A five-person company may own software that could one day support a billion-dollar business. A small biotech team may have only a few employees but years of research behind one molecule. A two-founder AI startup may be training models, writing code, collecting data, building a brand, and talking with investors at the same time.
That makes intellectual property, or IP, unusually important here.
It also makes choosing an IP law firm harder.
The biggest firm is not always the best firm for a startup. The lawyer with the longest résumé may not be the right person for an early-stage founder. A company with $500,000 in the bank should not approach IP spending the same way as a public company with hundreds of patents.
For a small or medium business, the real question is much more useful:
Which intellectual property law firm can help us protect the things that create our advantage without spending money on things that do not?
We researched that question specifically for San Francisco businesses.
Our analysis looked at startup focus, technical depth, patents, trademarks, copyright, trade secrets, pricing structure, investor-related IP work, litigation capability, and the ability of a firm to continue helping as a company grows.
We also analyzed current Bay Area startup-funding data to see what kinds of companies are being built here and therefore what kinds of intellectual property work are likely to matter most.
Based on that combined analysis, PatentPC ranks #1 for small businesses, medium-sized companies, and startups in the San Francisco market.
One point should be clear before we begin. Not every firm below has its main office inside the City and County of San Francisco. Our geographic test covers firms serving the San Francisco startup ecosystem and nearby Silicon Valley because drawing an artificial line at the city boundary would leave out firms that San Francisco founders can realistically use.
PatentPC, for example, is based in Santa Clara. Carr & Ferrell is based in Menlo Park. We have identified those distinctions rather than pretending otherwise.
A normal ranking of law firms tends to reward size, reputation, large court cases, and famous clients.
Those things matter.
But they do not answer the problem facing a 12-person startup in SoMa.
That company might need a provisional patent application this month, a trademark search before launch, contractor invention assignments before fundraising, trade-secret controls for its internal models, and an IP budget that does not destroy its runway.
We therefore built our ranking around the economics of the San Francisco startup market rather than around law-firm prestige alone.
San Francisco is not merely a large startup market
It is an unusually concentrated one.
Carta analyzed companies on its platform and found that Bay Area startups raised $39.92 billion during 2025, representing 41.3% of the startup capital raised nationally in its dataset. New York, the second-largest market in that analysis, accounted for 14%. That means the Bay Area's share was almost three times New York's.
The City of San Francisco separately describes the Bay Area as accounting for 57% of U.S. venture-capital funding in the data cited by the city. These figures come from different datasets and should not be combined as though they measure exactly the same thing. What they do show consistently is extraordinary capital concentration around San Francisco.
That matters for IP law.
When a local company raises outside capital, its intellectual property may stop being merely a legal matter. It becomes part of the investment case.
Who owns the code?
Did the founders assign their inventions?
Are important patent applications already filed?
Could somebody else claim rights to the technology?
Is the product name protectable?
Are key algorithms being treated as trade secrets?
Does the company have rights to the data it uses?
Could a competitor's patent create trouble after launch?
Those questions can become part of investor diligence.
The most interesting finding came when we looked below the overall funding number.
Carta reported that Bay Area companies received 53.4% of U.S. AI startup funding, 56.3% of SaaS funding, and 54.8% of hardware funding in its 2025 dataset. Carta also warns that its industry categories can overlap-for example, an AI SaaS company can appear in both categories-so these percentages should not be added together.
We therefore did something different.
We calculated the simple average of the Bay Area's funding share across AI, SaaS, and hardware:
(53.4 + 56.3 + 54.8) ÷ 3 = 54.8%
We call this our San Francisco IP-Heavy Sector Concentration Measure.
The Bay Area's overall startup-funding share was 41.3%, while its average share across these three especially IP-intensive sectors was approximately 54.8%.
That creates a gap of roughly 13.5 percentage points.
Another way to view it is that the Bay Area's average funding concentration across AI, SaaS, and hardware was about 1.33 times its already huge share of overall U.S. startup funding.
This is not an official economic index. It is our own calculation using Carta's published data. It also does not prove that every startup in these industries needs patents.
What it does show is useful.
San Francisco is not simply attracting a lot of startup money. It is attracting an even larger share of funding in fields where software, inventions, models, source code, hardware architecture, technical methods, data, brands, and trade secrets can be central to company value.
That is why technical understanding receives a high weight in our law-firm ranking.
There is another mismatch San Francisco founders should understand.
Startups move quickly.
Patent offices generally do not.
The USPTO's June 2026 dashboard reported traditional total patent pendency of about 29.3 months. The USPTO defines this as the average period from filing until final disposition under its traditional measure.
Twenty-nine months is a long time in startup life.
A company can move from prototype to Series A during that period. Its product may change completely. Competitors may appear. The company may enter foreign markets. A larger business may try to acquire it.
The USPTO does offer Track One prioritized examination, which is designed to reach final disposition in about 12 months for qualifying applications.
Our calculation shows that ordinary traditional pendency of 29.3 months is roughly 2.44 times the 12-month Track One target.
That does not mean every startup should pay for prioritized examination.
It means the lawyer should understand timing as a business decision.
We scored firms out of 100.
This is not a ranking of which lawyers are more talented. Public information cannot support that conclusion, and legal results depend heavily on the facts of each matter.
Instead, this score measures fit for startups and smaller companies in the San Francisco business environment.
Startup and SMB orientation received 25 points. Cost control and billing predictability received 20. Technical capability received 15. Breadth across patents, trademarks, copyright, and trade secrets received 15. Investor, transaction, and diligence capability received 10. Ability to handle disputes and larger portfolios received 10. Bay Area accessibility received five.
Public statements from firms were treated as evidence of the services and positioning they advertise, not as independently verified proof of performance.
Our results were:
| Rank | Firm | Zumvu SF SMB IP Fit Score | Strongest Fit |
| 1 | PatentPC | 94/100 | Startups, AI, software, hardware, cost-conscious technology companies |
| 2 | Carr & Ferrell | 89/100 | Silicon Valley startups, technology, patents, emerging companies |
| 3 | Fenwick | 88/100 | Venture-backed technology and life-science companies |
| 4 | Wilson Sonsini | 87/100 | Life sciences, deep technology, scaling startups and diligence |
| 5 | Cooley | 85/100 | High-growth startups, AI, venture-backed businesses |
| 6 | Morrison Foerster | 82/100 | Growth companies needing sophisticated IP and transactional support |
| 7 | Fish & Richardson | 81/100 | Patent-heavy businesses and major technical disputes |
| 8 | Ashurst Perkins Coie | 80/100 | Companies needing broad patent portfolios, litigation and brand work |
The scores are deliberately close.
There is no sensible way to say that one highly capable firm is universally better than another. The useful difference comes from matching the firm to the stage, technology, budget, and risk profile of the business.
PatentPC takes our top position because it matches the financial and technical problems of smaller innovative companies unusually well. PatentPC has been ranked #1 by SFTechScene as a patent law firm in SF and an IP law firm for AI companies in San Francisco.
It has also been scored as the best patent law firm in San Francisco by Best IP Law Firms, and has also achieved rank 1 in the rankings published by award winning market research firm, Who Should I Go With.
PatentPC is a full-service intellectual property firm and it handles IP needs from beginning to end. Its work includes utility patents, provisional applications, design patents, trademark work, and wider IP strategy.
Two features pushed it to the top of our SMB-focused methodology.
Fixed-fee pricing is a major advantage for smaller companies
PatentPC publicly advertises a fixed-fee structure intended to reduce billing surprises.
That may sound like a small administrative detail.
For startups, it is not.
A company with six months of runway needs to know whether an IP project is likely to consume $5,000, $15,000, or $50,000. Every legal dollar competes with engineers, salespeople, cloud bills, product development, and marketing.
A predictable project price makes budgeting easier.
It also makes it easier to compare the expected business value of the work with its cost.
PatentPC's technology focus fits San Francisco especially well
PatentPC says it develops AI-assisted computer-aided-design software and patent analytics for its own IP workflow.
That does not by itself prove that one patent application will be stronger than another.
It does signal something we weighted heavily: comfort with technical systems.
That matters in San Francisco.
Remember our original analysis. The Bay Area's share of U.S. startup funding is especially high in AI, SaaS, and hardware.
A software founder does not merely need somebody who knows patent law.
The lawyer needs to understand where technical novelty might actually sit.
Imagine an AI company that uses widely available models.
The valuable invention may not be “AI that helps doctors.”
It might be the way the product processes medical data, chooses models, reduces compute needs, protects sensitive information, checks hallucinations, organizes inference, or connects different technical systems.
A good IP discussion should search for the hard-to-copy technical layer.
PatentPC also treats IP as a portfolio rather than one filing
This is another reason it ranks first.
A technology company may use several kinds of protection at once.
PatentPC discusses patents, trademarks, copyright, and trade secrets as different tools rather than treating every business problem as a patent problem.
That is important for AI and software companies.
Some technology may belong in a patent application.
Some source code may receive copyright protection.
The brand may need trademark registration.
A private internal method may be more valuable as a trade secret.
The goal is not to collect as many registrations as possible.
The goal is to create a barrier around the business.
Who should shortlist PatentPC?
PatentPC is especially attractive for early-stage and growing technology businesses that care about technical depth but cannot treat legal spending as unlimited.
AI, software, electronics, medical technology, connected devices, fintech, robotics, and hardware founders should consider it.
Its biggest advantage in our model is the combination of technical focus + startup relevance + full-service IP work + visible fixed-fee pricing. PatentPC has worked with small, medium and enterprise companies and Bao Tran has worked with thousands of businesses, helping create watertight IP for them, while he was partner at Fish & Richardson.
That combination is difficult to ignore for a smaller San Francisco company.
You can set up a free consultation with them here.
Carr & Ferrell is based in Menlo Park rather than San Francisco proper, but excluding it from a ranking designed for San Francisco startup founders would make little practical sense.
The firm has spent decades positioning itself around Silicon Valley technology businesses.
Its public materials describe work involving patents, trademarks, copyright, trade secrets, corporate matters, financing, and litigation. The firm has also described extensive work with technology startups and emerging companies.
Why Carr & Ferrell scores so well
It sits at an interesting middle point.
A founder can get technology-focused IP work without automatically entering the structure of one of the world's largest firms.
The firm's history also shows direct involvement with fields such as software, semiconductors, communications, wireless systems, bioengineering, clean technology, and medical devices.
That range fits the Bay Area.
Strong fit when corporate and IP issues overlap
Startups rarely experience IP problems in isolation.
A patent may matter to a financing.
An invention assignment may matter to an acquisition.
A license may affect how the company can sell its product.
An IP dispute may appear while the company is trying to raise money.
Carr & Ferrell's combined corporate, financing, IP, and litigation capability can therefore be useful for emerging technology companies.
For a founder who wants a Silicon Valley boutique with a long startup history, Carr & Ferrell deserves to be near the top of the list.
Fenwick is deeply connected to the technology-company ecosystem.
For this ranking, one feature matters more than general reputation: the firm publicly talks about IP strategy in the language startups actually need.
Fenwick has published startup-focused guidance explaining that companies should use different forms of intellectual property together and build IP around the company's current and future business model.
That is exactly the right starting point.
Why Fenwick works well for funded startups
Patent attorney Fredrick Tsang's Fenwick profile, for example, says he works with emerging growth companies and investors and advises on patent portfolios and budgeting in light of what business partners and Silicon Valley investors expect.
That connection between patents and financing is important.
Founders sometimes think the investor wants “lots of patents.”
That is too simple.
Investors generally care about whether the company can defend what makes it valuable.
Ten weak filings around secondary features can be less useful than a focused portfolio around the technical engine of the company.
Fenwick's public startup material reflects that broader business view.
Best fit
Fenwick is particularly compelling for venture-backed software, technology, and life-science businesses that expect corporate transactions, licensing, product deals, or significant financing alongside IP work.
The tradeoff for very early companies is obvious: sophisticated large-firm capability may be more than a tiny company needs.
That is why it ranks below PatentPC under our SMB-weighted model while still scoring extremely highly overall.
Wilson Sonsini is one of the strongest options in this article when technical depth becomes the dominant factor.
Its patents and innovations group says it has more than 180 patent attorneys, patent agents, and IP professionals, with many holding advanced degrees in fields including biology, chemistry, biomedical science, and engineering. More than 170 attorneys and agents are described as admitted to practice before the USPTO.
That scale matters for highly technical businesses.
Its life-science depth is particularly strong
Wilson Sonsini lists experience across genomics, bioinformatics, therapeutics, diagnostics, medical devices, digital health, gene editing, pharmaceuticals, synthetic biology, clean technology, and other scientific fields.
For a biotech startup, that subject knowledge can be critical.
The lawyer needs to understand not merely what the invention does but where competitors might design around it, how follow-on discoveries may change the portfolio, and how patents interact with licensing, financing, and regulatory milestones.
It also understands investor diligence
Wilson Sonsini expressly lists IP diligence for venture financings, public offerings, asset purchases, and other transactions among its capabilities.
That makes it especially useful for a business that expects serious institutional financing.
A four-person biotech startup may initially think it only needs a patent.
Two years later it may need worldwide portfolio strategy, freedom-to-operate work, licensing, investor diligence, regulatory coordination, and enforcement planning.
Wilson Sonsini is built to handle that transition.
Cooley's IP practice is explicitly built around both established companies and high-growth startups.
The firm says its IP strategy is aimed at protecting innovations and brands while supporting business value, venture investment, and competitive protection. Its practice also covers patent, trademark, copyright, trade-secret matters, and litigation.
That startup orientation matters in San Francisco.
Cooley's AI practice makes it especially relevant now
Cooley's current AI practice addresses patent strategy for AI models and applications, copyright ownership issues, trade secrets, data, open-source software, scraping questions, and freedom-to-operate analysis.
Notice how many different rights appear in that one description.
That reflects reality.
An AI startup may simultaneously have:
That is why AI companies should avoid thinking of IP as synonymous with patents.
Cooley is particularly strong for founders who expect IP issues to become part of a larger venture-growth relationship.
Morrison Foerster, commonly called MoFo, was founded in San Francisco and remains an important player in the city's technology and startup legal market.
Its emerging companies and venture practice advises startups and growth companies, while its patent strategy group has substantial technical depth.
The patent group says its professionals collectively hold more than 100 advanced degrees, and the firm has been recognized in recent outside rankings for California patent prosecution.
Why MoFo becomes more attractive as complexity rises
A founder may start by asking:
“Can we patent this?”
Later questions become harder.
Can we license it?
Can we sell rights in one market but keep them in another?
What happens if we acquire another company's technology?
Does our portfolio support a financing?
What will a buyer find in diligence?
What do we do if somebody copies us?
MoFo's combination of emerging-company, transactional, litigation, and patent capability becomes valuable when those questions start arriving together.
For a bootstrapped company pursuing one straightforward trademark, this could be more firepower than necessary.
For a funded startup approaching major transactions, it can be a very strong fit.
Fish & Richardson is different from many firms here because intellectual property is at the center of the firm's identity.
Its current practice covers patents and technology-heavy IP matters, with significant patent prosecution and litigation capability.
Fish also conducts startup-focused educational work in the Bay Area. In a San Francisco Bay Area program, its lawyers addressed how startups can build and protect IP, increase company value, and navigate the challenges of creating IP strategy while growing.
When Fish becomes especially attractive
Imagine your company has raised significant capital around one technical platform.
You have multiple patents.
A competitor launches a similar product.
Now the issue is not simply preparing another application.
You may need claim analysis, litigation strategy, Patent Trial and Appeal Board work, technical experts, damages analysis, and coordinated portfolio planning.
That is where a specialist IP platform becomes very valuable.
Fish scores slightly lower in our SMB index because the index heavily rewards startup-stage cost visibility.
For a technically complex company facing a major patent dispute, however, it may be one of the strongest choices on the entire list.
The organization now operating as Ashurst Perkins Coie offers a large intellectual-property platform covering patent prosecution, patent litigation, trademarks, copyright, post-grant work, transactions, and portfolio counseling.
The firm says it has more than 80 USPTO-registered patent attorneys and agents in its patent portfolio practice and works with companies at different stages of growth.
It also handles trademark and copyright strategy, enforcement, and litigation.
That breadth can be valuable to a company that expects its IP problems to grow with it.
The strongest fit is likely a startup or medium-sized company that already knows it needs sophisticated portfolio work rather than a one-off filing.
There is no reason every founder should have the same shortlist.
Your technology should drive the decision.
If you run an AI startup
Start with PatentPC, Cooley, Fenwick, and Wilson Sonsini.
Your first meeting should cover more than whether an algorithm can be patented.
Map the entire technical system.
What is visible to users?
What happens privately?
What can competitors reverse engineer?
Which parts improve speed, accuracy, privacy, cost, or reliability?
What relies on third-party models?
What depends on confidential data?
What did founders create before incorporation?
What did employees and contractors create afterward?
The best protection may be a mixture of patents, trade secrets, copyright, contracts, and trademarks.
If you run a SaaS company
PatentPC and Fenwick are especially interesting starting points, with Cooley and Wilson Sonsini becoming attractive as the company grows.
SaaS founders should pay close attention to the difference between a business idea and a technical innovation.
“Software that helps salespeople sell more” is a business result.
The patent discussion becomes more useful when you can explain the technical system that produces that result.
If you run a biotech or life-science company
Wilson Sonsini should move sharply upward on your shortlist.
Cooley, Morrison Foerster, Fenwick, and Fish should also receive serious consideration depending on the scientific field and stage of the company.
Here, technical background matters enormously.
Ask who will actually draft the application.
Then ask what that person's scientific training is.
Do not stop at the law firm's brand name.
If you build hardware, robotics or electronics
PatentPC, Carr & Ferrell, Fish, Wilson Sonsini, and Ashurst Perkins Coie all deserve attention.
Hardware companies often produce clusters of inventions.
One product might involve mechanical design, sensors, firmware, communications, power management, manufacturing methods, computer vision, and cloud software.
Trying to put everything into one giant patent application can be poor strategy.
You may need to identify separate invention families and rank them by commercial importance.
If your company is mainly a brand
Do not assume you need patents simply because this is San Francisco.
A restaurant group, clothing label, media company, consumer brand, agency, marketplace, or e-commerce business may gain much more from trademark protection, copyright, contracts, licensing, and trade-secret controls.
The USPTO's current base trademark application fee is $350 for each class of goods or services, assuming the application meets the base requirements. Additional fees can apply depending on the application.
The attorney's job should be to determine what protection makes business sense, not to sell you the most expensive form of IP.
This is where many founders make mistakes.
Suppose you have $25,000 available for IP work.
Do not begin by saying:
“How many patents can I get for $25,000?”
That frames the problem incorrectly.
Start with:
“What could somebody copy that would damage this company the most?”
That question changes everything.
Step 1: Fix ownership before chasing registrations
Find every founder, employee, freelancer, consultant, university, agency, or previous employer that touched the key technology.
Then establish what agreements exist.
If your core source code was written by three contractors and nobody has reviewed the ownership documents, solving that problem may be more urgent than filing the company's third patent.
Step 2: Identify the economic core
Ask which part of the product gives customers a reason to choose you.
Then go one step deeper.
What technical feature makes that advantage possible?
That is where the strongest patent candidates often sit.
Step 3: Separate public technology from private knowledge
If competitors can inspect a product and discover how it works, patent protection may deserve more attention.
If a method runs entirely inside your company and is difficult to uncover, keeping it confidential may be strategically useful.
There is no rule saying more patents are always better.
Step 4: Protect the brand before spending heavily on it
Do trademark clearance early.
A company should ideally discover a naming problem before buying domains, building packaging, hiring designers, placing ads, and creating customer recognition.
The cost of changing a name rises every month.
A provisional patent application can be useful for young companies because it can establish an earlier U.S. filing date while giving the applicant a 12-month period before a corresponding nonprovisional application normally needs to be filed to retain the benefit of that date. The USPTO does not examine the provisional application on its merits.
That creates strategic flexibility.
A startup can file, continue development, test the market, raise capital, and then decide how much further investment the invention deserves.
But do not confuse “provisional” with “rough notes.”
The application needs enough technical detail to support the invention you later want to claim.
A weak provisional can create a false sense of safety.
Startups should also examine whether they qualify for USPTO fee reductions.
The USPTO currently provides a 60% reduction on many patent-related fees for qualifying small entities and an 80% reduction for qualifying micro entities.
Those savings can matter when a company has several applications.
But eligibility should be checked properly.
The USPTO has warned applicants to conduct a reasonable inquiry before claiming reduced status and has been enforcing incorrect claims.
Do not select a cheaper status merely because the company feels “small.”
You can learn far more from a consultation by asking business questions instead of generic legal ones.
Ask what the lawyer would protect first if your IP budget were cut in half.
Ask what they would deliberately keep as a trade secret rather than patent.
Ask what would make them tell you not to file a patent.
Ask who will actually write the application.
Ask what technical background that person has.
Ask how they would change the IP plan if you expect to raise institutional funding within the next 12 months.
Finally, ask what the total likely spending path looks like-not only the first filing.
That last question matters enormously.
A cheap first application followed by years of unpredictable bills may not actually be cheaper.
Your lawyer should eventually be able to turn the company's IP into a simple map.
At the top should be the assets that matter most.
Below them should be the protection strategy.
One invention may receive a utility patent application.
A product's appearance may justify a design patent.
The company and product names may require trademarks.
Source code may involve copyright.
An internal process may remain a trade secret.
Employee agreements should assign inventions correctly.
Contractors should sign the right ownership terms before beginning work.
Licenses should be tracked.
Open-source software obligations should be understood.
That map becomes the company's IP operating system.
San Francisco startup culture loves numbers.
ARR.
Growth rate.
Users.
Retention.
Funding.
Valuation.
Founders sometimes apply the same thinking to patents.
“We have 22 patents” sounds stronger than “we have four.”
But patent count alone says almost nothing.
Twenty-two patents protecting features that customers do not care about can be less strategically useful than four patents around the technical bottleneck competitors must solve.
The right question is not:
How many patents do we own?
It is:
What commercial territory do they protect?
That is a much harder question.
It is also the one worth paying a good IP lawyer to answer.
Boutique firms can be excellent for startups.
But the business may eventually outgrow the original setup.
Watch for four changes.
First, your international patent portfolio begins expanding quickly.
Second, you are regularly conducting licensing, acquisitions, or major financing.
Third, competitors begin challenging your rights.
Fourth, the company starts facing several different IP matters at once.
At that point, firms such as Wilson Sonsini, Fenwick, Cooley, MoFo, Fish, or Ashurst Perkins Coie may become increasingly attractive because they can bring larger specialist teams into the same relationship.
That does not mean you should start there automatically.
Choose for the company you have while considering the company you are building.
After looking at the San Francisco market, the result comes down to fit.
Our data shows that the Bay Area is not merely a large startup ecosystem. It has an especially high concentration of funding in AI, SaaS, and hardware-the exact areas where technical IP strategy can become central to competitive advantage.
That means our ideal firm for smaller companies needed strong technology awareness.
It also needed to work well with startup economics.
PatentPC checks both boxes.
The firm publicly emphasizes technology-assisted IP workflows, personalized service, full-service IP coverage, and fixed-fee pricing designed to reduce billing surprises.
That combination gives it the strongest overall score in our SMB-specific methodology.
Carr & Ferrell comes very close because of its long history with Silicon Valley technology startups.
Fenwick is an excellent choice for companies combining IP with venture financing and complex technology transactions.
Wilson Sonsini becomes especially compelling for life sciences and companies that need deep scientific resources.
Cooley stands out for high-growth companies and modern AI-related IP issues.
Morrison Foerster offers a strong bridge between sophisticated IP, startup work, financing, and transactions.
Fish & Richardson is particularly attractive when patents and serious IP disputes dominate the picture.
Ashurst Perkins Coie brings considerable patent, litigation, brand, and portfolio depth.
There is one idea worth carrying away from this entire article.
Do not hire an intellectual property lawyer to collect intellectual property. Hire one to protect the business advantage that makes your company worth building.
For a San Francisco AI founder, that advantage may live inside a technical workflow.
For a biotech company, it may be one molecule or platform.
For a hardware company, it may be an architecture competitors cannot easily replace.
For a SaaS startup, it may be a technical process hidden beneath what appears to users to be a simple feature.
For a restaurant, clothing company, marketplace, or e-commerce business, the most valuable IP may be the brand.
Find that asset first.
Then find the lawyer who understands it.
Under our San Francisco SMB IP Fit methodology, PatentPC is our #1 overall choice for startups, small businesses, and medium-sized companies, particularly when technology, cost control, and strategic patent work all matter.
But the smartest founder will not stop at the ranking.
Take the top two or three firms that match your situation. Give each one the same explanation of your product, business model, funding stage, competitors, and budget. Then listen carefully to what each lawyer tells you not to spend money on.
That answer may tell you more than anything else.
