Font and Design Asset Checklist: Inventory, Licence Scope, Embedding and Web Use, Audit Response, and Substitution Planning

By Casey Scott McKay · 2025-12-30

This checklist builds a licensed asset position from the only starting point that works: an inventory of what the organisation actually serves, embeds, installs, uses in its logo, and runs through rendering pipelines. Phase one builds that register. Phase two joins it to licences and records the limits that will be exceeded — seats, web tiers, domains, print runs, territories, and terms. The middle phases cover open font obligations that apply where no fee was paid, stock licence restrictions and release scopes, and the contractor and agency arrangements through which most unlicensed assets enter an organisation. The closing phases handle demand letter response, enterprise negotiation, substitution planning, generated assets, audio, and the purchasing route that keeps the register current after the first effort.

IP and Technology > Copyright | Checklist | Published 30 December 2025 - Updated 29 January 2026 | Casey Scott McKay - marksy.us

Summary. This checklist builds a licensed asset position from the only starting point that works: an inventory of what the organisation actually serves, embeds, installs, uses in its logo, and runs through rendering pipelines. Phase one builds that register. Phase two joins it to licences and records the limits that will be exceeded — seats, web tiers, domains, print runs, territories, and terms. The middle phases cover open font obligations that apply where no fee was paid, stock licence restrictions and release scopes, and the contractor and agency arrangements through which most unlicensed assets enter an organisation. The closing phases handle demand letter response, enterprise negotiation, substitution planning, generated assets, audio, and the purchasing route that keeps the register current after the first effort.

Keywords: font asset checklist · web crawl inventory · endpoint font report · embedded font enumeration · logo dependency · rendering pipeline use · seat and tier limits · application embedding licence · open font reserved names · stock licence restrictions · extended licence triggers · model and property release scope · contractor deliverable warranty · agency sublicensing · demand letter response · enterprise licence terms · substitution planning · generated asset provenance · audio library metering · purchasing route

How to use this checklist

| Field | Detail | |---|---| | Who runs it | Counsel with design, brand, engineering, procurement, and endpoint management | | When | On receipt of any demand; before a rebrand; annually | | Time required | Two to three weeks for a first pass; the crawl itself takes a day | | Gates | Register complete before responding to any demand; limits recorded against actual use | | Output | An asset register, a purchasing policy, a substitution plan, and a diligence-ready file | | Companion documents | Managing Fonts, Stock Assets, and Design Libraries and The Letters You Licensed |

The matter. A consumer business with a main site, six campaign microsites, two acquired brand properties, a mobile application, and a customer document generation service has received a letter from a foundry asserting unlicensed use of one of its typefaces across the web estate, the application, and the company's logo, and demanding a retrospective fee plus a forward enterprise licence. The company holds three desktop licences bought by a designer who left two years ago. The logo was drawn in the same face and is registered. Nobody has crawled the estate. The design team also uses stock imagery bought on a personal subscription by a freelancer, and one image appears on merchandise.

Phase 1. Build the register

[ ] Crawl every web property for served fonts and assets. Main site, microsites, campaign pages, acquired properties, support sites, and anything on a subdomain. Why. Web fonts are visible in the page source, which is exactly how the foundry found them. Trap. Crawling the main site only.
[ ] Run an endpoint report for installed fonts. Why. It is the seat number a foundry will assert and the one companies cannot answer. Trap. Counting design team machines rather than the estate.
[ ] Enumerate embedded fonts in shipped applications, mobile apps, games, and distributed documents. Why. Embedding distributes the font software itself and usually needs a separate licence. Trap. Treating an application as a use of the designer's desktop licence.
[ ] Identify the logo dependency. Why. Logo use is separately treated in many foundry licences, particularly where the logo is registered. Trap. Assuming the artwork's ownership answers the licence question.
[ ] Find rendering pipeline uses. Document generation, personalisation, print-on-demand, email rendering. Why. These are font uses and are almost never licensed. Trap. Classifying a document service as software rather than as a font use.
[ ] Extend the same exercise to stock, icons, templates, and audio. Why. They arrive through agencies, contractors, subscriptions, and individual purchases. Trap. An inventory that covers fonts and stops.
[ ] [Gate] Do not respond to any demand before the register is complete. Why. A denial contradicted by the company's own page source is expensive. Trap. A deadline in the demand letter treated as binding.

Phase 2. Join assets to licences and record the limits

[ ] Record per asset: source, licence, permitted uses, limits, expiry, actual use. Why. The register's value is in the join. Trap. A list of assets with no licence column.
[ ] Record seat limits and compare to the endpoint report. Why. Growth, reorganisation, and virtual desktop deployment break seat licences silently. Trap. A seat count from the purchase date.
[ ] Record web tiers and permitted domains and compare to traffic. Why. Exceeding a page view tier or adding a domain is the most common web breach. Trap. A tier appropriate at launch and exceeded by an order of magnitude.
[ ] Record application embedding permissions per title. Trap. One licence assumed to cover multiple applications.
[ ] Record logo permission expressly. Trap. A registered wordmark drawn from a face with no logo licence.
[ ] Record document embedding permissions, distinguishing subset-for-viewing from editable full embedding. Trap. Editable embedding in a widely distributed document.
[ ] Record modification rights, noting that modification frequently ends support and updates. Trap. A face modified for language support with no permission.
[ ] Check bundled licences arriving with operating systems, design subscriptions, and printer drivers. Why. They typically permit use only within that product. Trap. Conceding a use that a bundled licence in fact permits.

Phase 3. Open licences and provenance

[ ] Record open licensed fonts with their conditions. Why. Free is not unconditional: no sale of the software alone, renaming where the name is reserved, and notice travelling with redistribution. Trap. Open fonts omitted from the register because no fee was paid.
[ ] Check reserved font names against any modification. Why. A subsetted, patched, or extended open font may need renaming before distribution. Trap. A subset generated by a build tool and shipped under the original name.
[ ] Treat bundling into an application as redistribution and satisfy the notice obligations. Trap. Notice obligations satisfied in a licence file nobody ships.
[ ] Treat self-hosted web delivery as distribution with the same obligations. Trap. Notice omitted because the font is "just on the server".
[ ] Verify provenance for downloaded fonts. Why. Aggregator sites host fonts uploaded by parties with no right to distribute them, and a bundled licence file proves nothing. Trap. A commercial face obtained free and used in good faith.

Phase 4. Stock assets and releases

[ ] Classify each licence as rights-managed or royalty-free and record the restrictions. Why. Rights-managed licences are use-specific and expire; royalty-free licences carry print run, resale, standalone-use, logo, and sensitive-subject restrictions. Trap. Royalty-free read as unrestricted.
[ ] Identify uses requiring an extended licence. Merchandise resale, unlimited print runs, sublicensing to clients. Trap. An image on merchandise under a standard licence.
[ ] Apply a standing prohibition on stock assets in logos and marks. Why. Nearly every licence prohibits it. Trap. A mark filed incorporating a licensed illustration.
[ ] Respect editorial-only designations. Trap. An unreleased image cleared for commentary used in advertising.
[ ] Check release scope, not merely existence. Why. A model release permitting general commercial use may exclude health, political, adult, and endorsement contexts; property releases cover recognisable private property and artwork in frame. Trap. "Model released" treated as a complete answer.
[ ] Assess visible third-party marks in imagery separately from the release. Trap. A product or logo in frame in an advertisement suggesting affiliation.
[ ] Read the provider indemnity's limits. Caps, exclusions for modified assets, out-of-licence uses, and notification conditions. Trap. Modification that voids the cover.

Phase 5. Contractors, agencies, and acquisitions

[ ] Ask who bought each asset. Why. A large share are bought by people who are not the ultimate user. Trap. A register that records the asset and not the purchaser.
[ ] Check whether agency licences permit sublicensing. Why. They usually do not, leaving the client unlicensed and the agency in breach. Trap. A campaign delivered and reused for years.
[ ] Check freelancer subscriptions. Why. Personal subscriptions permit client work only within stated limits, and the deliverable may carry assets the client cannot use. Trap. A brand built on a designer's personal licence.
[ ] Treat deliverables as containing the assets. Why. A layout file contains the fonts and images used to build it, and passing them on may be an unpermitted distribution. Trap. Source files handed over as a courtesy.
[ ] Add the contractor clause. Identification of every third-party asset used, confirmation the licence permits the client's intended use, and a warranty that the deliverable does not require an unheld licence. Trap. An intellectual property clause covering only originally created material.
[ ] Request a substitution list at concept stage. Trap. A brand launched on a face the client cannot licence economically.
[ ] Put asset licences into acquisition diligence. Why. An acquired design estate is built on licences that do not transfer. Trap. A post-closing discovery.

Phase 6. Responding to a demand

[ ] Acknowledge, do not concede, and request the evidence. Which font, which uses, which pages or files, and the basis for the counts asserted. Why. Demands are frequently overstated and occasionally misidentify the font. Trap. An admission in the first reply.
[ ] Complete the register first. Trap. Responding within the letter's deadline.
[ ] Sort into three categories. Covered by an existing licence; covered but exceeded; unlicensed. Why. The negotiation differs entirely for each. Trap. One answer for the whole demand.
[ ] Check bundled and open licences before conceding any use. Trap. Paying for a use already permitted.
[ ] Price the substitution. Why. What replacement costs and how long it takes determines the negotiating position. Trap. Negotiating without knowing the alternative.
[ ] Negotiate retrospective and prospective together. Why. Willingness to buy forward is the leverage on the past. Trap. Settling the past and then negotiating the future separately.
[ ] Fix the purchasing process in the same exercise. Trap. A second letter in three years.
[ ] Record the settlement scope against the register. Why. The next demand will come from a different foundry about a different face. Trap. A settlement filed with the litigation papers and never linked to the assets.

Phase 7. Enterprise terms and substitution

[ ] Ask for one agreement covering every use with one renewal date. Trap. Seven purchases with seven expiries.
[ ] Ask for headcount rather than seat counts. Why. It survives reorganisations and virtual desktop deployment. Trap. A seat licence in an organisation that changes shape annually.
[ ] Ask for affiliate and contractor coverage. Trap. An acquisition creating an instant breach.
[ ] Ask for perpetual rights in what has shipped. Why. A lapse that renders published applications and documents unlicensed is commercially unacceptable. Trap. A pure subscription with no perpetual carve-out.
[ ] Ask for audit terms with notice, scope limits, and a cure period. Trap. An unlimited audit right.
[ ] Identify a metrics-compatible alternative for each critical face. Trap. Researching substitutes under time pressure.
[ ] Design the system for substitution. Style tokens, centralised declarations, and templates referencing a variable rather than a font name. Trap. A font name hard-coded across hundreds of templates.
[ ] Convert the logo to owned outlines. Why. It removes a permanent dependency in an afternoon. Trap. A registered wordmark that still depends on a live licence.
[ ] Consider a bespoke face where the dependency is deep. Trap. Enterprise licensing over a decade costing more than commissioning would have.

Phase 8. Generated assets, audio, and the purchasing route

[ ] Record generated assets separately. Service, terms version, date, human contribution, screening performed. Why. There may be no copyright to acquire, which means no exclusive right and no ability to stop identical use. Trap. A logo or signature visual generated and assumed to be owned.
[ ] Read the provider's ownership, commercial use, indemnity, and training data terms, and note the indemnity's conditions. Trap. An indemnity voided by modification.
[ ] Screen output against third-party rights. Why. Output resembling an existing work, a recognisable person, or a protected mark raises the same claims. Trap. Absence of copyright treated as absence of risk.
[ ] Meter audio by media, territory, term, and platform. Why. A track licensed for internal training is frequently not licensed for a campaign, and platform audio libraries are licensed for that platform only. Trap. Content repurposed across channels with the audio intact.
[ ] Clear composition and recording separately for commercial music, and both for samples. Trap. A synchronisation licence treated as a master licence.
[ ] Route all asset purchases through one channel with a required intended-use field. Why. The register decays unless the purchasing route captures the same information. Trap. A register built once and never fed.
[ ] Publish a short decision guide and a prohibition list for designers. Trap. A policy document nobody reads in place of a one-page list.
[ ] Require a licence reference on every asset in the shared design system. Trap. An unlabelled asset used hundreds of times.
[ ] Re-run the crawl annually and review at renewal. Trap. Subscription lapses discovered after they have left work in an ambiguous position.

Phase 9. Working the example matter

[ ] Crawl all nine web properties before replying. Main site, six microsites, and two acquired properties. Why. The foundry has already done this and its evidence is the company's own page source; the company needs the same picture. Trap. Replying on the basis of the three desktop licences it knows about.
[ ] Establish what the three desktop licences actually permit. Why. They almost certainly cover neither web delivery nor application embedding nor logo use, but they may cover more than the foundry's demand assumes. Trap. Conceding all three uses because the licence count is obviously insufficient.
[ ] Treat the document generation service as a separate use. Why. A rendering pipeline producing customer documents uses the font on a server, and neither a desktop nor a web licence covers it. It is also, on the foundry's likely evidence, visible in the generated output. Trap. Omitting it from the audit because nobody thought of it as a font use.
[ ] Deal with the logo dependency now. Why. The wordmark is registered and drawn from the face, which is the single most expensive element of the demand and the one with the longest tail. Establish whether the licence permits logo use, and if not, negotiate it as part of the enterprise agreement rather than as a separate concession. Trap. Redrawing the logo under pressure.
[ ] Price the substitution across the estate. Nine sites, one application, one document pipeline, and a registered logo. Why. That number is the company's entire negotiating position. Trap. Negotiating without it.
[ ] Trace the departed designer's purchases. Receipts, licence keys, and the accounts they were bought under. Why. Licences bought on a personal account may not be transferable, which changes the analysis. Trap. Assuming the company holds licences it cannot evidence.
[ ] Address the acquired properties separately. Why. Assets on acquired sites were licensed by the acquired business, and those licences did not transfer. Trap. Treating the whole estate as one licensing history.
[ ] Deal with the freelancer's stock imagery and the merchandise use. Why. A personal subscription almost certainly does not permit client work on merchandise, and merchandise use requires an extended licence under nearly every stock licence. This is a second, independent exposure. Trap. Focusing on the font demand and missing the image on the product.
[ ] Negotiate one enterprise agreement covering everything. Desktop, web across all domains, application embedding, server rendering, logo use, affiliate and contractor coverage, and perpetual rights in what has shipped. Why. It resolves the demand and prevents the next one. Trap. Buying the specific licences the demand identified and leaving the rest unresolved.
[ ] [Gate] Fix the purchasing route in the same exercise. Why. The company's asset estate was built by a designer who left, a freelancer with a personal subscription, and two acquisitions. Without a route, it will rebuild itself the same way. Trap. A settlement with no process change.

Phase 10. Type in the brand system

[ ] Separate the mark from the font. Why. A wordmark is a trademark under 15 U.S.C. § 1051 whatever typeface drew it, and the artwork can be owned outright while the software remains licensed. Trap. Conflating ownership of the logo with a right to the letterforms.
[ ] Read the logo term in the licence specifically. Why. Some foundries permit logo use, some require an additional licence particularly where the logo is registered, and a few prohibit it. Trap. An assumption in either direction.
[ ] Check whether a design patent covers the face under 35 U.S.C. § 171, with 35 U.S.C. § 289 supplying the remedy. Why. It constrains competitors' imitations and the brand's own modifications. Trap. Modifying a face covered by a live design patent.
[ ] Record modification permissions. Why. Modification for a logo or for language support is permitted by some licences and prohibited by others, and frequently ends support. Trap. A modified face that can no longer be updated.
[ ] Build trade dress evidence where the brand system is distinctive, applying Wal-Mart Stores v. Samara Brothers and, for colour, Qualitex v. Jacobson Products. Trap. Evidence assembled during litigation.
[ ] Consider a bespoke face where the dependency is deep. Why. Owning the software outright removes the question permanently and produces a distinctive asset. Trap. A decade of enterprise licensing costing more than commissioning.

Phase 11. Documents this checklist should produce

[ ] An asset register joining every served, embedded, installed, logo, and pipeline asset to a licence and its limits, held in a system the business already uses. Trap. A spreadsheet on a legal shared drive.
[ ] A purchasing policy and prohibition list, one page, where designers work. Trap. A policy document in a compliance portal.
[ ] A contractor clause requiring identification of third-party assets and a warranty that the client can use the deliverable. Trap. An intellectual property clause covering only originally created work.
[ ] A demand response process beginning with acknowledgement and an audit rather than a substantive reply. Trap. A process written after the first letter.
[ ] A substitution plan identifying metrics-compatible alternatives for each critical face, with the estate cost estimated. Trap. A plan that lists alternatives without testing them.
[ ] A diligence-ready file containing the register, the licences, the settlement scopes, and the crawl outputs. Trap. A file assembled when the term sheet arrives.

Phase 12. Special situations

[ ] Publishing and long-form documents. Check embedding permission bits alongside the licence text, and treat print-on-demand as a rendering use at the printer. Trap. Editable full embedding in a widely distributed publication.
[ ] Broadcast and out-of-home. Some foundry terms require separate permission for broadcast, film, and large-format display. Trap. An agency treating a desktop licence as covering everything a designer produces.
[ ] Signage and physical products. A typeface cut, moulded, or etched is a physical rendering rather than a software distribution, and licences vary on whether anything beyond a desktop licence is required. Trap. An assumption in either direction on a large signage rollout.
[ ] Regulated typography. Where a document's typography is prescribed by regulation — labelling, financial disclosures, accessibility standards — the face may be constrained and the licence must cover the required rendering method. Trap. A brand face that cannot lawfully be used on a regulated label.
[ ] Accessibility requirements. Minimum sizes, contrast, and in some contexts specific font characteristics interact with brand typography and occasionally require a second licensed face. Trap. An accessible variant used without a licence.
[ ] Multilingual expansion. Adding Cyrillic, Greek, Arabic, or CJK support frequently requires additional families or extended licences, and CJK faces are licensed and priced quite differently. Trap. A market launch blocked by a font licence.
[ ] Events and short activations. Temporary installations tempt teams to use fonts without licences and are highly visible to scanning. Trap. A one-week activation producing a permanent record.

Phase 13. Audio and time-based assets

[ ] Record audio in the same register. Asset, source, licence, permitted media, territory, term, actual use. Trap. Audio omitted because it is bought by a different team.
[ ] Check media, territory, term, and audience metering on production music. Why. A track licensed for internal training is frequently not licensed for a campaign or a broadcast spot. Trap. Content repurposed with the audio intact.
[ ] Check redistribution prohibitions on sound effects libraries. Why. They catch anyone incorporating effects into a product or application. Trap. Effects shipped inside software.
[ ] Treat platform audio libraries as platform-specific. Trap. A platform-supplied track reused on a company website.
[ ] Clear composition and recording separately for commercial music, and both for samples. Trap. A synchronisation licence treated as a master licence.
[ ] Record performer terms for voice content, including reuse scope and any synthetic voice permission. Trap. An archived recording used to train a model under an old buyout.

Phase 14. Governance and cadence

[ ] Name an owner. Why. Asset licensing is not usually anyone's job, which is why it fails consistently. Trap. Shared responsibility across procurement, design, and legal.
[ ] Route purchases through one channel with an intended-use field. Why. The register decays unless the purchasing route captures the same information. Trap. A register maintained manually alongside an unchanged purchasing habit.
[ ] Build a pre-cleared library. Faces, images, icons, and audio licensed broadly and available without further approval. Why. Most requests are satisfied from it and the exceptions become manageable. Trap. A library with no maintenance owner.
[ ] Make the exception route same-day. Trap. A procurement cycle that designers route around.
[ ] Require a licence reference on every asset in the shared design system. Trap. An unlabelled asset used hundreds of times.
[ ] Report three measures. Assets with a matched licence; limits approaching breach — seats, tiers, print runs; unresolved provenance items. Trap. Reporting spend.
[ ] Re-run the crawl annually and review subscriptions at renewal rather than at expiry. Trap. A lapse that leaves published work ambiguous.
[ ] Keep the register where the business already works, with the licences attached and the terms version and date recorded. Trap. A register pointing at a contract repository nobody can search.

Phase 15. Proportion

[ ] Do four things regardless of scale. Crawl the web estate, record the four licence terms that matter, prohibit stock assets in marks, and add the contractor clause. Why. Together they take days and prevent most of what goes wrong. Trap. A comprehensive programme that is abandoned.
[ ] Concentrate on the critical faces and the logo. Why. That is where the dependency and the cost sit. Trap. Equal effort across a library of two hundred faces, most of which are unused.
[ ] Assign owners and dates to every open item. Trap. A report circulated for information.

Phase 16. Interface kits, icons, and code assets

[ ] Treat icon sets like fonts. Why. They are frequently licensed as font files with the same seat, embedding, and application distinctions. Trap. An icon font shipped in an application under a desktop licence.
[ ] Check redistribution terms on interface kits and templates. Why. Use as the basis of a product sold to others is the recurring breach. Trap. A template licensed for internal use forming the foundation of a client deliverable.
[ ] Check sharing permissions on design tool component libraries. Why. They are licensed to a team or organisation, and sharing a file containing them with a client or contractor may exceed the licence. Trap. A design file sent to an agency.
[ ] Record per-project and per-title terms on three-dimensional models and motion assets, including broadcast and product prohibitions. Trap. An asset licensed for a pitch used in a campaign.
[ ] Include code assets and snippets downloaded into products, with their licences and any copyleft obligations. Why. They reach the product they are compiled into. Trap. An open source review that covers dependencies and misses copied snippets.
[ ] Extend the purchasing route to engineering. Why. Code and component assets are acquired by a different team than design assets, and the route must capture both. Trap. A policy addressed only to designers.

Phase 17. Diligence readiness

[ ] Keep the register diligence-ready rather than assembling one. Why. Buyers now ask, and a business without a register is diligenced by crawling its own website. Trap. Assembly beginning at term sheet.
[ ] Be able to answer the font position. Faces, licences, limits, and whether the logo depends on a licensed face without a logo licence. Trap. A list of purchases in place of a position.
[ ] Be able to answer the stock position. Assets in current campaigns, extended licences where required, and any asset in a mark or on merchandise. Trap. An answer covering the current campaign only.
[ ] Be able to answer the release position by scope. Trap. "Model released" offered as the complete answer.
[ ] Be able to answer the contractor position. Whether deliverables were built on contractor subscriptions. Trap. An assumption that agency work is clean.
[ ] Be able to answer the open source and generated asset positions. Reserved names, notice obligations, provider terms, and whether the business holds exclusive rights in generated material at all. Trap. Generated assets presented as owned.
[ ] Price known gaps and disclose them. Why. A quantified gap with a remediation plan is priced; an undisclosed one discovered in diligence is priced worse. Trap. A warranty offered in place of evidence.

Phase 18. Questions the business will ask

[ ] "We bought the font, don't we own it?" No — the purchase licensed software on terms metering installations, web delivery, embedding, and frequently logo use separately. Trap. An answer that stops at "no" without explaining what to do about it.
[ ] "It came with our design software." Then it carries that product's terms, typically permitting use only within that product. Trap. A bundled font on a public website.
[ ] "It's free." Free is not unconditional. Reserved names, notice obligations, and redistribution conditions apply. Trap. Open fonts omitted from the register.
[ ] "The agency handled it." Ask whether the agency's licence permits sublicensing. It usually does not. Trap. A campaign asset reused for years.
[ ] "Can we put this image in our trademark?" Almost certainly not, under nearly every stock licence. Trap. A mark filed before anyone asked.
[ ] "We generated it, so it's ours." Possibly not, because copyright requires human authorship and an asset with no author confers no exclusive right. Trap. A brand-defining visual with no protection and no exclusivity.
[ ] "What is our biggest exposure?" The fonts nobody inventoried, followed by the assets bought by people who are not the ultimate user. Trap. An answer framed as a legal risk rather than as a purchasing failure, which is what it is.

Phase 19. Working with the design team

[ ] Start from their problem. Designers want the right face quickly, without a procurement queue, and without discovering later that they cannot use it. Why. An enterprise licence and a pre-cleared library solve their problem and the organisation's simultaneously. Trap. A policy presented purely as a restriction.
[ ] Explain the logo rule specifically. Why. It produces the most expensive mistakes and is the least intuitive. Trap. A general briefing that never mentions it.
[ ] Give them the substitution list. Why. A designer who knows which faces are available and which are problematic specifies accordingly, and specification creates the dependency. Trap. A list held by legal.
[ ] Do not treat asset acquisition as a compliance failure. Why. Designers buying fonts are doing their job; the failure is a purchasing route that does not capture what was bought and for what. Trap. A remediation exercise framed as an investigation.
[ ] Report back on what the register prevented. Trap. A process that produces obligations and never visibly produces benefits.

Phase 20. Budget

[ ] Free. The crawl, the endpoint report, the purchasing policy, the prohibition list, the contractor clause, and reading the four licence terms that matter. Trap. Skipping these because they produce no invoice and therefore no urgency.
[ ] Low cost. Converting the logo to owned outlines, identifying substitution candidates, adding a licence reference field, and registering the marks that matter. Trap. Deferring the logo conversion indefinitely.
[ ] Moderate. An enterprise font agreement, extended stock licences where required, and a bespoke face for a deep dependency. Trap. Buying the specific licences a demand identified rather than the agreement that prevents the next one.
[ ] Expensive. A retrospective settlement negotiated without an audit, a substitution project under time pressure, and a rebrand because the logo depended on an unlicensable face. Trap. Reaching this category because the free items were skipped.

Phase 21. Three closing tests

[ ] The traffic test. What typeface does the main website serve, under what licence, with what page view tier, and how close is current traffic to it. Why. Four questions answerable in minutes with a register and by nobody else, and the fourth decides whether a demand letter is coming. Trap. An answer that requires asking the agency.
[ ] The library test. Open the shared component library and pick an asset at random. If it has no licence reference, every subsequent use of it is unmanaged, and there will be hundreds. Trap. A library curated for quality and never for rights.
[ ] The logo test. Would the logo survive the loss of the font licence it was drawn from? Why. Converting to owned outlines is an afternoon's work standing between the brand and a permanent dependency on somebody else's commercial terms. Trap. A registered wordmark still tied to a live subscription.

[ ] And one for the acquisition pipeline. Add asset licences to the standing diligence request list, so that the next acquisition arrives with its font and stock position disclosed rather than discovered. Why. Acquired estates are the second largest source of unlicensed assets after departed employees. Trap. A diligence list that covers registered rights and omits licensed ones.

Outcome. A register joining every served, embedded, installed, logo, and pipeline asset to a licence and its limits; open licence obligations recorded; stock restrictions and release scopes captured; a contractor clause in the standard engagement; a demand response process that begins with an audit; enterprise terms negotiated before a letter arrives; a substitution plan for the critical faces; and a purchasing route that keeps the register current.

The five things people get wrong

One. Responding to the demand before the audit. The letter arrives with a deadline designed to prevent exactly the inventory that should happen first, and a denial contradicted by the company's own page source destroys the negotiating position permanently. The correct first response acknowledges receipt and asks for the evidence.

Two. Forgetting the rendering pipeline. Document generation, personalisation, print-on-demand, and email rendering all use fonts, and almost nobody licenses them for it because nobody thinks of a document service as a font use. It is one of the two categories a foundry finds most reliably.

Three. Treating the logo as settled. A registered wordmark drawn from a licensed typeface depends on a licence term that many foundries treat separately and some prohibit. Converting the logo to owned outlines takes an afternoon and removes the dependency permanently, and almost no organisation has done it.

Four. Assuming free means unconditional. Open licensed fonts carry reserved font names, notice obligations, and redistribution conditions. An estate assembled from several open sources has assembled several sets of obligations, none of which is in anybody's register.

Five. Letting people who are not the user do the buying. Agencies that cannot sublicense, freelancers using personal subscriptions, and acquired businesses whose licences did not transfer account for most unlicensed assets in most organisations. It is fixed by a clause in the engagement terms and a line in the diligence request list.

Key Authorities at a Glance

| Authority | Proposition | |---|---| | 17 U.S.C. § 101 | Definitions; computer program | | 17 U.S.C. § 102 | Software as literary work | | 17 U.S.C. § 103 | Derivative works; scope | | 17 U.S.C. § 106 | Exclusive rights | | 17 U.S.C. § 107 | Fair use | | 17 U.S.C. § 109 | First sale | | 17 U.S.C. § 117 | Copies made in using a program | | 17 U.S.C. § 411 | Registration precondition | | 17 U.S.C. § 504 | Statutory damages | | 15 U.S.C. § 1051 | Registration of marks | | 15 U.S.C. § 1052 | Refusals | | 15 U.S.C. § 1114 | Infringement | | 15 U.S.C. § 1125 | False designation | | 35 U.S.C. § 171 | Design patents | | 35 U.S.C. § 289 | Total profit remedy | | Feist Publications v. Rural Telephone Service | Originality | | Google LLC v. Oracle America | Functional elements and fair use | | Star Athletica v. Varsity Brands | Separability | | Wal-Mart Stores v. Samara Brothers | Product design | | Qualitex v. Jacobson Products | Non-traditional marks | | Andy Warhol Foundation v. Goldsmith | Purpose and character | | Typeface design protection | Design and software distinction | | Open font licence conditions | Open licensing obligations | | Font licence audits | Enforcement practice | | Stock licence restrictions | Asset licence scope | | Model and property releases | Release requirements | | Right of publicity elements | Likeness claims |