Seen & Unseen

SHEDDING LIGHT ON computer vision

For marketers, visual content has quickly become the primary canvas for telling stories and influencing consumers. But this surge in visuals makes it that much harder to stand out amongst countless images and videos — and avoid some brand-unsafe associations. Fortunately, marketers don’t have to fly blind: Computer vision — a technology that processes and understands visual information — can help them see the right opportunities in a highly visual digital world.

This is how many media pros said visual content is important to their marketing strategy:

80%

said it wasn’t important at all.

At this point there is nothing we’re doing that isn’t visual.

— Erin Rech, head of digital, Initiative

Marketers deploy it in many different ways:

But it has its own share of challenges lurking on the periphery.

72%

images & videos on owned-and- operated platforms

29%

images & videos in social posts

36%

augmented social graphics

19%

images & videos in display ads

48%

insufficient viewability

41%

contextual irrelevance

42%

lack of brand awareness

The challenge is the sheer scale of the images that are out there. You’re talking about [scanning through] millions of photos a day. It makes that passive monitoring very difficult.

— Tom Dunmore, former technology journalist and co-founder, Mediablaze

To brighten some of those dark spots, marketers look to cutting-edge technologies to gain an advantage.

360-degree video

Virtual reality

Augmented reality

Wearables

Computer Vision

Computer vision can help you build a visual index of key images that you would want to run ads next to based on what’s appearing on screen.

— Erin Rech, head of digital at the IPG-owned agency Initiative

Computer vision tools utilize deep learning — that’s the tech that lets Snapchat detect human faces for dog ear filters or recognize other patterns.

Some high-profile brands are already putting the technology to use.

Maybelline

L’Oreal

Served up contextually relevant ads next to images related to beauty and personal care.

canon

Placed contextually relevant ads alongside online imagery related to cameras and videography — and more than doubled the average industry benchmark.

Dick’s Sporting Goods BOunty New balance Jaguar nike

Prevented ads from running on site and apps where audiences are unlikely to be interested.

The bad news: Just 12% of marketers use computer vision today.

Looking for reach? This tech plugs into major DSPs, so all of your buying gets visually smarter.

Only 10% use it specifically for ad targeting.

Meaning many marketers could end up left in the dark.

Discover the unseen opportunities all around you.

That means 88% still haven’t opened their eyes to the opportunity...

When you switch on computer vision, you start to...