Patent Portfolio Strategy for Growing Tech Companies: A 2026 Playbook
The short answer. A patent portfolio strategy for a mid-sized technology company in 2026 should do four jobs at once: deter competitors from copying your core technology, give you leverage in cross-licensing or settlement, raise the floor on your valuation in a financing or acquisition, and create defensible margin around your most profitable products. Companies that get this right typically file between 5 and 30 new applications a year, run a freedom-to-operate (FTO) review before each major launch, and budget 0.5%–2% of R&D spend on patent prosecution and maintenance. Companies that get it wrong file too late, file too broadly, or treat patents as a cost line item instead of a strategic asset.
This guide walks through how growing technology companies — Series A through pre-exit — should think about patent portfolio strategy in 2026: what's changed, what to file, what to drop, how much to budget, and how to be ready when a buyer or acquirer asks "show me your IP."
It's written for founders, VPs of Engineering, GCs, and CFOs of mid-sized companies in software, AI/SaaS, biotech, cleantech and energy, and consumer products — the kinds of clients Lynch LLP works with every day.
Why "Growing" Companies Need a Different Patent Strategy Than Startups
The patent strategy that worked at the seed stage — file a provisional, hope it covers the demo, focus the budget on hiring — stops working somewhere around the Series A or first $5–10M of revenue. Three things change at once:
The threats get bigger. Once you have real revenue, you become worth suing. Once you have a recognizable product, you become worth copying. The companies that take notice of you are the ones with their own patent portfolios.
The budget gets bigger — but so does the scrutiny. Patent spend that was invisible at seed-stage becomes a board-level line item. The CFO wants to know what each application is doing for the company. "Filing because we should" is not an answer.
The exit becomes real. Acquirers, strategic partners, and Series B/C/D investors will run IP diligence. A weak or messy portfolio doesn't just lower your valuation — it can kill a deal or trigger an indemnity carve-out you spend years living with.
The companies we see succeed at this stage stop treating patent filings as a series of one-offs and start treating them as a portfolio with a job description.
The Four Jobs of a Mid-Stage Patent Portfolio
Every patent in a growing company's portfolio should do at least one of these four things. If a filing doesn't, it probably shouldn't be filed.
1. Defensive — Stop Copycats. Patents on the specific features customers buy you for. The thing your sales team demos. The user-visible behavior that gets screenshotted. These are the patents that make a competitor pause before cloning your product.
2. Offensive — Create Leverage. Patents on building blocks that other companies in your space need. These are the ones that matter in a cross-licensing negotiation, a settlement, or — for the more aggressive — an enforcement campaign. Most growing companies don't enforce, but having enforceable patents changes how everyone treats you.
3. Valuation — Justify the Multiple. Patents that strategic acquirers or Series B/C investors can point to as proof that you have something defensible. This is especially true in AI/SaaS and biotech, where buyers want to see that the technology — not just the team — is protected.
4. Margin — Secure the Premium. Patents on the technical implementations that let you charge more than commodity competitors. In hardware and energy, this often means manufacturing process patents. In SaaS, it's the system architecture and data pipeline. In biotech, it's the formulation or method of treatment.
A healthy mid-stage portfolio has filings in at least three of these four buckets. A portfolio that's all "defensive" is fragile. A portfolio that's all "offensive" is expensive and rarely gets used. The mix is what matters.
How Many Patents Should a Mid-Sized Tech Company File Per Year?
There is no single right number — but there are useful benchmarks. We see growing technology companies cluster into the following bands:
| Stage | Typical revenue / size | New U.S. apps / year | Annual IP spend (USD) |
|---|---|---|---|
| Late Series A / Early Series B | $2M–$15M revenue | 3–8 | $60K–$180K |
| Series B / Growth | $15M–$75M revenue | 8–20 | $150K–$450K |
| Series C / Pre-exit | $75M+ or 200+ FTE | 15–40 | $300K–$1.2M+ |
| Public / Mature | $250M+ revenue | 25–100+ | $1M–$5M+ |
These ranges are heavily industry-dependent. A 200-person AI/SaaS company often files fewer applications than a 50-person biotech, because biotech inventions are more discrete and more valuable per filing. A consumer-products company with strong design IP may file more design patents than utility patents.
The better question is not "how many patents per year?" but "what fraction of our patent-eligible innovations are we capturing?" If your engineers are inventing 30 protectable things a year and you're filing on three of them, the issue isn't budget — it's your invention-disclosure pipeline.
The Lynch LLP Mid-Stage Portfolio Framework
When we onboard a mid-stage client, we use a four-question framework before we file anything new. We've found these questions catch most of the strategic mistakes companies make on their own.
1. What is the business asset this filing protects? If you can't name a product, a feature, a customer-visible behavior, or a moat, the filing probably isn't worth the spend. "We thought it was cool" is not a business asset.
2. Who is the adversary? A patent is only as useful as the people it deters. Are we worried about VC-backed competitors, BigCo strategic plays, offshore copycats, or all three? Different adversaries call for different filing strategies — and very different international filing decisions.
3. What is the lifecycle of the underlying technology? A patent that takes three years to issue isn't useful if the technology is obsolete in eighteen months. For fast-moving spaces (most software, much of AI), we lean toward continuation strategies and provisional bridges so the claim language can adapt.
4. What is the exit horizon? A company eighteen months from a strategic acquisition needs a different portfolio than a company seven years from a possible IPO. Acquirers care most about: clean chain of title, no inventor disputes, key applications still pending (so they can shape claims post-acquisition), and reasonable foreign coverage.
These four questions should be asked at every quarterly portfolio review. The companies we work with that run this discipline rarely get IP-due-diligence surprises.
Foreign Filing Strategy: PCT and the National Phase
Most growing technology companies underspend on foreign filing — and a meaningful minority overspend. The PCT (Patent Cooperation Treaty) gives you a 30-month window from your earliest priority date to decide where, internationally, your invention is worth protecting. Used well, the PCT is the most cost-effective way to keep options open while you learn where your customers actually are.
Where to file is a business question, not a legal one. The countries where your competitors manufacture, where your customers pay, where your acquirers will look, and where enforcement actually works — those are the countries that belong on your foreign-filing list. The rest do not.
A defensible default for a U.S.-headquartered, mid-sized technology company in 2026 looks something like:
| Tier | Jurisdictions | Cost (per application, approx.) | When it makes sense |
|---|---|---|---|
| Tier 1 (almost always) | United States | $15K–$30K through allowance | Always |
| Tier 2 (frequent) | EPO (Europe), China | $20K–$50K combined | Selling in or competing in EU/China |
| Tier 3 (selective) | Japan, South Korea, Canada, Australia | $4K–$15K each | Industry-specific (semis, autos, biotech, consumer) |
| Tier 4 (rare, strategic) | India, Brazil, Mexico, GCC, ARIPO | $3K–$10K each | Specific manufacturing, distribution, or enforcement reasons |
If you're not already running this kind of jurisdiction-by-jurisdiction analysis at PCT national phase, your spend is almost certainly miscalibrated. Lynch LLP's International Protection team handles national phase entry for clients across the U.S., Europe, China, Japan, Korea, and most of the rest of the world.
A practical note on the U.S. as a foreign jurisdiction. Many of our clients have customers, but no employees, in the United States. For founders reading this from outside the U.S.: the U.S. remains the single most valuable jurisdiction for software, AI/SaaS, and consumer products patents because it's the largest enforceable market and home to the deepest pool of willing licensees and acquirers.
Freedom to Operate (FTO): When to Run It, What It Costs
A freedom-to-operate analysis answers a different question than a patentability search. Patentability asks "can we patent this?" FTO asks "if we ship this product, are we going to get sued?"
For growing technology companies, FTO matters most at three moments:
Before a major product launch. Especially a new product line, not just a feature release. The cost of a basic FTO is small relative to the cost of an injunction.
Before a financing round or acquisition. Buyer's counsel will ask. The right answer is "we ran FTO on our top three products in [date] and refreshed it last quarter." The wrong answer is "we've never done one."
When you receive a demand letter, infringement notice, or "courtesy copy" of a competitor's patent. This is not the moment to start.
A targeted FTO on a single product line — focused on the specific technical features the product implements — typically costs $10K–$40K and takes three to six weeks. A comprehensive FTO across a multi-product portfolio runs higher and longer. The output is a clearance opinion that ranks the patents-of-concern by risk and identifies design-around opportunities for the highest-risk ones.
The biggest mistake we see: companies treat FTO as a "yes/no" deliverable and walk away when the answer is "yes, with two yellow flags." The yellow flags are the part that matters. They're the patents you should be re-checking every six to twelve months as your product evolves.
Defensive Posture: Preparing for IPR and Litigation
Inter Partes Review (IPR) — the proceeding at the U.S. Patent Trial and Appeal Board where one party challenges another's patent claims — is the single most important post-grant tool in 2026. Mid-sized technology companies should understand IPR as both a sword and a shield.
As a shield. If your company gets sued for patent infringement in U.S. district court, you generally have a one-year window from being served to file an IPR petition challenging the asserted patent. IPR is statistically the highest-leverage defense available — a meaningful percentage of instituted petitions result in some or all challenged claims being canceled. For a mid-sized company that can't afford a multi-year jury trial, IPR can short-circuit the whole dispute.
As a sword. IPR is also a way to clear competitors' patents preemptively, before they're asserted against you, and before they're used to block your product. This is rarer in mid-market companies than it should be — usually because the cost feels high relative to the immediate ROI.
As something to be defended against. If you are the patent holder, IPR is what's coming if you ever try to enforce. Your prosecution strategy should anticipate this from day one. Strong, well-drafted, well-supported claims survive IPR. Weak claims do not. The cost of getting drafting right at filing is a tiny fraction of the cost of losing claims at IPR five years later.
The takeaway for portfolio strategy. When you write claims, write them as if you'll have to defend them at IPR. When you evaluate a competitor's patent, evaluate it as if you might want to challenge it. And when you build the portfolio, build it deep enough that losing one or two patents in IPR doesn't gut your position.
IP Due Diligence: Getting M&A-Ready
By the time an acquirer's counsel sends you their IP diligence checklist, the portfolio decisions are already made — you're just collecting evidence of them. Mid-sized technology companies that consistently transact well share a few habits.
- Clean chain of title. Every inventor on every patent has a properly executed assignment to the company. Every contractor who touched anything patentable signed an IP assignment before they wrote a line of code. This sounds basic; it is the single most common diligence problem we see.
- Inventor list integrity. The named inventors actually invented the claims. Founders, executives, and friends-of-the-CEO are not on patents they didn't conceive. Inventor disputes can invalidate an entire patent.
- No undisclosed open-source contamination. Especially in AI/SaaS, where models and training pipelines often have surprising provenance. Acquirers in 2026 are asking detailed questions about training data, open-source licensing, and AI model derivation. If your portfolio rests on technology with murky provenance, you have a problem that legal alone can't fix.
- A documented FTO posture. Even one prior FTO on the company's flagship product is reassuring; zero is alarming.
- A reasonable foreign filing footprint. Acquirers want to see that you've been thoughtful about international protection — not necessarily expensive about it.
- A plausible-looking litigation/IPR record. Either no disputes, or disputes that were handled professionally and concluded reasonably.
The companies that score well on these dimensions get higher valuations and shorter diligence periods. The companies that don't get re-trade demands, indemnity carve-outs, escrow holdbacks, and — sometimes — broken deals.
Common Mistakes Mid-Sized Companies Make
A few patterns we see repeatedly:
- Filing too late. The single most expensive mistake. Public disclosure, customer demos, or a published academic paper before filing can permanently destroy patent rights in most countries outside the U.S. The "we'll patent it later when we have more money" instinct is the most common way growing companies destroy IP value.
- Filing too broadly. The opposite mistake. Filing on every engineering decision is expensive, dilutes attention, and produces a portfolio that's hard to defend. File deeply on what matters.
- Letting the patent attorney drive the strategy. Patent counsel should execute the strategy. The strategy should be set by the company's leadership, informed by counsel. If your law firm is deciding what's important to your business, that's a sign of a misaligned engagement — not a sign your law firm is doing a great job.
- Optimizing for "issued patents" as a vanity metric. A patent that issues but covers something nobody copies is a wasted filing. A patent that gets one strong claim through prosecution and accurately covers what your competitors are actually doing is worth ten "issued" patents.
- Confusing patent filings with trade secrets. Patents publish. If your real moat is something better kept secret — model weights, training data composition, manufacturing process specifics — patenting may actively destroy the moat.
- Not budgeting for maintenance fees. U.S. utility patents require maintenance fee payments at 3.5, 7.5, and 11.5 years from issuance. Foreign jurisdictions require annuities every year. A portfolio of 40 issued patents in 8 jurisdictions can quietly accumulate $50K–$150K/year in maintenance costs. Plan for it; prune accordingly.
Budget Benchmarks: What a Mid-Sized Tech Company Should Spend on IP
The most useful frame for IP budget is "as a percentage of R&D." Mid-sized technology companies typically spend:
- 0.5%–1.0% of R&D on IP at the conservative end (defensive-only posture, modest foreign filing).
- 1.0%–2.0% of R&D for companies with active offensive intent, real foreign filing programs, or higher-value-per-patent industries (biotech, cleantech, hardware).
- 2%+ of R&D for portfolios with significant litigation or licensing activity, or for companies in the late-stage run-up to an exit.
If your IP spend is below 0.5% of R&D, you almost certainly have a coverage gap. If it's above 3% without a specific reason (active litigation, licensing program, M&A prep), the spend is probably miscalibrated.
A separate-but-related question is how that spend gets structured. Mid-market companies increasingly prefer fixed-fee or capped-fee arrangements for prosecution work, hourly for advisory and dispute work, and budgeted-by-jurisdiction for foreign filing. Boutique IP firms — including Lynch LLP — are typically more flexible on fee structure than BigLaw.
Frequently Asked Questions
How much does it cost to file a U.S. patent in 2026?
A standard U.S. utility patent application drafted, filed, and prosecuted to allowance typically costs $15,000–$30,000 in legal fees plus USPTO fees, depending on complexity, claim count, and the number of office actions. Most applications go through one to three office action responses before allowance. Provisional applications run lower — typically $3,000–$8,000 — and serve as a one-year placeholder. Small-entity and micro-entity discounts at the USPTO can reduce government fees by 60% or 80% respectively.
How long does it take to get a patent issued?
The average pendency at the USPTO in 2026 is roughly 24–36 months from filing to issuance, varying by technology center. AI and software applications often run longer; mechanical and biotech can be faster. The Track One prioritized examination program can accelerate issuance to 12 months for an additional fee.
How many patents does a Series B tech company typically have?
A typical Series B technology company has filed 5–20 U.S. applications and has 2–8 issued patents, with a similar count of pending applications. AI/SaaS companies often skew lower; deep-tech and biotech skew higher.
What's the difference between a provisional and a non-provisional application?
A provisional application is a one-year placeholder that establishes a priority date but is never examined and never issues as a patent. A non-provisional application is the real, examined application. Most companies file provisionals when they're moving fast and need a priority date now, then convert to a non-provisional within twelve months. Provisional applications must contain enough technical disclosure to support the eventual non-provisional claims — a thin provisional offers thin protection.
Should we file in China?
Probably yes if (a) you sell into China, (b) your competitors manufacture in China, or (c) you ever expect to license, partner with, or be acquired by a China-based company. The cost of EPO and China together is meaningful but defensible for most growing technology companies. The cost of skipping them and discovering the gap five years later is usually much larger.
When should we run a freedom-to-operate analysis?
Before a major product launch, before a financing or M&A event, and immediately after receiving any kind of patent demand letter or notice. A targeted FTO on one product runs $10,000–$40,000.
Can software and AI inventions be patented in 2026?
Yes — but it requires careful claim drafting. The USPTO's 2024 guidance on AI inventions and subsequent updates have created a workable framework for AI/software patentability when claims are tied to specific technical implementations rather than abstract ideas. See our deeper treatment in Why 2026 Is the Best Year Ever to File AI Patents.
How do we know if our existing portfolio is strong or weak?
A useful rule of thumb: a strong portfolio has at least one issued patent reading directly on each of your top revenue-driving products, defensible foreign coverage in your top three commercial markets, no clouded chain of title, and at least one pending application that allows for continued claim refinement. A weak portfolio is missing two or more of those.
What's the difference between a boutique IP firm and BigLaw for patent work?
Boutique IP firms like Lynch LLP typically offer lower hourly rates, more attorney continuity (you know who's drafting your application), more flexible fee structures, and deeper specialization in IP. BigLaw firms offer broader practice areas (corporate, employment, M&A under one roof), larger teams for litigation, and brand recognition that some acquirers value. For most mid-sized technology companies focused on IP outcomes rather than full-service legal coverage, a boutique is the better economic and strategic choice.
How long does IP due diligence take in an M&A transaction?
Typically 2–6 weeks of focused work, depending on portfolio size and how organized the company's records are. A clean, well-documented portfolio can be diligenced in days. A messy one can stretch into months and put deal timing at risk.
Working with Lynch LLP
Lynch LLP is an intellectual property and business law boutique serving startups and mid-sized technology companies across Southern California and beyond. Our patent attorneys hold engineering degrees in mechanical and electrical engineering and have built portfolios for clients in AI and software, biotech, energy and cleantech, and consumer products.
We work with growing companies on:
- Building and pruning patent portfolios with a clear job description for each filing.
- PCT and foreign filing strategy, including national phase entry across the EU, China, Japan, Korea, and the rest of the world.
- Freedom-to-operate analyses tied to product launch and financing milestones.
- Inter Partes Review (IPR) — both as petitioner and patent owner.
- Patent and trademark litigation and enforcement.
- IP due diligence for financings and M&A.
If you're wrestling with any of the questions above, we'd be glad to talk.
Book a ConsultationThis article is for general informational purposes and is not legal advice. Patent strategy depends on the specifics of your technology, your business, and your competitive environment. For advice on your situation, please contact a qualified patent attorney.