Hiring developers in San Francisco

What the published data says about the San Francisco-Oakland-Fremont, CA technical labour market, and what it means when you are deciding whether to hire here, hire elsewhere, or build the team differently.

The local market in numbers

The Bureau of Labor Statistics counts 69,030 software developers employed in the San Francisco-Oakland-Fremont, CA metropolitan area, at a median annual wage of $186,640. The national median for the same occupation is $135,980, which puts San Francisco about 37% above the country as a whole. Among the 377 US metropolitan areas where this occupation is separately published, San Francisco ranks 2 by median wage.

Those two numbers answer different questions and are often confused. The wage figure tells you what a seat costs. The employment figure tells you how likely you are to fill it. A market can be expensive and deep, expensive and thin, cheap and deep, or cheap and thin, and the hiring strategy that follows is different in each case. Reading only the wage column is the most common way a hiring plan goes wrong before it starts.

What the wage distribution shows

In San Francisco the 10th percentile sits at $128,090 and the 90th at $273,400, a spread of about 2.1 times. The middle half of the market, between the 25th and 75th percentiles, runs from $163,500 to $219,670. That band is the number worth carrying into a budget conversation, because it describes where most actual offers land rather than where the single midpoint sits.

That spread is narrower than the national spread of about 2.6 times. A wide local spread usually means the metro contains several different markets under one label: a well-paid cluster attached to a few large employers, and a much broader base of internal IT and agency work that is priced quite differently. When you benchmark against a local median without knowing which of those clusters your role sits in, the benchmark is close to meaningless.

One more caution on reading these figures. They are wages paid by employers to people on a payroll in this metro area. They are not contractor rates, they do not include equity, bonus or benefits, and they say nothing about what the same work costs when it is done by a team somewhere else. They are the right baseline for the question “what would this cost if we hired locally” and the wrong baseline for almost every other question.

Pay across every technical occupation

Software development is one of several technical occupations the Bureau publishes for this metro. The full set is worth reading because roles compete for the same people: a strong back-end developer in San Francisco can be recruited into a data or security title, and the pay in those adjacent columns is part of what you are bidding against.

Blank cells are where the Bureau does not publish an estimate for this metro, usually because the sample is too small to release.

Every one of these occupations is separately published for this metro, which is itself a signal: the Bureau suppresses estimates where the sample is too small, so a complete set indicates a technical labour market with real depth across specialisms rather than a single dominant employer.

Reading each technical occupation in San Francisco

The table above is a summary. Each row behind it describes a different hiring problem, and the gap between the local and the national figure is rarely the same from one row to the next. Where a metro is expensive in one occupation and ordinary in another, that difference is usually the clearest signal available about which employers are actually competing for people here.

Software Developers

The Bureau counts 69,030 people in this occupation in the San Francisco-Oakland-Fremont, CA area, at a median of $186,640 against a national median of $135,980, which is about 37% above the national figure. The middle half of the local market runs from $163,500 to $219,670. A location quotient of 2.68 means the work is considerably more concentrated here than the size of the local economy would predict, so sourcing should be straightforward and closing will be the harder half. The local band runs from $128,090 at the 10th percentile to $273,400 at the 90th, so the title alone sets the budget only very loosely.

Web Developers

The Bureau counts 1,420 people in this occupation in the San Francisco-Oakland-Fremont, CA area, at a median of $152,450 against a national median of $92,650, which is about 65% above the national figure. The middle half of the local market runs from $115,170 to $179,950. A location quotient of 1.33 means the work is more concentrated here than the size of the local economy would predict, so sourcing should be straightforward and closing will be the harder half. The local band runs from $80,190 at the 10th percentile to $216,060 at the 90th, so the title alone sets the budget only very loosely.

Web and Digital Interface Designers

The Bureau counts 6,630 people in this occupation in the San Francisco-Oakland-Fremont, CA area, at a median of $163,810 against a national median of $104,000, which is about 58% above the national figure. The middle half of the local market runs from $110,100 to $211,120. A location quotient of 3.83 means the work is considerably more concentrated here than the size of the local economy would predict, so sourcing should be straightforward and closing will be the harder half. The local band runs from $81,120 at the 10th percentile to $230,700 at the 90th, so the title alone sets the budget only very loosely.

Computer Programmers

The Bureau counts 2,440 people in this occupation in the San Francisco-Oakland-Fremont, CA area, at a median of $130,730 against a national median of $100,390, which is about 30% above the national figure. The middle half of the local market runs from $77,710 to $166,830. A location quotient of 1.73 means the work is more concentrated here than the size of the local economy would predict, so sourcing should be straightforward and closing will be the harder half. The local band runs from $74,110 at the 10th percentile to $174,770 at the 90th, so the title alone sets the budget only very loosely.

Software QA Analysts and Testers

The Bureau counts 5,880 people in this occupation in the San Francisco-Oakland-Fremont, CA area, at a median of $134,630 against a national median of $104,300, which is about 29% above the national figure. The middle half of the local market runs from $97,320 to $168,050. A location quotient of 2.06 means the work is considerably more concentrated here than the size of the local economy would predict, so sourcing should be straightforward and closing will be the harder half. The local band runs from $75,100 at the 10th percentile to $205,990 at the 90th, so the title alone sets the budget only very loosely.

Data Scientists

The Bureau counts 10,460 people in this occupation in the San Francisco-Oakland-Fremont, CA area, at a median of $170,110 against a national median of $120,230, which is about 41% above the national figure. The middle half of the local market runs from $130,430 to $212,970. A location quotient of 2.61 means the work is considerably more concentrated here than the size of the local economy would predict, so sourcing should be straightforward and closing will be the harder half. The local band runs from $99,170 at the 10th percentile to $272,430 at the 90th, so the title alone sets the budget only very loosely.

Information Security Analysts

The Bureau counts 3,730 people in this occupation in the San Francisco-Oakland-Fremont, CA area, at a median of $162,310 against a national median of $129,180, which is about 26% above the national figure. The middle half of the local market runs from $115,000 to $201,690. A location quotient of 1.28 means the work is more concentrated here than the size of the local economy would predict, so sourcing should be straightforward and closing will be the harder half. The local band runs from $61,860 at the 10th percentile to $246,070 at the 90th, so the title alone sets the budget only very loosely.

Computer Systems Analysts

The Bureau counts 11,180 people in this occupation in the San Francisco-Oakland-Fremont, CA area, at a median of $134,460 against a national median of $105,850, which is about 27% above the national figure. The middle half of the local market runs from $104,460 to $170,700. A location quotient of 1.41 means the work is more concentrated here than the size of the local economy would predict, so sourcing should be straightforward and closing will be the harder half. The local band runs from $91,480 at the 10th percentile to $209,150 at the 90th, so the title alone sets the budget only very loosely.

Database Architects

The Bureau counts 1,600 people in this occupation in the San Francisco-Oakland-Fremont, CA area, at a median of $175,860 against a national median of $139,500, which is about 26% above the national figure. The middle half of the local market runs from $163,160 to $212,170. A location quotient of 1.56 means the work is more concentrated here than the size of the local economy would predict, so sourcing should be straightforward and closing will be the harder half. The local band runs from $133,050 at the 10th percentile to $237,930 at the 90th, so the title alone sets the budget only very loosely.

Database Administrators

The Bureau counts 1,470 people in this occupation in the San Francisco-Oakland-Fremont, CA area, at a median of $127,760 against a national median of $104,620, which is about 22% above the national figure. The middle half of the local market runs from $92,810 to $163,940. A location quotient of 1.37 means the work is more concentrated here than the size of the local economy would predict, so sourcing should be straightforward and closing will be the harder half. The local band runs from $66,750 at the 10th percentile to $210,810 at the 90th, so the title alone sets the budget only very loosely.

Computer Network Architects

The Bureau counts 3,350 people in this occupation in the San Francisco-Oakland-Fremont, CA area, at a median of $166,760 against a national median of $134,050, which is about 24% above the national figure. The middle half of the local market runs from $128,690 to $211,740. A location quotient of 1.22 means the work is more concentrated here than the size of the local economy would predict, so sourcing should be straightforward and closing will be the harder half. The local band runs from $121,160 at the 10th percentile to $242,150 at the 90th, so the title alone sets the budget only very loosely.

Network and Computer Systems Administrators

The Bureau counts 3,910 people in this occupation in the San Francisco-Oakland-Fremont, CA area, at a median of $129,680 against a national median of $99,130, which is about 31% above the national figure. The middle half of the local market runs from $100,850 to $163,550. A location quotient of 0.81 means the work is less concentrated here than the size of the local economy would predict, so a local-only search in this specialism will be the slow part of the plan. The local band runs from $82,760 at the 10th percentile to $193,250 at the 90th, so the title alone sets the budget only very loosely.

Computer and Information Systems Managers

The Bureau counts 23,830 people in this occupation in the San Francisco-Oakland-Fremont, CA area, at a median of $232,890 against a national median of $175,140, which is about 33% above the national figure. The middle half of the local market runs from $203,840 to $313,220. A location quotient of 2.33 means the work is considerably more concentrated here than the size of the local economy would predict, so sourcing should be straightforward and closing will be the harder half. The local band runs from $155,660 at the 10th percentile to $335,770 at the 90th, so the title alone sets the budget only very loosely.

Computer Occupations, All Other

The Bureau counts 16,930 people in this occupation in the San Francisco-Oakland-Fremont, CA area, at a median of $161,700 against a national median of $116,580, which is about 39% above the national figure. The middle half of the local market runs from $102,340 to $206,460. A location quotient of 2.55 means the work is considerably more concentrated here than the size of the local economy would predict, so sourcing should be straightforward and closing will be the harder half. The local band runs from $63,340 at the 10th percentile to $255,390 at the 90th, so the title alone sets the budget only very loosely.

How concentrated the work is

The location quotient for software developers in San Francisco is 2.68. The measure compares the share of local employment in an occupation against the national share, so 1.00 means exactly what the size of the local economy would predict. At 2.68, this market has substantially more of this work than its size would suggest.

High concentration cuts both ways. It means candidates exist and that they are already employed, which shortens sourcing and lengthens closing. Expect competing offers, expect counter-offers, and expect notice periods to be honoured rather than bought out. Low concentration means the reverse: less competition for the people who are there, and fewer of them, so the binding constraint moves from persuasion to supply.

Expressed another way, there are about 29.1 software developers for every 1,000 jobs of any kind in this metro. That density figure is the one to quote when someone asks whether the local market can absorb a team of five or a team of fifty, because it scales with the size of the economy rather than against a national average.

What a local seat actually costs

A salary figure is not a budget. The number that belongs in a plan is the employer-side cost of keeping one person in one seat for a year, and it is reliably higher than the wage line that gets quoted. The arithmetic below uses the published software developers median for San Francisco and the statutory 2026 employer rates, so every figure in it can be checked.

Employer share of OASDI at 6.2% up to a wage base of $184,500, plus Medicare at 1.45% with no cap. Source: Social Security Administration, Contribution and Benefit Base, retrieved 2026-09-25.

That is the floor, not the total. On top of it sit federal and state unemployment insurance, workers' compensation, health cover, retirement contributions, equipment, software licences, office cost where there is an office, and the recruitment fee or internal recruiter time that brought the person in. None of those are in the wage figure and all of them are real.

There is one more line that almost never makes it into a comparison, and it is usually the largest. A seat that stays empty for three months has cost you a quarter of the year's output from that role. Whatever the wage, the delivery cost of the vacancy is the same, and in a market where a search genuinely takes that long it dominates the difference between one hiring option and another. When teams tell us a local hire and a remote team came out close on cost, the time to fill is nearly always the variable that was left out.

Working across time zones from San Francisco

If any part of the team is not in Pacific, the working day is a design decision rather than an accident. San Francisco runs on Pacific time. The question that matters is not how many hours of difference there are, but how many hours of deliberate overlap the team commits to and what is required to happen inside them.

The pattern that works is a fixed overlap window of two to four hours, published and treated as real. Inside it go the things that cost a day when they are done asynchronously: code review on anything non-trivial, decisions with more than one reasonable answer, and the conversation when somebody is stuck. Outside it goes focused work. Teams that skip the fixed window do not save the meeting time; they pay it back several times over in waiting.

Two practices decide whether this works. The first is writing decisions down where the other half of the team will find them, because a decision made in a corridor in San Francisco does not exist for anyone who was asleep. The second is that nobody should ever be blocked overnight on a permission, an environment or an access request; those are the failures that turn an eight-hour offset into a lost day, and they are entirely preventable.

How the San Francisco market has moved

One year of data describes a market. Three years starts to describe a direction. These are the published figures for software developers in the San Francisco-Oakland-Fremont, CA area across the last three releases.

Source: BLS Occupational Employment and Wage Statistics, metropolitan area files for May 2023, May 2024 and May 2025.

Over those two years the median wage moved up 8%, from $172,340 to $186,640. Employment moved down 18% over the same period, from 83,920 to 69,030. Pay rising while employment falls usually means the composition of the market is shifting rather than that everyone got a raise: junior and mid-level roles thinning out while senior roles hold. If you are hiring at the experienced end, expect the local band to be firmer than the headline suggests.

One caveat belongs with this table, and it is the Bureau's own. These estimates are produced from overlapping survey panels and the methodology is revised between releases, so they are not designed to be read as a clean time series. A two-point difference between years is noise. A move of the size seen here is worth noting as a direction rather than quoting as a growth rate.

Nearby markets worth comparing

Hiring decisions are rarely made between one city and the rest of the country. They are made between a handful of markets a team is already willing to consider. These are the metros closest to San Francisco on that basis, with the same figures side by side.

Percentages compare each metro against San Francisco, not against the national median.

Differences of a few percent between neighbouring metros are not worth planning around; they fall well inside the error on any single hire. Differences of twenty percent or more are worth planning around, and so is a large gap in employment, because that is the number that decides whether a shortlist exists at all.

The industries that shape hiring here

Wage tables describe the price of labour. They do not describe what the labour has been doing, and that is usually the part that decides whether a candidate fits. The anchors of the San Francisco economy are venture-backed SaaS, fintech, AI and machine learning and marketplaces.

Venture-backed SaaS

Developers who have come up through venture-backed SaaS in this market tend to arrive with domain knowledge that is genuinely hard to buy, and with working habits formed by that sector's constraints: its release cadence, its tolerance for risk, its review and approval culture, and the compliance obligations it carries. Both halves travel with the candidate. When the sector you are hiring into shares those constraints, that background is worth paying for. When it does not, the adjustment is real and it is worth asking in the interview how they would work differently.

Fintech

Developers who have come up through fintech in this market tend to arrive with domain knowledge that is genuinely hard to buy, and with working habits formed by that sector's constraints: its release cadence, its tolerance for risk, its review and approval culture, and the compliance obligations it carries. Both halves travel with the candidate. When the sector you are hiring into shares those constraints, that background is worth paying for. When it does not, the adjustment is real and it is worth asking in the interview how they would work differently.

AI and machine learning

Developers who have come up through AI and machine learning in this market tend to arrive with domain knowledge that is genuinely hard to buy, and with working habits formed by that sector's constraints: its release cadence, its tolerance for risk, its review and approval culture, and the compliance obligations it carries. Both halves travel with the candidate. When the sector you are hiring into shares those constraints, that background is worth paying for. When it does not, the adjustment is real and it is worth asking in the interview how they would work differently.

Marketplaces

Developers who have come up through marketplaces in this market tend to arrive with domain knowledge that is genuinely hard to buy, and with working habits formed by that sector's constraints: its release cadence, its tolerance for risk, its review and approval culture, and the compliance obligations it carries. Both halves travel with the candidate. When the sector you are hiring into shares those constraints, that background is worth paying for. When it does not, the adjustment is real and it is worth asking in the interview how they would work differently.

The same concentration explains where the competition for a candidate comes from. In a market anchored by a small number of large employers in one sector, you are usually bidding against those employers directly, and against their compensation structure rather than against another startup.

Where the talent comes from

The main academic feeders into this market are UC Berkeley, University of San Francisco and San Francisco State University. Pipelines matter most for junior hiring: graduate intake is where a local university genuinely changes the supply you can reach. At senior level, people are recruited out of other companies rather than out of programmes, and the relevant pipeline is the set of local employers large enough to have trained them.

This is worth being blunt about when a hiring plan leans on a university reputation. A strong local computer science department improves your odds of finding good juniors and has almost no effect on your odds of finding a senior engineer who has already run the kind of system you are building. Those are two different searches with two different sources, and conflating them is a common reason a plan that looked reasonable produces nobody.

Where hiring plans go wrong in this market

These are the failures we see most often in San Francisco specifically. They are not generic hiring advice; each one follows from something in the figures above.

Benchmarking against a national median

Local pay for this occupation runs about 37% above the national figure. A band built from national data will sit below the local 25th percentile, which does not produce hard negotiations, it produces silence. Candidates at this level read an under-market offer as a signal about the company and stop replying.

Running a slow process in a fast market

With the work this concentrated, anyone worth hiring is in more than one conversation. A two-week gap between final interview and offer is not diligence here, it is a withdrawal. Decide the panel, the decision-maker and the offer range before the first interview, not after the third.

Writing the brief around a technology instead of a problem

A specification that lists ten technologies and no description of the system filters for people who match keywords rather than people who can do the work. It is also the main reason shortlists in this market get rejected late, because the mismatch only becomes visible in a technical conversation.

Costing the seat and not the vacancy

The loaded-cost table above is the easy half. The expensive half is the months the role is open, and that cost is identical whichever option you eventually choose. Any comparison between hiring locally and building the team another way that leaves out time to fill has already reached the wrong answer.

Treating one metro as one market

The wage spread here is wide enough that San Francisco contains several distinct markets under a single label. Benchmarking a role against the metro median without knowing which of those clusters it belongs to gives a number that is precise and wrong.

What this means in practice

Pulling the figures above together, these are the practical consequences for a team hiring into San Francisco.

• Move quickly once you have a shortlist. In a market this concentrated, the candidates worth hiring are in more than one process, and a week of internal deliberation frequently costs the hire.
• Budget for a counter-offer. Expect the current employer to respond, and decide in advance what you will do about it rather than improvising under time pressure.
• Benchmark against the local band rather than the national median. An offer built from national figures will be uncompetitive here by a margin large enough that strong candidates will not counter, they will simply stop replying.
• Separate the two questions. Decide what the work genuinely requires to be done in this time zone or in this building, and treat everything else as a search that does not have to be local at all.

Hiring by technology in San Francisco

Published pay and market conditions for San Francisco-Oakland-Fremont, CA, read against the specific technology you are hiring for.

• Machine Learning developers in San Francisco
• AI Engineering developers in San Francisco
• Node.js developers in San Francisco
• Python developers in San Francisco
• Java developers in San Francisco
• .NET developers in San Francisco
• PHP developers in San Francisco
• Laravel developers in San Francisco
• Django developers in San Francisco
• Go developers in San Francisco
• Spring Boot developers in San Francisco
• Microsoft Azure developers in San Francisco
• Rust developers in San Francisco
• Ruby on Rails developers in San Francisco
• WordPress developers in San Francisco
• Shopify developers in San Francisco
• JavaScript developers in San Francisco
• GraphQL developers in San Francisco
• SQL developers in San Francisco
• Data Engineering developers in San Francisco
• Data Science developers in San Francisco
• MySQL developers in San Francisco
• WooCommerce developers in San Francisco
• Magento and Adobe Commerce developers in San Francisco
• Drupal developers in San Francisco
• QA Automation developers in San Francisco
• Salesforce developers in San Francisco
• React developers in San Francisco
• Next.js developers in San Francisco
• TypeScript developers in San Francisco
• Angular developers in San Francisco
• Vue.js developers in San Francisco
• React Native developers in San Francisco
• Flutter developers in San Francisco
• iOS and Swift developers in San Francisco
• Android and Kotlin developers in San Francisco
• AWS developers in San Francisco
• DevOps developers in San Francisco
• Kubernetes developers in San Francisco
• Terraform developers in San Francisco

Frequently asked questions

What does a software developer cost in San Francisco?

The Bureau of Labor Statistics puts the median annual wage for software developers in the San Francisco-Oakland-Fremont, CA metropolitan area at $186,640, which is about 37% above the national median of $135,980. The middle half of the local market sits between $163,500 and $219,670. Those are wages only: employer taxes, benefits, equipment and recruitment cost sit on top, and a vacant seat costs the delivery schedule rather than the payroll line.

How large is the San Francisco developer market?

The Bureau counts 69,030 software developers employed in the San Francisco-Oakland-Fremont, CA metropolitan area. Employment counts matter more than wage figures when you are estimating how long a search will take, because a high median in a thin market usually means a long search rather than an expensive one.

Is software development concentrated in San Francisco?

The location quotient for software developers in this metro is 2.68. A value above 1 means the occupation is more concentrated here than in the country as a whole. At 2.68, San Francisco the work is clearly concentrated here. The practical consequence is that candidates are available but so are competing offers, and counter-offers late in a process are common.

Should we hire locally in San Francisco or work with a remote team?

It depends on what the work needs rather than on cost alone. Work that requires being in a room with regulated data, physical hardware or a specific client site has to be local. Most product engineering does not. The honest comparison is not a local salary against an hourly rate, but the fully loaded cost of a local seat, including the months it stands empty, against the cost and the coordination overhead of a team in another time zone.

How does the San Francisco time zone affect working with a distributed team?

San Francisco runs on Pacific. The workable pattern for most teams is a committed overlap window rather than full-day coverage: two to four hours when everyone is online for review, decisions and anything that would otherwise take a day of messages. Teams that try to run without a fixed overlap lose more to waiting than they ever save on rate.

Which industries drive technical hiring in San Francisco?

The local anchors are venture-backed SaaS, fintech, AI and machine learning and marketplaces. That mix matters when you read a CV: a developer whose experience comes from the dominant local industry brings domain knowledge you may value, and habits formed in that industry that may not fit a different kind of product.

Where does the local talent pipeline come from?

The main feeders are UC Berkeley, University of San Francisco and San Francisco State University. University pipelines matter most for junior hiring and least for senior hiring, where people are recruited out of other companies rather than out of programmes.