Take back control.

Protection solutionsDMCA takedown serviceGoogle copyright removal serviceCopyright monitoring servicePublisher anti-piracy serviceCopyright protection for creatorsImage licensing & attribution serviceSoftware piracy takedown service
Knowledge hubCopyright protection guidesCopyright essentialsTakedown field guidesContent protection
Company & standardsAbout LINK-BUSTERFees & service scopeResponsible enforcementContact

Independent copyright protection.
Hong Kong · United States

How to measure copyright takedown results

Measure takedown results with defined denominators and separate records for findings, submissions, rejections, delistings, source removals, and disputes.

Three distinct trays for paper records beside a restrained abstract bar-chart sculpture, no numbers or claims Conceptual imagery generated with OpenAI. Not documentary evidence.

Define the denominator before showing a rate

A transparent outcome taxonomy makes reports useful and marketing claims defensible.

A reproducible takedown reporting model

A useful report names the population, time period, action, decision maker, and evidence for each outcome. It does not turn submissions into removals.

Work completed

Count discoveries, reviewed URLs, and submitted requests as different activities.

Recipient decisions

Record accepted, rejected, pending, disputed, and restored states from the provider record.

Verified outcomes

Distinguish search delisting, source removal, and an inaccessible page whose cause is unknown.

Define the denominator

A dashboard showing a removal percentage should explain what it counts. Unique reviewed URLs, all discoveries, and submitted notices are different populations.

Specify the reporting period and whether pending cases are included. Preserve enough underlying records to reproduce the calculation.

Write a data dictionary before calculating a rate. Define a unique work, finding, URL, case, submission, recipient decision, and verified outcome. State how redirects, duplicate mirrors, parameters, multiple works on one page, and repeated observations are handled. Without these rules, two reviewers can produce different totals from the same case register while both believe they counted removals.

Use separate outcome categories

Discovery means a location was flagged. Reviewed means a person assessed it. Submitted means a notice was sent through a recipient channel.

Delisted means a search outcome, while source removed means the material is no longer available at the verified location. Rejected, withdrawn, disputed, and inaccessible deserve their own categories.

Use event dates and current-state fields for different questions. Event history answers how many submissions or decisions occurred during a period. Current state answers how many locations are presently removed, delisted, disputed, restored, or pending. Keep both. Overwriting a restored URL as if the earlier removal never happened loses operational history, while reporting the old removal as current overstates results.

Do not turn correlation into revenue proof

Sales changing after enforcement does not by itself prove a causal effect. Releases, promotions, price changes, and seasonality may also matter.

Present commercial observations with their limitations. Avoid claiming that every detected download would otherwise have become a sale.

Choose denominators that match the question. Submitted URLs can measure provider outcomes; reviewed findings can measure analyst decisions; all discoveries can show queue conversion. Publish the numerator, denominator, observation window, pending treatment, and last verification date beside the percentage. If a cohort is young, show its pending share rather than comparing it directly with an older completed cohort.

Audit the claims you publish

Before publishing a crawl-volume or removal statistic, specify the metric, period, deduplication method, and evidence source. Page fetches are not unique pages, and submitted URLs are not successful removals. LINK-BUSTER does not publish a billion-pages-per-day claim without operational telemetry that supports it.

Audit a sample back to source evidence. For each selected row, find the original work, reported URL, submission copy, provider reference, decision, and independent verification. Recalculate the published total from exported records and document exclusions. A metric that cannot be traced to individual rows should stay internal until the collection or definition problem is fixed.

Sources & further reading

This guide draws on the primary references below. Requirements vary by recipient and jurisdiction; check the current official process before acting.

General information, not an individualized legal opinion. For a disputed claim or litigation deadline, consult qualified counsel. Read our editorial and correction policy.

More from content protection

Publisher anti-piracy workflow: a practical guide

Build a publisher anti-piracy workflow that tracks titles, editions, rights, licensed distribution, suspected copies, priority, and verified outcomes.

Read guide

Protect an online course without targeting your students

Protect online courses by mapping lessons and source files, distinguishing legitimate student access from redistribution, and preserving lesson-level evidence.

Read guide

Software piracy: document the claim without downloading risk

Document suspected software piracy safely with product, version, licence, listing, and visible-page evidence - Without downloading or running unknown files.

Read guide

Ready to protect
what you created?

Choose ongoing monitoring, help with a specific infringement, or an attribution and licensing assessment.

Choose a service