Advertising has an AR problem

Generated by AI with Midjourney

At last year's at Meta Connect conference, Mark Zuckerberg spent 10m12s talking about augmented reality (15% of the keynote). This year that dropped to 4m03s (5%) with the main part an update about their partnership with Ray-Ban. Note the second last sentence from Rocco Basilico, the Chief Wearables Officer for Essilor Luxottica:

Zuckerberg: So, this is the first product that we built together [speaking about the Ray-Ban Stories glasses], but we've got bigger plans. Wanna give a hint of where we're going?

Basilico: Well, we are working together on a new set of glasses to build a portal to the metaverse. It's an exciting multi-year project.

A portal to the Metaverse is the interesting part. It suggests Meta is thinking of the Metaverse as a central hub to their other products. Certainly the Metaverse gets the vast majority of talk time at Connect conferences.

It’s possible Meta is focused on VR and the Metaverse because of the technological challenges of AR (miniaturization mainly), but there’s another, more troubling reason… they don’t know how to make advertising work in AR.

Print and early internet advertising

The first ever banner ad on the internet, circa 1994. It had a clickthrough rate of 44% (compared to 0.05% today)

Advertising on the early internet mimicked its predecessor, print. Ads were placed around the webpage content just like the ads around the articles in newspapers and magazines. Because the ads were on the periphery, they were visually flashy or provocative to draw people’s attention. This style of ads worked early on when they were novel (the first banner ad had a clickthrough rate of 44%) but eventually began converting poorly (less than 0.05% today) as users learned to ignore them.

It was Facebook that boosted advertising again when they invented the News Feed. This algorithmically generated list of posts with an infinite scroll was a format only possible online which had an effectively infinite amount of content to show and space to show it. Into this feed, they placed ads which could be personalized to each person, making them relevant and higher converting. And because they were in the feed (not on the periphery) they were in the user’s line of sight and impossible to completely ignore. Mobile only magnified the benefits by allowing ads to be fullscreen and even harder to ignore. The ad-in-feed became the internet’s native ad format.

No more screens

Augmented reality is different from our current devices in one fundamental way — there’s no screen to look at. AR glasses use a series of lenses to reflect light from a hidden screen in a way that makes the interface appear overtop the world in front of us.

In an app like Instagram, Meta controls everything that shows on the screen. That rectangle is a blank canvas they place the feed into, along with any ads. With AR glasses, the canvas isn’t blank, it’s the world the user is looking out at. There’s no feed and hence no place for ads.

In AR concept videos, ads are placed right into the user’s field of view, attached to buildings and products. These are no different from billboards today, just smarter and personalized. Artist Keiichi Matsudo showed how this will look if you believe that “marketers ruin everything.”

A bus ride in an AR dystopia.

The first attempts at advertising in augmented reality will be as unsophisticated as the banner ads in the early internet. Early adopters will engage out of curiosity but over time the ads will become ineffective. As companies try different things, a native format will emerge, one that’s effective for marketers and acceptable for consumers.

AR and direct response

Compelling AR demos center around helping people complete a task (video)

People don’t like advertising — they ignore, avoid, or pay not to see it if possible. For this reason, the dystopian vision of Matsudo could never happen because users won’t tolerate it, they’ll take the glasses off or never buy them. The brand advertising that puts up virtual billboards and giant dinosaurs in the city won’t work outside of novelty demos. People don’t want to feel like they’re on an acid trip all day.

Unsolicited direct response advertising — informing people about store hours, deals, or sales — could be helpful but annoying at scale. No one wants to feel like they’re walking around a carnival all day.

The most likely type of advertising is the solicited direct response — i.e. your AR glasses showing useful information you asked for or helping you do some task. Google is best positioned here because they’re the company people go to with questions. They also have the richest database of information about the real world. Meta is quite poorly positioned, only knowing what you might like or want based on a profile.

With control over the stack, and no need for advertising, Apple is well positioned to make AR a consumer success.

Apple's opportunity

Apple’s business model has always been selling devices that make people’s lives better. They have never depended on advertising and always focused on helping people accomplish tasks. Apple would be just fine if people their glasses lay dormant all day except if someone launched an app to do something.

Apple’s control of the software stack adds to their strong position. They could limit what apps could show in AR to ensure a good user experience. It would be in their power to prevent ads from popping up altogether, for example. This could be done explicitly in their Terms of Service (and enforced by app review), and implicitly with the development frameworks they allow 3rd party developers access to.

With control over the stack, and no need for advertising, Apple is well positioned to make AR a consumer success.

Even though mainstream AR is still several years away, there are enough concept videos, prototypes, and adjacent experiences to help us see what challenges are coming. For Mark Zuckerberg and Meta, the fact that their existing advertising model isn't a natural fit means these challenges will be significant, and their pointing AR experiences into the Metaverse suggest they know it.

Mark Rabo