP, IBM, Cisco, and Intel, to name a few) pay considerably less than FAANG and its peers, which tend to be younger and sell consumer-facing products and services.
My job at a smaller company was much more soul sucking than my jobs at larger companies
The default view is for median TC which is resistant to outliers.
I'm surprised NYC is 190k vs Bay Area at 263k and Seattle at 240k. Maybe there's just more non-tech industry software jobs pulling down the median?
I believe it's the outsized proportion of FAANMG employees. Bay Area should be obvious but Seattle has gobs of $250k comp Amazon and Microsoft employees, much more than outposts in NYC.
I wonder how skewed the underlying data is, too. Perhaps they're somehow overcounting people at the higher levels and undercounting people lower on the totem pole. The idea of $250K being a median across all levels -anywhere- is kind of astonishing if true. Yes, we all know a few people making $400K at Facebook who have a vacation home in Aspen and drive two Ferraris (they always tell HN how common it is), but is it really that many to drive the median so high?
A person who only makes $400k TC at Meta won't even have a first home in the Bay Area, much less another one in Aspen.
Plenty of people with $400k TC at Meta have Bay Area homes. $400k TC doesn’t mean “I just got the job where’s my $2m house.” But over a career (Facebook is 20 years old) it does.
The median home-owning household in the bay area doesnt make anywhere close to $400k/yr. Its expensive, but not that expensive.
I don't think that's a good way to view it. How about the median home buying household? In my town the median sale price this year has been $1.6 million, considering interest rates and property taxes that amounts to $10k/mo housing costs, which is over half the take-home pay of a $400k gross income in California. This is not even to mention the fact that nobody will lend you a mortgage based on the equity portion of your TC, they will usually only count the salary portion and discount the equity portion.
Napkin math:

$400k * 0.6 = $240k post tax

$240k - $36k rent = $204k

$204k - $50k annual expenses = $154k

So if you just work that job for 3-4 years you'll easily have a heavy down payment for a house, assuming no promo, no raises.

If you work the job for 7-8 years you can buy a house (more like a townhouse) all in cash. Not going to be the best house on the market, but it's doable.

This is all not considering investing in the stock market, raises, being frugal with your expenses, living with roommates to save up a bit, etc. there's lots of ways to parlay more money than most other people will ever make annually into a home

Your belief in stationary housing prices is charming.
Read the last part of my comment. I left out many natural ways one's earnings and investments and savings could increase over time also. The math still maths
It’s a selection bias for people that are already interested in compensation and willing to divulge it. If you’re on the site, then it perfectly fits you. But it wouldn’t be representative of all software engineers from a census perspective.
This is the answer. Most people don't know about levels.fyi.
I know many SWEs at boring non-FAANG, non-unicorn companies that make around the medians shown here so it seems roughly representative, anecdotally. The competitiveness of the market for good engineers has forced every company to at least pretend to try to compete on compensation. It didn't used to be this way but it has really compressed wages upward because you simply won't be able to hire anyone vaguely qualified otherwise.
No idea how it is in the US, but if you have a look at Zürich, Switzerland, something like 20% of the data points are from Google, maybe 50% from FAANG. Clearly very heavily biased towards high-paying multinational companies.
Surprised to see $155k as the median for Montgomery-Selma. It's a very low cost of living area, which in the tech industry is justification to pay low salaries regardless of the low desirability of the area. Going to guess it's mostly defense related developer jobs associated with Gunter.
$155k in that part of Alabama gets you a real nice house, big lot, kids in private school, etc.
With apologies: a choropleth map, not a heatmap. And the granularity is unfortunately quite too low, but I appreciate that sometimes your geodataset is limiting. But with that pedantry out of the way: it would be awesome to be able to normalize based on cost of living.

Not that it’s an excuse: I do find it kind of odd that I do the same job as another remote engineer and yet I’m paid a fraction? But to be fair I don’t have student loans and paid off my house in a few years, so cost of living, cost of education, etc. can reveal practical opportunity even if it enshrouds any definition of fairness.

Thanks for the pedantry re. choropleth vs heatmap.

I learned something today because of it.

Not really "across the US" because it's only the lower 48. AK, HI, US territories don't appear to be included. We're used to being forgotten so it's not a huge deal, but I figured I'd take this small opportunity to bitch about it :-D

If you're curious, SWE pay in AK is pretty low. I'd guess median in the 80k.

If I say I have been to McDonald's all across the country, that does not mean I have been to EVERY single McDonald's restaurant. Just various ones across the country
> If I say I have been to McDonald's all across the country, that does not mean I have been to EVERY single McDonald's restaurant. Just various ones across the country

Sure, but I would argue that you're speaking imprecisely then and using a cliche or phrase that isn't technically accurate at best, and is actually misleading at worst. And when requested for clarity, you should say, "well almost all." Either that or we have different definitions of what the word "all" means[1].

If someone said "I have been to all the states in the US" would you expect that they have been to AK and HI?

[1]: The MW definition matches my understanding: https://www.merriam-webster.com/dictionary/all

Yes. If someone said they had been to all the states I’d expect they had been to Alaska and Hawaii. On the other hand McDs all across the country I’d interpret as having been to a lot of McDs in different states with a wide geographical distribution.
You're only hurting yourself by looking up a dictionary for these sorts of things. Words have varying contextual meaning that can range from the dictionary definition to it's polar opposite. Just take the best possible interpretation and move on. Almost nobody cares about your "technically accurate".
Note the “all” in the example binds to “across”, not “McD’s”.

“All McD’s” - quite precise; literally all the McD’s.

“McD’s all across the country” - a much looser group, implies various McD’s spanning roughly coast to coast (hence “all across the country”), but not necessarily including the westernmost and easternmost McD’s - just a decent distribution of west-to-east, perhaps with no giant gaps (like missing the whole Midwest).

The "Not Enough Data" should really be a different color, instead it is very close to the lowest bracket which makes it confusing
(comment deleted)
Impressive Missoula, MT has a median higher than many metropolitan areas like Austin. One of the factors of the house market explosion in western MT.

In a related note, I was checking for tech meetups (at meetup.com) in Missoula and Bozeman and except for Montana programmers, there's no much there. There are a few slack communities but nothing specific for technologies or other groups.

I noticed Missoula as an outlier, too. Anyone have a good explanation?

My completely uninformed guess is that a bunch of highly-paid engineers moved there during the pandemic for some reason I don't understand, rather than anything inherent to the tech jobs market in Missoula. If so, why Missoula (vs., say, Jackson Hole)? And if not, is there another plausible explanation?

Gas and oil industry
Missoula has been low-profile fashionable with the tech crowd for a long time. It is basically a slightly smaller Boulder.
(comment deleted)
I'm surprised as well, google did not tell me much there about the high wage either, still have no clue
This $263k median in the Bay Area is making me sad.
On the other hand, most LinkedIn jobs published last week (e.g. full time/hybrid in SF) seem to be much closer to 180K-200K.
That’s base compensation. They aren’t reporting equity, which is usually at least 1/3 to 1/2 total compensation.
I think it's certainly incorrect (having known lots of people on both sides of that number, there are far far more below). Another comment thread suggested that startup equity is being taken at face value, which might justify the number but is totally ridiculous
(comment deleted)
levels.fyi has been of good use for the industry at providing tools to navigate the incosistencies with leveling across companies. Somewhat similar to what Leetcode did as well (not saying Im happy with the standard).

There's a lot more refinement that's needed for levels.fyi data:

1. Data goes stale pretty quickly. Salaries are on a downtrend now and many averages don't reflect it yet.

2. Data is overreported in the few popular reigons and companies. Bay area/FAANGetc

3. Values are inflated with stocks that aren't public companies.

4. Lots of companies are following weird vesting schedule now and that calculation isn't the simplified 4 year average of stock value.

The website makes money by selling salary negotiation, so they have a bit of incentive to inflated salaries a bit
Ann Arbor gets the call out over Detroit. That's rough... and the pay is higher in Traverse City? I'm so confused by these results.
This is a nice market where you'll be hired if you're better, so it's also an average quality of engineer map.
I found it interesting that Madison, WI is on par with Chicago, and pay is better than Milwaukee. Digging in a little it's because global tech companies have offices in Madison (also Epic, but even though I was in Toronto for 20 years and thus out of date, I never knew Epic to be a high paying company), and Milwaukee seems to be regionals (Uline & Kohls - and retail rarely pays top dollar for tech talent).
I’ve commonly heard people referring to Epic as a company that pays quite well for the area. They also have a reputation for being kinda crappy to work for, so I never pursued a position there. All the other jobs and offers I’ve had in the Madison area pay quite poorly compared to remote jobs with companies based in the Bay Area.
Are we talking Epic health software or Epic MegaGames?
Epic Systems, the purveyor of medical records management software.
I'm getting the sense there's a heavy sampling bias towards larger companies. I looked at my company which is on 1 metro area and it isn't super accurate. The titles they list don't even match up with what we call them.
We're working on improving our taxonomy right now and would love some more detailed feedback. Do you mind emailing me at <my hn username> @levels.fyi?
Can levels.fyi stop showing non-liquid pre-IPO startup equity as part of total comp please?

You'd be surprised at how difficult it is to get liquidity for that stuff, often there are limits to the amount you can liquidate, some can't sell on private marketplace, some can only sell every once in a blue moon event, etc. This is all without mentioning that the "valuation" itself is typically pretty speculative.

Levels.fyi is treating this equity the same as public company RSUs, which is not the same at all.

How do you think it should be treated? I think at the individual granular data point level adding a tag or note about the equity not being immediately liquid is a good start. But I don't think it'd be a good idea to weigh the stock differently since that can depend on so many things. For example SpaceX and some other private companies do offer regular liquidity and I would consider their equity close to liquid.

Appreciate the feedback though, and definitely agree we can work on how we display the data and make it more clear.

Add another dropdown so we can color code by Base salary only, Stock only, etc.
Yeah that make sense, will work on adding for this heatmap
Maybe 3 categories?

1. Salary (straightforward, on regular schedule, and you'll get it)

2. Bonuses and RSUs (various vesting rules, and ways you can never see it)

3. Startup stock and (worse) stock options (probably worthless, vesting rules, and you might need an advisor to make sure you don't exercise and come out with a big negative)

Is levels.fyi biased towards high pay jobs?
Yes. And extremely biased towards tech jobs. So take their data with a grain of salt.

If you use Levels.fyi data in a salary negotiation anywhere outside of SF or Seattle, you will probably get laughed at.

I used to be in the 90th percentile in my area 2 years ago. Then AIv2 (rebranded as genAI, LLM) pumped with VC money via low interest pushed my TC to 70-75th percentile.

I could start chasing the $ again, but at this point I’m nearly financially independent and can almost just say fuck it.

This was really cool to look at!

But I still don't understand how non-tech people afford to live in SF. Wouldn't they be priced out?

People who bought houses at any point in the last decades are house-rich and have a low tax basis thanks to Prop 13. Otherwise, there's an active, ongoing exodus of non-tech people, as well as increasing cost of living and longer commutes for people who stay.
Someone needs to serve the burgers though. Are you saying those people commute two hours each way?
Oh, definitely. One of the worst commutes in Northern California is the Altamont Pass, which Interstate 580 traverses. As early as 4:30am there is a long line of cars filled with commuters who live in more affordable Central Valley places like Stockton, Modesto, and Tracy, commuting to either Silicon Valley or San Francisco. These are generally not FAANG software engineers; they do all sorts of other work, such as cleaning offices, preparing food, stocking aisles, teaching students, building Teslas (Tesla has a shuttle that goes from Modesto to its Fremont factory), policing Bay Area communities, fighting fires, and a lot more.
The megacorps also have buses that run to distant towns. People have informal roommate situations midweek and ride the bus home for the WFH+weekend.
Yes - check out Brentwood, Gilroy, etc where middle class Bay Area residents are commuting from. The expense and shortage of non-tech labor is also pushing things like ghost kitchens, coffee robots, self-checkout kiosks, etc.
This actually sounds a lot like Vancouver in Canada.
There are ways people without FAANG-level salaries are able to afford San Francisco:

1. Purchasing when market prices were lower, or inheriting a home. San Francisco has been expensive for decades, but it didn’t always require FAANG-level salaries to afford purchasing a home there.

2. Living in a rent-controlled unit and avoiding evictions (e.g., if the landlord wants to sell, move in, or redevelop the unit).

3. Qualifying for government-subsidized housing, in the form of either Section 8 (voucher-based housing assistance), below market rate rentals, or below market rate properties for purchase. Many Bay Area municipalities have a local housing authority that provides more information about local subsidized housing programs.

4. Shared living situations, whether it’s with family, friends, or strangers, helps reduce housing costs at the expense of needing to share space with others. I know many people in the Bay Area who wouldn’t be able to afford to live here without some type of shared accommodations.

5. Some employers subsidize housing expenses. For example, some universities in the Bay Area offer housing assistance to tenure-track faculty members, ranging from down payment assistance to zero-interest mortgages. There are some universities that sell homes to faculty and staff at below-market prices, with the stipulation that those properties get sold to other faculty and staff once they are put up for sale.

The increasing homeless population says yes.
So dumb to pay according to where you live.

Fun fact: VCs own lots of corp and residential real estate. They want to drive people to live in their areas, pay more but houses cost more and it’s just a big con

It's not just VCs. Do you think your manager wants their home value to go down? They just paid $2.5 million for that 70 year old box of aluminum wiring, lead paint and asbestos and they sure as hell want to sell it for more when they retire.
People say this as if there’s some conspiracy.

But I think the uncoordinated explanation is just that annd top executives often operate in a world where informal quick access to a bunch of powerful people is important. And in that universe it simply makes sense to have people be in the same place.

And beyond that, in their eyes… why wouldn’t you want to live in these wonderful cities? Why wouldn’t you want to be in the office and work through things? These are people who self select through working being their life so they barely conceptualize alternatives

In the same way you can’t conceptualize why someone would want to be in the office, they can’t conceptualize why you wouldn’t.

What's juvenile is not knowing how prices get set, which leads to an oversimplified world view, which leads to thinking that location-based pay is "dumb". To price something, you take what it costs to make the thing, throw that out the window, and make up a number based on vibes. If you're lucky, that number is bigger than what it costs to make it, and the business grows. but if you've made crap, or any of a million other reasons why people aren't buying your stuff, then you have to sell it for less than it costs to make it, and the business might be in trouble.