ProductAdLens · AI video ad pre-testing

Know how your video ad will perform before you spend on media.

AdLens uses AI to predict how viewers will notice, remember and respond to your ads on YouTube, Instagram, Facebook and other social platforms. You get a clear score, scene-by-scene analysis and guidance on what to fix in minutes, not weeks.

  • No consumer panel needed
  • Scene-by-scene analysis
  • Benchmarked per platform

The problem

Most video budgets are committed on gut feel.

Traditional testing is slow

Consumer panels and agency pre-tests take weeks and are priced per study, so most video ads go live without ever being tested.

Every film becomes ten edits

One campaign turns into 6s, 15s and 30s cuts in 9:16, 1:1 and 16:9. Nobody has time to judge them all by eye.

The verdict comes after the spend

View-through and brand-lift numbers only tell you what didn't work once the media money is already gone.

What AdLens measures

Six pillars. One impact score.

Every video is scored on the KPIs that predict real-world impact. Each KPI gets a red, amber or green rating, with an explanation of why it matters and how to improve it.

Attention

Will the video stop the scroll and hold viewers?

  • Attention heatmaps from predictive eye-tracking
  • Engaging first two seconds
  • Scene pace and scene motion
  • Predicted viewer drop-off curve

Branding

Will viewers actually register the brand and the pack?

  • Brand cut-through, scene by scene
  • Brand in the first seconds and the last scene
  • Packaged product attention over time
  • Product cut-through at the start and end

Processing ease

Is it simple to follow and easy to remember?

  • Ad recall potential
  • Visual simplicity and clutter
  • Enough time to read on-screen text
  • Right aspect ratio and sound setting for the platform

Emotional engagement

Does it create a reaction strong enough to stick?

  • Human presence, including in the opening scene
  • Activation potential of the visuals
  • Activation potential of the on-screen text

Persuasion

Does it move people to act?

  • Call-to-action cut-through
  • Call-to-action in the voice-over
  • Engagement potential

Strategic fit

Does it say what you meant it to say?

  • Message fit across visuals, on-screen text and voice-over
  • Brand values fit across visuals and text

How it works

From upload to a launch-ready decision in four steps

01

Upload & brief

Upload the video and tell us the brand, market, target platform and, optionally, the messages, brand values and voice-over script.

02

AI analysis

AdLens breaks the video into scenes and predicts attention, branding, recall and engagement for each one.

03

Score & benchmark

Get an overall impact score, a score for each of the six pillars and a red, amber or green rating for every KPI, benchmarked against the platform.

04

Improve & share

Follow the guidance on how to fix each weak KPI, re-test the new edit, and share a presentation-ready report with your team.

Scene-by-scene analysis

See the exact second viewers drift away

AdLens maps attention, branding and engagement across every scene, so editors know exactly what to cut, move or reshoot.

  • Predicted drop-off curve: the share of viewers still watching at each second
  • Brand and pack visibility in the first seconds and the last scene, when recall is formed
  • Scene pace and motion compared with the platform benchmark
  • Whether on-screen text stays up long enough to be read
  • Sound-off readiness for feeds that autoplay on mute
Predicted viewer drop-off · 20s video
% of viewers still watching Brand cuts through
0%25%50%75%100%0:000:050:100:150:20Steepest drop-off 0:06–0:0943% reach the end

Illustrative sample. Real reports are generated for each video.

Platform benchmarks

Scored for where the ad will actually run

Aspect ratio, scene pace and sound setting are checked against the best practice for each platform. A cut that works on YouTube In-Stream is judged differently from one made for Reels.

YouTube In-StreamYouTube ShortsInstagram ReelsInstagram FeedInstagram StoriesFacebook FeedFacebook StoriesTikTok

Who it's for

Built for the people who sign off creative

Brand managers
Walk into approval meetings with evidence instead of opinions, and pick the strongest edit with confidence.
Creative & media agencies
Pressure-test films before client review and back up creative recommendations with data.
Social & performance teams
Choose the cut most likely to hold viewers before you put budget behind it.
Consumer insights teams
Screen every video quickly and save full consumer research for the strategic big bets.

Why AI pre-testing

Test every edit, not just the hero film

Turnaround
Traditional: Typically weeks
AdLens: Minutes
What gets tested
Traditional: A few finished films
AdLens: Every cut, length and aspect ratio
When in the process
Traditional: Late, after production
AdLens: From rough cut to final edit
Cost model
Traditional: Priced per study
AdLens: Test and re-test as often as you need
Output
Traditional: A research deck
AdLens: Scores, scene-by-scene charts and how-to-improve guidance

Built from real FMCG work

We've run pre-launch creative validation at enterprise scale

Before building AdLens, our team automated the creative-validation pipeline behind one of India's largest FMCG portfolios. That work cut cycle times for batches of hundreds of assets roughly in half. AdLens is built on what we learned there about volume, brand and market context, and getting answers to marketers faster.

Read the Hindustan Unilever case study
  • Bulk testing
    Validate full campaign batches, not one video at a time
  • Presentation-ready reports
    A shareable report for every video, ready for the approval meeting
  • Workflow integration
    Connect to your asset library, agency hand-off and approvals

FAQ

Questions we usually get

Something else on your mind? Ask us directly.

What exactly does AdLens measure?

AdLens scores a video ad on six pillars: attention, branding, processing ease, emotional engagement, persuasion and strategic fit. Each pillar is made up of specific KPIs, such as brand cut-through in the first seconds, a predicted viewer drop-off curve and call-to-action visibility. The pillars roll up into one overall impact score.

How does it work without a consumer panel?

AdLens uses AI models, including predictive eye-tracking, to simulate how viewers watch and process a video scene by scene. That makes it possible to test every edit in minutes instead of recruiting a panel for each one.

Which platforms and formats does it support?

AdLens is built for video ads on social and online-video platforms: YouTube In-Stream and Shorts, Instagram Reels, Feed and Stories, Facebook Feed and Stories, and TikTok. Checks like scene pace, aspect ratio and sound setting are benchmarked against the platform you select.

What do we need to provide?

The video file, plus the brand, market and target platform. To get the strategic-fit scores, also add the messages you want to land, your brand values and the voice-over script. Hindi and other non-English scripts are fine.

Does it replace consumer research?

No, and it isn't meant to. Most video edits never get a research budget. AdLens gives those decisions evidence, so you can shortlist faster and save full research for the strategic, high-stakes work.

Can it fit into our existing creative workflow?

Yes. Carufus builds system integrations for a living, so AdLens can be connected to your asset library, agency hand-off or approval process. That includes bulk testing for high-volume campaigns.

Is our unreleased creative kept confidential?

Yes. Pre-launch creative is treated as confidential. We're happy to walk your team through data handling during the demo.

How do we get started?

Book a demo and bring a few past video ads whose in-market results you already know. We'll run them through AdLens, so you can judge the predictions against outcomes you trust.

Test your next video before it costs you.

Bring a few past ads whose results you already know. We'll run them through AdLens live, so you can judge it for yourself.