# IP Budgeting: What Protection Should Cost at Each Company Stage > How much should a company spend on IP? Realistic annual IP budget ranges from bootstrap to enterprise, where budgets blow up, and the cost-control levers. Guide | Author: Lidiia Levitska | Source: Intellectual Property Law (outsideipcounsel.com) Canonical: https://outsideipcounsel.com/guides/ip-budgeting-by-stage/
Quick answer: A sensible IP budget scales with company stage: roughly $1K–$5K per year bootstrapped (trademark and copyright registrations, a provisional patent, assignment hygiene), $15K–$50K at seed (one to three nonprovisional patents, watch services, early foreign decisions), $50K–$250K when scaling (patent families, foreign nationalization — the budget killer — and the first maintenance fees), and at enterprise scale the discipline flips from filing to pruning. The two rules that matter most: budget for the whole life of each asset, not just the filing, and spend on what competitors would actually copy and you'd actually enforce. This is general education, not legal advice — have an attorney licensed in your jurisdiction review your specific situation.
Your patent attorney quotes $12,000 for a nonprovisional application and you approve it — not realizing you've actually committed to $30,000-plus over the asset's life once office actions, issue fees, maintenance fees, and the inevitable "should we file this in Europe?" conversation arrive. That gap between the sticker price and the lifecycle cost is why companies need an **IP budget**, not just a price list. This guide is the planning view: what a company should spend at each stage, where budgets actually die, and the levers that control costs. For the per-item price sheet — what each individual filing costs — see [how much does IP protection cost?](/guides/how-much-does-ip-protection-cost/); this guide assumes those prices and asks how to allocate across a portfolio and a roadmap. Both sit under the [IP strategy and portfolio management guide](/guides/ip-strategy-and-portfolio-management/). ## Why budget by stage instead of by asset? Because IP spending is a **portfolio of multi-year commitments**, and the right portfolio changes as the company does. A pre-revenue startup filing five nonprovisionals is usually burning runway on assets it can't yet exploit; a Series C company still relying on one aging provisional is naked. Stage-based budgeting forces the two questions that per-item pricing hides: 1. **Lifecycle cost:** every filing is the down payment on a 20-year stream of prosecution, issue, maintenance, and possibly foreign costs. 2. **Opportunity cost:** every dollar on patent #6 is a dollar not spent on the trademark watch or the trade-secret program. One framing note on the ranges below: they assume outside counsel at typical U.S. rates, and USPTO fees under the schedule effective January 2025, where **small entities get a 60% discount and micro entities 80%** on most patent fees. If you qualify (small entity: generally under 500 employees and no obligations to large entities; micro entity: additional income and filing-history limits), claim it — it roughly halves-to-quarters your government-fee line. ## Bootstrap stage: ~$1,000–$5,000 per year Pre-funding, the goal is **hygiene and optionality**, not a portfolio: - **Trademark.** A federal application at the USPTO base fee of **$350 per class** (plus surcharges for custom goods/services descriptions, and attorney help if budget allows). Your name is the asset most likely to be worth something at every future stage, and rebranding after traction is far costlier than filing early — the [trademark case-law archive](/topics/trademarks/) is full of disputes that a $350 filing would have prevented. - **Copyright registrations.** **$45–$65 each** online for key assets (core code releases, flagship content). Cheap, and registration is a prerequisite to a U.S. infringement suit under [17 U.S.C. § 411(a)](https://www.law.cornell.edu/uscode/text/17/411). - **A provisional patent application** if the technology genuinely warrants it — the USPTO fee is modest (about $130 for small entities under the 2025 schedule), and even attorney-drafted provisionals typically run in the low thousands. It buys a 12-month option on a filing date. - **Assignment hygiene: the non-negotiable.** Present-assignment IP clauses from every founder, employee, and contractor. This costs almost nothing now and is the single most common diligence failure later. If you're deciding which of these even applies to your business, start with [which IP protection do you need?](/guides/which-ip-protection-do-you-need/) ## Seed / funded stage: ~$15,000–$50,000 per year With runway comes the first real portfolio decisions: - **One to three nonprovisional patent applications**, typically $10,000–$25,000 each in attorney fees to draft and file, plus USPTO fees (about $800 filing/search/examination for a small entity under the 2025 schedule) — and budget another few thousand per application for **office-action responses**, because first rejections are the norm, not a crisis. - **The foreign question looms.** A **PCT application** (a few thousand dollars) keeps international options open without committing to any country yet. Whether to file it at all is a strategy call — see [international IP protection](/guides/international-ip-protection/). - **Trademark expansion**: additional classes, a **Madrid Protocol** filing for one or two key markets if you're going international early, and a **watch service** (a few hundred dollars a year) to catch confusingly similar applications. - **Counsel relationship.** This is the stage where ad-hoc legal help stops scaling; see [when to hire IP counsel](/guides/when-to-hire-ip-counsel/) for structuring it without enterprise overhead. Investors at this stage expect to see hygiene complete and core technology filed on — the checklist view is in [IP audit before you raise](/guides/ip-audit-before-you-raise/). ## Scaling stage: ~$50,000–$250,000 per year This is where budgets break, for two predictable reasons: **Reason one: foreign nationalization.** The PCT clock runs out at roughly **30 months from your earliest priority date**, and you must pick countries. Each national-phase entry means local filing fees, translations (often the largest single cost — thousands of dollars for Japanese, Chinese, or Korean), and local associate counsel. Entering five or six jurisdictions commonly costs **tens of thousands of dollars per patent family**, and then each country charges annual annuities forever after. The 30-month decision is the most expensive checkbox in IP — treat it as a board-level portfolio decision, not a docketing reminder. Default answer for most companies: **fewer countries than you think** — protect where you sell meaningfully, where competitors manufacture, and almost nowhere else. **Reason two: maintenance fees begin.** U.S. utility patents require fees at **3.5, 7.5, and 11.5 years after grant** — under the 2025 USPTO schedule, **$2,150, $4,040, and $8,280** at large-entity rates (small entities pay 40% of that). More on the triage discipline below. Also at this stage: **family strategy** (continuations that keep an application alive to chase competitors' design-arounds — powerful, but each continuation is another full prosecution budget), broader trademark coverage, and the beginnings of docketing/annuity infrastructure. ## Enterprise stage: pruning is the strategy Mature portfolios invert the problem: the marginal filing matters less than the **carrying cost of everything already granted**. A 500-patent portfolio can easily consume high six figures annually in U.S. maintenance fees and foreign annuities alone. The disciplines that matter: - **Annual portfolio review** mapping every asset to a current product, a competitor's product, or a licensing program — anything mapping to none of the three is a pruning candidate. - **Annuity management services**, which pay foreign renewals in bulk at lower per-payment cost than law firms. - **Monetization of the tail**: licensing or selling patents you'd otherwise abandon — see [how to license or sell a patent](/guides/how-to-license-or-sell-a-patent/) — recovers some sunk cost, though thin markets mean abandonment is often the honest answer. ## Where IP budgets actually die Across stages, the same three killers account for most blowups: 1. **Foreign filings** — covered above; the multiplier nobody models at filing time. 2. **Office-action churn.** Every rejection-response cycle costs low thousands in attorney time. Applications drafted with weak claims or filed into crowded art can rack up three, four, five rounds. The fix is upstream: search before filing and draft claims worth fighting for — see [patent search before filing](/guides/patent-search-before-filing/). 3. **Litigation reserves — or their absence.** Patent litigation defense routinely costs seven figures through trial. You don't budget to *fund* litigation at early stages; you budget to *avoid* it (clearance work) and you understand that IP insurance and contingency arrangements exist. A demand letter should never be the first time the CFO hears the word. ## Cost-control levers that actually work - **Flat-fee arrangements** with outside counsel for defined tasks (applications, responses, trademark filings) — converts an unpredictable line into a predictable one, and most firms will do it. - **Prioritized claims:** file fewer applications with better claims on the inventions that score highest on detectability and competitor appetite, instead of spreading budget thin. - **Provisional cascades:** use provisionals as cheap 12-month options; convert only the ones the business still cares about at month 11. Letting a provisional lapse costs almost nothing; abandoning a half-prosecuted nonprovisional wastes five figures. - **Abandonment discipline:** a written keep-or-kill review before every maintenance fee and annuity, not auto-pay. - **Annuity services** for foreign renewals once you have more than a handful. - **Entity-status vigilance:** claim small/micro entity discounts when eligible — and update status when you outgrow it, because falsely claiming the discount can render fees underpaid and jeopardize the patent. ### The maintenance-fee triage discipline The USPTO's escalating fee curve ($2,150 → $4,040 → $8,280) is an intentional re-justification mechanism. Use it as designed: - **At 3.5 years:** pay unless the patent maps to nothing — it's cheap and the technology's fate is often still unclear. - **At 7.5 years:** the real gate. Does this patent still read on anyone's product, yours or a competitor's? If no one would notice it expiring, let it. - **At 11.5 years:** pay only for patents you would plausibly **enforce or license** in the remaining term. At $8,280 (plus equivalent foreign annuities across a family), sentimentality gets expensive. ## The allocation heuristic: copyable and enforceable When ranking what to protect, one two-part test outperforms elaborate scoring models: **spend on what competitors would actually copy AND what you would actually enforce.** An invention nobody would copy doesn't need a patent; an invention you'd never sue over (or couldn't detect infringement of) doesn't reward one either — it may be better held as a trade secret or simply shipped. Everything in the budget should trace to a name-able competitor and a plausible enforcement or licensing story. ## What investors expect to see at each round - **Pre-seed/seed:** hygiene — assignments signed by everyone, trademark filed, provisional(s) on the core insight. Total historical spend of a few thousand dollars is completely respectable. - **Series A:** nonprovisionals on the core technology, a PCT if international markets are real, watch services, and evidence of *decisions* (a disclosure log, a filing rationale). - **Series B and beyond:** a portfolio that maps to the roadmap, foreign coverage in actual markets, maintenance-fee discipline, and no orphaned or encumbered assets. What kills diligence isn't a small budget — it's incoherence: patents unrelated to the product, missing assignments, or a lapsed core asset nobody noticed. Run the internal check before someone runs it for you: [IP audit before you raise](/guides/ip-audit-before-you-raise/) covers the pre-raise sweep, and [IP diligence for fundraising and M&A](/guides/ip-diligence-for-fundraising-and-ma/) shows the process from the money's side of the table. ## The bottom line An IP budget is a curve, not a number: low thousands for bootstrap hygiene, tens of thousands through seed as the first real filings land, potentially low-to-mid six figures while scaling — driven almost entirely by the 30-month foreign-filing decision and the onset of maintenance fees — and then a pruning discipline at maturity. Budget lifecycle costs at commitment time, claim your entity discounts, treat every renewal as a fresh decision, and let one heuristic allocate the dollars: protect what competitors would copy and you'd enforce. Companies that do this spend less than their peers and diligence better. --- *This article is general legal information for educational purposes only. It is not legal advice, does not create an attorney-client relationship, and may not reflect the most current law in your area. Government fees change and budget questions turn on specific facts. For advice about your situation, consult an attorney licensed in your jurisdiction.* ## Frequently asked questions ### How much should a startup spend on IP? A bootstrapped startup can cover the essentials for roughly $1,000–$5,000 a year: a federal trademark application (USPTO base fee $350 per class plus attorney help), copyright registrations at $45–$65 each, a provisional patent application if the technology warrants it, and clean assignment agreements. A funded seed-stage company typically lands in the $15,000–$50,000 per year range once it's prosecuting one to three nonprovisional patent applications and watching its trademarks. The right number depends less on stage labels than on whether competitors would copy what you're building. ### Why do foreign patent filings cost so much? Because you pay separately in every country: national filing fees, translation costs (often the single biggest line item — a full technical translation into Japanese or Chinese can run thousands of dollars per application), local associate attorneys in each jurisdiction, and then annual maintenance annuities in each country for the life of each patent. A PCT application defers the decision to about the 30-month mark, but entering national phase in five or six countries commonly costs tens of thousands of dollars per patent family — which is why the 30-month decision is where disciplined companies prune hardest. ### What are patent maintenance fees and when are they due? U.S. utility patents require maintenance fees at 3.5, 7.5, and 11.5 years after grant to stay in force. Under the USPTO fee schedule effective January 2025, the large-entity amounts are $2,150, $4,040, and $8,280 respectively — about $14,470 over a patent's life — with 60% discounts for small entities and 80% for micro entities. The escalating structure is deliberate: it forces owners to re-justify each patent as it ages. Treating each fee event as a keep-or-abandon decision, rather than an auto-pay, is one of the highest-leverage budget disciplines available. ### What do investors expect a company to have spent on IP? Investors care about coverage, not spend for its own sake. At seed, they expect hygiene: assignments from every founder, employee, and contractor, a trademark filing, and provisionals or early filings on core technology. At Series A and B, they expect a deliberate portfolio matched to the product roadmap, evidence someone is making protect-or-pass decisions, and no gaps a competitor or troll can exploit. A company that spent $30,000 thoughtfully often diligences better than one that spent $300,000 filing indiscriminately.