Changelog

What shipped, and when.

Where AdCrime came from, and everything we have shipped since. Newest first. High level on purpose: what it does for you, not how it works.

39 releases · 9 in September 2026

September 2026

9 releases

The ad as it was servedNEW

Reports now show the intercepting ad with the display URL and ad text it ran with, where the search captured them, instead of a version rebuilt from the landing domain. When an affiliate finding's ad puts your brand name in its headline, or calls itself official in English, German, French, Spanish, Italian or Dutch, a new card on the report says so: programme terms commonly restrict both, separately from keyword bidding. A discount it advertised is shown as a fact, never counted against it. The card reads earlier reports too, from the headline.

Was your own ad on the page?NEW

Each report now says whether your own ad ran in the same searches as the intercepting one, and how often the intercepting ad was placed above it. Filter the reports list to a single scan and a strip sums it up market by market: in how many searches with other advertisers your own ad was missing, and in how many it sat below one of them. Recorded for scans run from 27 September 2026.

An action for every account in the case fileIMPROVED

The downloadable case file used to give a recommended action for the first three accounts only. Now every account that can be suspended gets one, with any overflow on pages of its own. Repeat captures of one self-hosted account count as one account and one action, and ads we classified as competitors no longer appear among the offenders.

Check one of your codes before you sign upNEW

Under the instant exposure check on the landing page, a brand we already hold captures on can paste one code it issued and learn whether that code has been captured on an intercepting page, how many times, on how many operator pages, across how many markets, and since when. Which channel let it out stays in the coupon leak monitor, where your codes and their channels live.

No cap on keywordsIMPROVED

Paid plans no longer limit how many keywords you monitor. A check is one keyword in one market, once: the plan cards show the workload your included checks cover, and the scan builder shows what each run costs before you save.

Coupon leaks gets its own pageIMPROVED

Coupon monitoring moved out of the roster into a page of its own: how many of your codes are circulating, which channel let them out, every capture that carried one, and the codes we captured that your list does not know about yet, with one-click add. The list editor sits underneath the facts it feeds.

Geo split badge in the reports listIMPROVED

Offenders that served a sweep differently by market now wear a badge in your reports list, with the markets that answered both ways named on hover. Until now that finding only surfaced when you opened a report; now you can see it across the whole list at a glance.

API reference and rate limitsIMPROVED

The read-only reports API now has a public reference at /docs/api: authentication, every filter, every record field, error codes, and copy-paste examples. Each key gets a published rate limit with X-RateLimit headers on every response, so scripts can pace themselves instead of guessing.

Exposure breakdown by emailIMPROVED

The instant check now goes one level deeper. If your brand is in our register with confirmed interceptions, a company-domain email unlocks the full breakdown: which markets are being intercepted and how hard, and the most active operators, masked. It shows on the page and lands in your inbox. The free scan still names them.

August 2026

13 releases

Instant exposure checkNEW

Type your domain on the landing page and see your exposure in seconds, straight from our live register: confirmed interceptions, how many operators, how many markets, and how you compare to the thousands of brands we have swept. The full report still names names; the instant check tells you whether you need it.

Request checks from your dashboardNEW

Running low mid-investigation? Ask for more checks from the billing page with one line of context. Requests are reviewed by a person, usually within a day, and the answer lands on the same card.

Smarter scan setupIMPROVED

The scan builder now suggests the keyword patterns that actually catch interceptions, ranked by measured yield across our whole register, and flags the patterns that have never produced a confirmed catch before they cost you a slot.

Bidders: everyone else on your keywordsNEW

A new dashboard page lists every advertiser we captured bidding on your keywords without a proven affiliate route: real competitors, coupon sites, and keyword collisions, each shown with the exact keyword it ran on. Domains with a confirmed record elsewhere in our register wear a badge, in counts only.

CSV export and a read-only APINEW

Your reports, in your tools: one-click CSV export of any filtered view, and read-only API keys for pulling the same records as JSON or CSV from your own scripts and dashboards. Keys are shown once, stored only as fingerprints, and revocable anytime.

Account history in every documentNEW

Case files and evidence files now state persistence: when an account has been on your terms across multiple sweeps, the document says since when, how many captures, and how many separate sweeps. An incident reads differently from a standing arrangement, and now the paperwork says which one you have.

Coupon-leak monitoringNEW

Keep the list of codes you actually issued, with the channel each went out through. When a scan captures one of them on an offender's page, the report names it a leak and says which channel it escaped from. Codes we cannot match stay labeled honestly; we never guess which of your codes are real.

Geo-differential serving, called outNEW

Some offenders show the affiliate route to certain markets and a clean page to others. Reports now name that behavior, market by market, and when the same market answered both ways the report says so plainly: the difference is in the serving, not the location.

The complaint kit, completedNEW

Every confirmed detection now ships ready to file: the network's compliance form fields pre-filled from the evidence, and when the captured ad text carries your mark, a second rail prepares Google's trademark complaint with a statement of the capture facts.

Operator intelligenceNEW

Operators got their own page: the accounts behind your detections, clustered with a timeline each, plus a reach line when an account also appears in our confirmed register on other brands. Counts only; no other brand is ever named.

Evidence files, one detection at a timeNEW

Any single detection can now be exported as a standalone evidence file: a reviewer-ready PDF with the captures, the click path, and the filing steps, plus a ZIP evidence pack. Full case files download straight from the dashboard, no waiting on us.

The recovery clockNEW

Confirmed detections now carry a recovery window: how long your network's own rules give you to reverse commissions, and how many days remain, anchored to the report's own dates so a printed file stays honest.

Plans that speak in checksIMPROVED

One check is one keyword in one market, once. The scan builder's calculator now shows exactly what a configuration costs per run and per month, and sub-daily monitoring cadences arrived for the higher tiers.

July 2026

3 releases

Sonar 2.0: public betaNEW

Our detection engine's biggest upgrade: coverage across every major search market, more reliable detection across languages and regions, and fresh case files in minutes once you are monitoring. Included free in Pro.

The State of Affiliate Brand-Bidding 2026NEW

Our first public data report from the detection corpus: thousands of branded-search ads examined across fourteen countries, with the patterns, the networks, and the concentration of offenders laid out in full.

Sharper publisher-ID extractionIMPROVED

Classification now resolves the static publisher account more consistently across networks, so more detections ship with the exact ID your affiliate manager forwards to compliance.

June 2026

2 releases

Intelligence HubNEW

A public library on reading affiliate-fraud evidence: redirect chains, publisher IDs, and tracking parameters, explained for affiliate managers.

Report dossier polishIMPROVED

PDF case files got a cleaner evidence layout: publisher and merchant details surfaced up front, tighter page flow, and a forwarding-ready summary page.

May 2026

1 release

Email digests redesignedIMPROVED

Scan digests, enforcement updates, and onboarding mails moved to a single dossier-style template with clearer verdicts and case references.

April 2026

1 release

Coverage across more marketsIMPROVED

Monitoring expanded to more of the countries where interception actually happens, so a brand's blind spots abroad get watched as closely as its home market.

March 2026

1 release

More reliable detectionIMPROVED

A reliability pass meant fewer missed ads and steadier results run to run, especially on the busy coupon and promo queries where interception clusters.

February 2026

1 release

AdCrime goes liveNEW

The free brand scan opens to the public: tell us your brand, and we return a full case file of the affiliates bidding on it, with evidence, within 24 hours.

December 2025

1 release

First brands monitoredMILESTONE

A closed pilot put the engine on real brands' terms for the first time, running quietly and shipping evidence to a handful of programs.

November 2025

1 release

Confidence scoringNEW

The classifier learned to grade every detection: confirmed interception with hard attribution on one side, competitors and legitimate ads cleared on the other, and the uncertain cases labeled honestly.

October 2025

1 release

Evidence capture, end to endMILESTONE

The full case file came together: the ad as served, the redirect chain from click to landing page, the tracking parameters, the publisher ID, and screenshots of both ends.

August 2025

1 release

From script to systemMILESTONE

The rough prototype became a real, repeatable detection engine, able to run a brand's terms and follow every ad to where it actually went.

May 2025

1 release

Committed to building itMILESTONE

Nights-and-weekends research turned into a real project. If affiliate managers could not see this problem, we would build the thing that showed them.

February 2025

1 release

Mapping the networksMILESTONE

Months of digging into how the affiliate networks encode a publisher's account, so that a captured link could be turned back into a name a compliance team could actually act on.

November 2024

1 release

The first prototypeMILESTONE

A rough tool that followed a single affiliate's redirect chain by hand, just to prove the trail from a hijacked click back to the account being paid could be captured at all.

August 2024

1 release

The idea takes shapeMILESTONE

After watching affiliates bid on a brand's own name and quietly skim commission on customers it already had, the question that started AdCrime: why can nobody see who is doing this?

Questions about a release? mario@adcrime.com