Building a Brand-Protection Program: 7-Step Framework

Share

publisher: QSDEFENDER Brand Protection Advisory Team

Issue Time: Oct 30,2026

Summary

A manufacturer-grade 7-step framework for building a brand protection program: quantify SKU-level exposure, fix the legal and customs floor, choose physical and digital identity layers, engineer a label that survives the line, run server-side verdicts with clear key custody, connect supply-chain checkpoints and channel alerts, then operate a quarterly KPI loop. Includes the ISO 22383:2020 replacement for the withdrawn ISO 12931 and honest boundaries on what benchmarks can prove.

Most brands do not start a brand protection program framework because they want one. They start because a distributor is undercutting them by 40%, a customer posts a photo of a label that does not scan, or customs calls about a container nobody can account for. By then the response is usually reactive: one takedown, one new hologram, one legal letter. This guide lays out a seven-step framework that a mid-size manufacturer can actually run — with the deliverable, the owner, and the evidence each step must produce before the next one starts.

Set of security labels combining holographic, VOID and serialized QR elements on a single sheet

A program is a stack of decisions, not a catalogue order: each layer answers a different attack.

The 7-Step Brand Protection Program Framework at a Glance

A program differs from a project in one respect: it has an operating loop. Authentication technology is a component inside that loop, never the whole of it. The seven steps below are sequenced because each one constrains the next — if you pick a label before you know which SKUs lose money to fakes, you will buy the wrong label at the right price.

Step	Question it answers	Deliverable	Typical owner
1. Quantify exposure	Which SKUs, channels and markets actually lose money?	Ranked risk register with revenue at risk	Finance + brand
2. Fix the legal floor	What can we enforce, and where?	Trademark coverage map, contract clauses, customs filings	Legal / IP counsel
3. Choose the identity layer	How will a unit prove it is ours?	Layer specification: physical + digital	Brand protection + packaging
4. Engineer the label	Will it survive our line and our customers?	Approved artwork, substrate and test report	Packaging engineering
5. Build the verification chain	Who decides a scan is genuine, and how fast?	Server-side verdict logic, key custody plan	IT / security
6. Connect supply chain and channel	Where do units change hands, and who sees it?	Checkpoint map, alert rules, evidence pack	Supply chain + sales ops
7. Run the operating loop	Is it working, and what changes next quarter?	KPI dashboard, quarterly review, incident playbook	Program lead

Step 1: Quantify Exposure Before You Buy Anything

Start with a risk register, not a supplier shortlist. For each SKU, score four inputs: unit price and margin, channel complexity (how many hands between factory and shelf), geographic exposure (where your product is copied or diverted today), and regulatory consequence (what happens if a fake injures someone). Multiply the first two to get an exposure weight, then rank.

The macro numbers justify the exercise but should never be used as your own estimate. The OECD and EUIPO joint report Mapping Global Trade in Fakes (2025 edition, based on 2021 customs data) put global counterfeit trade at USD 467 billion, or 2.3% of world imports, with the EU absorbing USD 117 billion — 4.7% of its imports, roughly double the global rate. Those figures describe the market, not your loss. Your loss has to come from your own returns, distributor complaints and test purchases.

Output: a ranked list of no more than ten SKUs, each with an annual revenue-at-risk figure and a named owner.
Gate: do not proceed for any SKU where the annual loss estimate is below roughly three times the projected label cost — that ratio is our working rule of thumb, not an industry standard.
Common mistake: protecting the highest-priced SKU first. Copycats usually attack the fastest-moving mid-price item because volume, not margin, pays their tooling.

Step 2: Fix the Legal Floor First

Authentication without enforceable rights produces evidence you cannot use. Three items belong in this step:

Trademark coverage in every market where the product is made, distributed or sold. Rights are territorial: an EU registration does nothing in China, the US or Brazil. The WIPO Madrid System lets one application designate many member states, which is usually cheaper than filing country by country.
Contract terms that make technology operational. Territory definitions, receipt-scanning obligations, audit rights, penalty clauses and a clause stating who owns the serial data. A scan alert without a contractual consequence is a chart, not a control.
Customs engagement. In the EU, rights holders file an application for action so customs can detain suspect consignments; in the US, CBP recordation plays the same role. Both require a product guide showing officers how to tell genuine from fake — ideally in the local language.

Standard-status check: if a supplier cites ISO 12931 as the basis for its authentication design, that standard was withdrawn in July 2021. It was replaced by ISO 22383:2020, which organises authentication elements into physical, material, digital and combined categories and explicitly excludes economic criteria — meaning the standard will not tell you whether a solution is worth its price. That judgement stays with you.

Step 3: Choose the Identity Layer

This is where most budgets are wasted, because the choice is usually made on unit cost alone. Decide by the verification moment: who scans, with what device, and what must the answer prove.

Array of serialized QR code labels each carrying a unique identity number

Serialization turns a label into an identity: the code is public, the record behind it is not.

Layer	Verification moment	Strength	Watch out for
Serialized QR	Consumer phone scan, anywhere	Zero hardware cost, works on any packaging, easy to reprint	Copied codes look identical until the server sees a duplicate scan
NFC (e.g. NTAG 424 DNA)	Phone tap at close range	AES-128 authenticated URL that changes every tap; tamper loop option	Requires server-side validation; per-unit cost is higher than print
UHF RFID	Warehouse and logistics reads, metres away	Bulk reads without line of sight; case and pallet level	Not a consumer-facing proof; tag performance depends on the surface
Overt physical (hologram, VOID)	Human eye, no device	Immediate deterrence and visible tamper evidence	Deterrence only; a determined copier can imitate the look

In practice a workable program combines at least two categories — a visible deterrent plus a serialized digital identity — which is the layered logic we describe in layered authentication. For a deeper comparison of frequency bands before you commit, see UHF vs HF vs NFC frequency choice.

Step 4: Engineer the Label to Survive the Line

The chip rarely fails first; the construction does. Four specifications decide whether the label still works after six months on a shelf:

Print quality: module size, quiet zone and contrast determine whether a phone camera can read the code under warehouse lighting. Undersized modules are the single most common cause of a "dead" QR code.
Substrate and adhesive: curved glass, low-surface-energy plastics and frozen cartons each need a different adhesive. A label that lifts at the edge is a tamper signal you did not intend.
Application method: hand application, semi-automatic and high-speed line application impose different tolerances on die-cut spacing and liner strength.
Durability tests: abrasion, humidity, temperature cycling and chemical wipe — run on the finished label on the finished pack, not on a sample sheet.

Label production line with die-cutting and roll handling equipment in operation

Print and die-cut tolerances are set on the line, not in the artwork file.

Print specification detail belongs in the artwork brief, and we cover it separately in the QR implementation playbook. If you are still comparing suppliers at this stage, the RFID and NFC buyer's guide lists the questions worth asking before a sample run.

Step 5: Build the Verification Chain Before You Print

A label is a claim; the server is the judge. Before the first production run, settle five decisions:

Where the verdict lives. Genuine, already-scanned, unknown and tampered are four different answers, and each should route to a different message and a different internal alert.
Key custody. For cryptographic tags such as NTAG 424 DNA, the AES keys can be held by you, by the label supplier, or split. Whoever holds them can mint identities; write the ownership and rotation rules into the contract.
Replay detection. Dynamic URLs and read counters let the server distinguish a first scan from a copied code — provided the counter state is synchronised and monitored.
Latency budget. A consumer will abandon a verification page that takes more than a few seconds. Treat the verdict response time as a product requirement, not an IT metric.
Failure modes. Offline verification is not a thing for encrypted tags. Decide what the page says when the network is down, and never let it say "genuine" by default.

Step 6: Connect Supply Chain and Channel

Unit identity only becomes intelligence when it is read at handovers. Four checkpoints cover most diversion cases: factory dispatch, distributor receipt, retail receipt, and the consumer scan. Each read produces a timestamp, a location and a device identity; the gap between expected and actual is the signal.

Alert rules should be few and defensible: a scan outside the authorised territory, an outbound-to-inbound mismatch, inventory that exists in the system but not in the warehouse, a serial scanned in two countries within an hour, and an abnormal burst of scans from one device. We describe the rule set and the evidence package in RFID against gray-market diversion.

Step 7: Run the Operating Loop

Programs fail quietly: the labels ship, the dashboard goes unread, and nobody notices that scan rates collapsed after month three. Six indicators are enough to keep it honest.

Indicator	What it tells you	Planning benchmark
Verification rate	Share of units scanned at least once	Category-dependent; track the trend, not the absolute
Duplicate-scan ratio	Share of serials scanned more than expected	A rising ratio is the earliest copy signal
Verdict distribution	Mix of genuine, duplicate, unknown, tampered	Unknown spikes usually mean a print or data error
Time to first detection	Days from counterfeit listing to internal alert	Shorter is better; set a target and review it quarterly
Enforcement latency	Hours from evidence pack to takedown request	Counterfeit sellers relist quickly; speed compounds
Revenue at risk	Exposure attached to open incidents	The number that justifies next year's budget

Review quarterly, refresh the risk register annually, and after every significant incident ask one question: which step in this framework would have caught it earlier? That answer, not the incident itself, is what improves the program.

Where Programs Fail

Across the categories we work in — pharmaceuticals and supplements, wine and spirits, cosmetics, automotive parts — the same five failures recur. Buying a technology before defining the loss. Treating a takedown as a strategy. Choosing a label nobody on the line can apply at speed. Leaving key custody undefined. And measuring activity (labels printed, listings removed) instead of outcome (revenue protected). The framework above exists mainly to make those failures visible early.

Two boundaries worth stating plainly. First, this is operational guidance, not legal advice; customs filings, contracts and regulatory obligations need counsel in each jurisdiction. Second, the benchmarks in the tables are planning aids drawn from general practice, not guarantees, and no client data appears in this article.