The estate of a beloved twentieth-century creator had a licensing program with real commercial momentum and a piracy problem expanding faster than the team could answer it. Unauthorised merchandise, unlicensed video, and a fast-growing tail of AI-generated derivatives were eroding margin on every legitimate licensee deal. The brief was straightforward to state and harder to execute: protect the legacy without smothering the fan culture that kept it alive.
A licensing program with a piracy tax
Before the engagement began, the estate ran a respected licensing program across apparel, home goods, publishing, and a small but profitable digital catalogue of remastered video. Revenue had grown for six consecutive years, but the in-house counsel team was spending most of its enforcement budget on reactive takedowns rather than program design.
The first audit told the story plainly. For every licensed unit sold through approved channels, the team estimated three to five unauthorised units circulating somewhere on the open or grey market. Some of those would never have been legitimate sales; many would have been. The program was paying a quiet tax that nobody had ever sized.
Leadership wanted three things in roughly this order: protect the integrity of the creator's voice, restore margin to existing licensees, and free in-house counsel to negotiate the next generation of partnerships rather than chase listings on weekend evenings.
Where the damage actually concentrated
We began by mapping abuse across the surfaces where the brand had measurable exposure: marketplaces (Amazon, Etsy, eBay, AliExpress), short-form video (TikTok and Reels), long-form video (YouTube and a handful of regional platforms), and the newer generative-AI surface where the creator's style and catchphrases were being prompted directly into image and video models.
Two patterns emerged inside a fortnight. On physical merchandise, five repeat sellers accounted for almost half of all unauthorised listings; on video, three uploader networks accounted for most of the infringing reach. That concentration is the single most useful finding any enforcement program can produce, because it converts an unbounded problem into a finite one.
The third pattern was the new one. Generative-AI derivatives were not yet large in absolute volume, but the growth curve was steep enough that the team could see where it would be in twelve months. The estate decided early that this surface needed a dedicated track rather than being bolted onto the existing workflow.
The enforcement architecture we built
For high-volume physical sellers, we moved from one-at-a-time takedowns to a tiered model. First-offence sellers received a standard notice and a short removal window. Repeat offenders triggered marketplace-level escalations. Persistent operators behind multiple storefronts were referred to platform trust and safety leads with documented patterns, and in two cases to local enforcement partners.
For video, the team graduated from per-clip DMCA filings to channel-level rights claims and monetisation recovery through Content ID and its equivalents. That change recovered a small but real revenue stream that had previously been leaking entirely to infringers and platforms.
For the generative-AI surface, we worked with two of the larger image platforms on name and style-based prompt rules. These are imperfect by design, but they raise the friction enough that casual misuse drops sharply, which is most of the volume. Tribute artwork, fan edits, and non-commercial creative work were left alone unless they crossed into impersonation or commerce — a distinction that was cultural as much as legal, and it mattered.
Outcomes after twelve months
By the end of the first year, unauthorised physical listings on the major marketplaces were down roughly seventy percent against the audit baseline. The estate's three largest licensees reported their cleanest fourth quarter in five years, with returns and customer service complaints related to counterfeit product falling in step.
On video, recovered ad revenue from monetisation claims covered roughly a third of the program's annual cost — not a primary goal, but a useful offset. Time spent by in-house counsel on reactive enforcement dropped by more than half, and the team redirected that capacity into two new licensing categories the estate had been deferring for two years.
The generative-AI track stayed small in revenue terms but proved its worth strategically. When a higher-profile incident hit a peer estate later in the year, the team already had documented relationships, escalation paths, and policy language ready to use.
What we learned worth carrying forward
Concentration is the friend of enforcement. In almost every brand protection program we run, the top handful of bad actors account for a disproportionate share of the damage. Naming and ranking them on day one beats any amount of broad-front activity.
Tone matters as much as policy. A heritage brand can over-enforce its way into looking petty very quickly, and the estate was clear from the start that it would rather miss a marginal infringement than alienate the community that kept the legacy meaningful.
Finally, the boring infrastructure pays off. A clean rights database, a single source of truth for approved licensees, and a documented escalation matrix sound like back-office hygiene; in practice they are what allow a small team to act decisively when something genuinely urgent hits.
Implications for other legacy brands
The estate's situation is unusual in its specifics but generic in its shape. Any rights holder with a long catalogue, a passionate fan base, and a growing licensing program will eventually face the same three-surface problem: physical commerce, video, and now generative AI. The right answer is rarely to spend more on enforcement; it is to spend more deliberately.
If we had to compress the lesson into one sentence, it would be this: enforcement programs work when they are designed to make the next month easier than the last, not when they treat every infringement as equally urgent.
Protecting a legacy brand is less about covering every surface equally and more about finding the two or three places where damage concentrates and closing them first. The rest of the program exists to keep those wins from quietly eroding while attention moves on.
