In short: RISE searches Google's Image Search Engine for semi-exact matches of your brand's images, finding infringements even when there's no brand name or keyword to search for.
Understanding RISE
RISE (Reverse Image Search Engine) extends semi-exact image matching technology to Google's Image Search Engine. It lets you detect infringing content across the internet, even when bad actors avoid using brand keywords or familiar markers.
Unlike basic reverse image searches that only detect exact matches, RISE uses similarity analysis to filter out false positives and surface actionable infringements.
1. Key capabilities
RISE expands your detection coverage beyond internal databases to monitor the wider web.
Bypass text evasion: identify infringing uses of images even when no brand names or keywords are present.
Find semi-exact matches: detect altered or slightly modified images across external search engines, rather than relying only on exact copies.
Broaden your search scope: extend infringement detection beyond internal platforms to cover the wider internet.
Reduce false positives: high-accuracy similarity matching means you review relevant content.
Improve detection: outperform standard, limited image search tools by capturing disguised visual assets.
Learn more: Image Match · Images
2. How RISE prioritises image searches
Prioritising means processing power and enforcement effort go where they matter most.
When large volumes of images enter the platform, RISE doesn't search all of them at once. Instead, it prioritises a targeted subset based on clear rules (for example, 15,000 high-value images out of a pool of nearly 400,000). It works through image groups in this order:
The "Urgent" pile: image groups flagged as needing a RISE update that haven't been searched yet. These are the highest priority: "check this immediately".
The "Repeat Offenders" pile: image groups already labelled as infringing. RISE prioritises these to catch new infringers trying to reuse the same images.
The "Brand New" pile: groups that have never been searched. Completely fresh images jump ahead of older ones.
The "30-Day Check-Up" pile: groups that haven't been searched in the past month. RISE checks them again for further infringements.
The "Human Request" pile: groups uploaded directly through the platform. If a user uploads an image to investigate it, RISE treats it as important and moves it up the queue.
The "VIP Product" pile: groups marked with a higher-priority label. These are images of a brand's most important products, such as current best-sellers, which take priority over items such as discontinued products.
The "Back of the Closet" pile: groups searched the longest time ago. As routine maintenance, RISE takes whatever has been in the "already checked" pile the longest and gives it a fresh look.
The "Coin Toss" pile: sorted alphabetically, or by ascending group ID. If thousands of images are tied on all the rules above, RISE uses alphabetical order or the group's ID number to break the tie and keep the queue moving.
Tip: Uploading an image yourself moves it up RISE's queue, so it's a quick way to investigate a specific image.
Learn more: How to Upload Images · Labels
3. Example: RISE in action
This is the kind of infringement keyword searches alone would miss.
The scenario: a bad actor creates a standalone e-commerce website to sell counterfeit luxury watches. To avoid detection by standard text-scraping tools, they leave the brand name, model name and any related keywords out of the website's text, and only upload pictures of the watches.
RISE in action: because RISE extends to Google's search engine, it scans the wider web for the brand's specific product shapes, designs and visual signatures. It detects the watch images on this obscure website and flags the URL for review, bypassing the infringer's text-evasion strategy completely.
Learn more: How to Find Infringing Domains & Websites · Websites
Quick reference
If you want to know… | Look at… |
What RISE can find that keyword searches can't | 1. Key capabilities |
Which images RISE searches first | 2. How RISE prioritises image searches |
What a RISE detection looks like in practice | 3. Example: RISE in action |
How Zeal matches images within platforms |