Brand Protection Guides

You can't defend a budget line you can't measure.

Infringement shows up in your numbers long before anyone files a takedown — in branded CPC, in conversion rate, in return reasons. How to find it there, and what to track once enforcement is running.

November 16, 2026· 9 min read·EnforceShield Team

TL;DR

  • Infringement is usually budgeted as legal spend because nobody has sized it. That framing caps the budget and hides the return.
  • You do not need a vendor's estimate to start. Branded search CPC, product-page conversion, return and complaint reasons, and marketplace Buy Box share already carry the signal.
  • Most published counterfeit-loss figures are global trade estimates. They are useless for a single brand's business case — build the number from your own data instead.
  • Once enforcement runs, four metrics tell you whether it works: time from detection to filing, removal rate, recurrence rate, and queue direction.
  • Beware the model where the bill rises with infringement volume — it prices enforcement highest exactly when you need it most.

Brand protection tends to sit in the budget next to trademark renewals and outside counsel. It gets reviewed once a year, defended weakly, and capped.

That happens for a specific reason: nobody has ever put a number on it. Marketing can show cost per acquisition. Operations can show fulfillment cost per order. Brand protection shows a count of takedowns, which answers a question finance never asked.

The work of this guide is to fix that — to size the leak with data a brand already has, and then to name the small set of metrics that show whether enforcement is actually working.

Skip the Global Statistics

The first instinct is to reach for a published figure. Global counterfeit trade in the hundreds of billions. Some percentage of world trade. Those numbers are real, and they are useless to you.

They describe cross-border trade in counterfeit goods across every category and every market. Your exposure depends on your category, your price point, how visible your marketing is, and which channels you sell on. Two brands with the same revenue can be an order of magnitude apart.

Worse, a borrowed statistic is easy for a CFO to dismiss, and dismissing it usually takes the whole argument with it. Build the number from your own data instead. It will be smaller and far harder to argue with.

Where the Leak Is Already Visible

Infringement shows up in operating metrics long before anyone files a takedown. Five places are worth checking.

Paid search. If cost per click on your own brand terms has been drifting up without a change in your bidding, somebody is competing with you for your own name. Impression share on branded terms tells the same story from the other side. This is the cleanest signal available, because branded search is otherwise stable.

Product-page conversion. Look for pages where traffic has held steady but conversion has softened. That pattern is consistent with buyers leaving mid-journey — comparison-shopping their way to a cheaper listing that looks like yours. It is not proof on its own; it is a place to look.

Customer service. Return reasons are an underused source. "Not as described", quality complaints about a formulation you did not change, warranty claims on units whose batch codes do not match anything you shipped. Support tickets asking whether a site is really yours. These are counterfeit units reaching real customers, and your support team already has the records.

Marketplaces. Buy Box share and the seller names appearing on your own listings. A falling share against unfamiliar sellers is the most direct measurement of diverted revenue you will find, because the unit is a sale you would otherwise have made.

Reviews. A cluster of negative reviews describing a product you do not make is counterfeit units pooling into your listing's rating. The damage here compounds, because the rating affects every future buyer.

None of these is conclusive alone. Together, and moving in the same direction over a quarter, they are a defensible business case.

Building the Number

Take each signal, value it conservatively, and show your work.

Branded search is the easiest: the CPC increase multiplied by branded clicks gives you an annual figure for what interception costs in media alone. Lost Buy Box share multiplied by the units that typically move through it gives a revenue figure. Complaint and return volume that traces to product you did not make gives both a refund cost and a support cost.

Two rules keep the number credible. State every assumption in the open, and present a range rather than a point estimate. A range that a skeptical finance lead can narrow is more persuasive than a precise figure they can reject outright.

Then add the part that does not appear in any of those lines: the newest products, which are usually the most exposed and the least covered, because the registration gap leaves new SKUs undefended for months after launch.

The Four Metrics That Matter Afterwards

Once enforcement is running, most reporting you will be offered is volume-shaped: notices sent, listings found, cases opened. Volume measures effort. It can rise while your position gets worse.

Four metrics measure outcome.

Median time from detection to filing, per channel. This is where a provider's internal approval queue becomes visible. A case that sits waiting for an analyst is a case running during your peak ad spend. Ask for the median and the slow tail, per channel, not a single headline number across all of them.

Removal rate. The share of filings that result in an actual removal. A falling rate usually means the evidence attached to filings is thinning, or the reporting account's standing on that platform has eroded. Either way it is an early warning, and it is invisible if you only count notices sent. As a reference point, EnforceShield's takedown success rate is 97.3% on decided cases across its client base.

Recurrence rate. The share of removed infringers that reappear within thirty days. High recurrence means enforcement is acting on listings while the other side operates as a network of accounts — which is what it looks like when takedowns stop sticking.

Queue direction. Whether open cases are growing or shrinking week over week. This single number tells you whether your current capacity matches your current arrival rate. A rising backlog is the earliest reliable signal that the setup has been outgrown, and it shows up months before anyone feels it.

The Cost Structure Is Part of the Measurement

One dimension belongs in the business case and is routinely left out: what the model costs when infringement spikes.

Hourly legal billing rises with every case. Per-takedown pricing rises with every removal. Volume-tiered software rises at the tier boundary. All three share a property worth naming plainly — they charge you most in the month your marketing worked hardest, because that is the month the copies arrive.

That creates a quiet incentive to under-enforce exactly when enforcement matters. Pricing tied to the brands and SKUs being protected rather than to the count of infringements removes it, which is why EnforceShield prices per brand and SKU, not per takedown. When you compare providers, model the bill at three times your current case volume, not at today's. The ranking often changes. The wider comparison is in the three architectures brands choose between.

And if the number you build turns out to be a handful of isolated cases rather than a steady flow, a continuous program is more than you need. Single infringements can be removed through an on-demand takedown, with no subscription.

What Good Looks Like

A brand with this running knows four things at any moment: how fast a detected case reaches a filing, what share of filings land, what share of infringers come back, and whether the queue is growing.

None of those requires a dashboard full of charts. They fit in a weekly line. And together they turn brand protection from a legal cost that gets defended once a year into an operating function with a number attached — which is the only version of it that survives a budget review.

Where the leak shows upData you already haveWhat a rise in it suggests
Paid searchBranded-term CPC and impression shareSomeone else is bidding on your name
Product pagesConversion rate on pages with stable trafficBuyers are finding a cheaper 'you' elsewhere mid-journey
Customer serviceReturn reasons, 'not as described', warranty claims on units you cannot traceCounterfeit units reaching real customers
MarketplacesBuy Box share, unfamiliar seller names on your ASINsHijacking or unauthorized distribution
ReviewsSudden negative reviews describing a product you do not makeCounterfeits pooling into your listing's rating

Frequently Asked Questions

How much do counterfeits actually cost a brand?

There is no credible universal answer, and you should be skeptical of anyone who offers one. The widely quoted figures — global counterfeit trade running into the hundreds of billions, a couple of percent of world trade — are estimates of total cross-border trade in counterfeit goods. They describe a global phenomenon, not your brand. They cannot tell you what you are losing, because your exposure depends on your category, your price point, how visible your marketing is and which channels you sell on. The number that matters is the one you build from your own data, and the article above lists where to find it.

How do I build a business case for brand protection?

Start with the leak you can evidence, not the one you can estimate. Branded-search cost increases where a third party is bidding on your name, conversion softness on pages whose traffic has not changed, return and complaint volume mentioning product you did not make, and lost Buy Box share are all measurable today. Put a conservative value on each, state the assumptions plainly, and present the range rather than a single figure. A defensible range beats a precise number nobody believes, and finance will trust it more.

What metrics should I track once enforcement is running?

Four. Median time from detection to filing, per channel — this is the number a vendor's approval queue shows up in. Removal rate, meaning the share of filings that actually result in removal, which tells you whether your evidence is good enough. Recurrence rate, the share of removed infringers that return within 30 days, which tells you whether you are fighting listings or operators. And queue direction: whether open cases are growing or shrinking week over week. Notices sent is not a metric; it is an activity count.

Why is 'notices sent' a bad metric?

Because it measures effort, not outcome, and it can rise while your position gets worse. A brand can double its filings and lose ground if the removal rate is falling, if the same infringers keep returning, or if the queue is growing faster than it is cleared. Volume is also the metric most likely to be reported to you by a provider whose pricing is tied to it, which is a reason to be careful with it rather than a reason to trust it.

Does enforcement pay for itself?

It depends on the shape of the exposure and the shape of the cost. The case for a continuous program is strongest when infringement volume is high and rising, because the value recovered scales while a fixed cost does not. When infringements are rare and isolated, a full program is more than you need: an on-demand takedown removes a single case with no subscription, and EnforceShield offers that as a separate service. What tends to break the maths is a cost structure that rises with case volume, because it puts the largest bill in front of you in the month with the most infringement.

How quickly should I expect to see an effect?

Removals on the most responsive channels come relatively quickly once filings start. The effect on your own numbers is slower and noisier, because branded CPC, conversion and return rates all move for many reasons at once. Give it a full quarter before reading the business-level metrics, hold the enforcement metrics to a much shorter cycle, and be suspicious of any provider quoting a revenue-recovery figure in the first week.

Get a clear picture of your current exposure.

A 30-minute review across marketplaces, social, search, ads and lookalike sites — what is live against your brand right now, and what it is plausibly costing you.