AI has broken a three-decade relationship.
Brand protection programs have tracked takedown volume for 30 years. Following that number made sense when enforcement meant placing humans in front of search results.
More people = more takedowns. That’s understandable math.
Now a single model can surface more counterfeit listings in a day than a full team used to find in a quarter. Which means the number keeps going up, and it tells you very little about whether your program is working.
Enter: the saturation rate.
On the Agenda
The Vision: Saturation rate is more important than takedown volume. Here’s why.
Word on the Street: The latest from the Patagonia lawsuit, and more.
Updates from HQ: Two new launches, plus two new pieces of industry thought leadership written by members of our team.
THE VISION
SATURATION RATE > TAKEDOWN VOLUME
Of the last 30 years of “standard” brand protection, Kaleigh Miller spent half of them watching brand protection get measured the same way — from inside Amazon, where she processed enforcement notices at global scale, to TikTok Shop, where she built the seller governance system from the ground up.
She's now at MarqVision, where she opens customer onboarding conversations with:
Your takedown count is going to drop. That's not bad news.
In the manual enforcement era, volume was a reasonable proxy for effort. More people running searches meant more enforcement. These days, with AI, a model can surface far more volume than any human team, which means chasing raw counts now means optimizing for the easiest targets rather than what might be the more consequential ones.
"Even though AI, in reality, is much faster and can probably find significantly more volume, we don't actually think that that's the right thing to do," Kaleigh says.
"The right thing to do is to find the most impactful listings, and to do that really well."
That's where saturation rate comes in.
Of the 10 results a customer sees when they search for your product on Amazon right now, how many are counterfeit?
Getting that number to zero (on a highly visible and impactful marketplace such as Amazon) is a different goal than running up an overall online takedown count — and it's the goal that maps to the customer's actual experience.
The transition isn't painless. When brands switch to AI-native enforcement, their takedown count often falls sharply while the model trains on their specific IP. A brand used to seeing 1,000 listings enforced per week might see 250 at the start. To a C-suite audience expecting continuous improvement, that reads as regression.
Miller's framing is that it doesn't have to be either/or.
"We can have saturation rate as the primary metric," she says. "We're happy to also share volumes so you can measure the percentage of takedowns. Both need to be able to see that value."
Saturation rate tells you whether the program is working and whether the market is actually getting cleaner, all by measuring and prioritizing the removal of the most high-impact listings. Volume simply tells you how active the program is, not where it’s making a difference.
Both numbers have a role. They're just not equal, and for 30 years, the industry has been treating the lesser one as the headline.
Learn more about MarqVision’s approach to brand protection (and get a free report) here.
WORD ON THE STREET
Your line of sight into what’s happening right now in the brand protection industry.
AI and cybersecurity claims are reshaping what in-house counsel need to worry about in 2026. Trademark exposure is increasingly in the mix. (World Trademark Review)
The Patagonia v. Pattie Gonia case is now headed to mediation. The enforcement principles it surfaces are worth understanding regardless of how it resolves. (Law.com)
The European Commission passed amendments to the EU AI Transparency Act this week, in part effectively attempting to curb AI imagery that is similar enough to copyrighted imagery to be deceptive, but not so similar to count as infringement across many platforms. While these amendments allow European brands to better protect against fraudulent AI imagery, it's up to marketplaces to interpret and respond to these infractions. As requirements evolve and shift globally, having strategic brand protection (that also cares about deep partnerships with marketplaces and platforms, as MarqVision does) is key. (European Commission)
UPDATES FROM HQ
Your workflow just got a whole lot easier. We recently launched the Marq AI Slack App, allowing you to enforce takedowns and pull performance reports right there in the program you’re already running. Learn more here.
Speaking of launches, we also just announced MarqRadar, a free brand risk scanning tool that scans across five channels. Find details here.
When it comes to the industry, our team is best in class. BP Strategy Consultant Conseulo Michetti and Director of Global Partnerships Martina Clochiatti recently shared their expertise in pieces for the World Trademark Review’s 19th annual guide. Read them each below:
Is your program asking "how many counterfeits did we remove?" or "what does a customer see when they search for my product right now?"
Those are different questions, and they require different numbers. The latter holds up in a C-suite conversation.
-
Thanks for reading, we’ll see you next time.
The MarqVision Team