Let’s Get Imaginative with Images: NetSeer’s Survey Explains How Publishers Monetize Visuals

Better Contextual Analysis and Programmatic Opportunites

We’ve written a bit in 2016 about visual monetization, and how much the marketplace has matured through smarter technology and growing interest and attention from publishers. There are factors we can discern anecdotally, or by looking at the broader digital market and extrapolating. But it’s more valuable to have fresh data around what’s driving these changes. 

To that end, NetSeer recently conducted a survey of publishers about the way they’re monetizing visual content, the challenges they’ve experienced, and what they want to do with it in the future. 

“We’re seeing a dramatic shift for publishers, acknowledging in-image as a viable way to enhance revenue,” NetSeer Head of Product Amir Bakhshaie tells us. “A year ago, most publishers didn’t really know about it. It validates how quickly the industry is growing.”

Digging deeper into that survey data, we tried to understand what the feedback can tell us about where visual monetization is going. Bakhshaie served as a guide through the survey results.

What’s Driving the Push

Visual monetization has been on the table for publishers for a long time. It’s been generally demonstrated through campaign reporting that user engagement is overall higher on the web with visual content than with text-based content. A number of vendors have jumped in gamely to try to monetize that engagement in one way or another.

But visual monetization strategies have been rising in priority among publishers—NetSeer’s survey indicated 90% of publishers rated in-image ads as either “important” or “somewhat important.” To see what’s pushed that number up, we can look at the evolution of technology and the pressures in the marketplace.

A clear driver here is the improvement in the technology. An obstacle to wider adoption has been the inability to understand what an image is communicating well enough to target effectively. Without this, publishers can’t justify charging premium rates and risk running afoul of brand contextual obligations (e.g., no airline ads on plane crash photos). The survey says 70% of publishers want ads laid onto their visual content to be highly relevant.

When it comes to understanding the context of images, Bakhshaie says, “Some [vendors] are very good, and pretty much all of the in-image guys are there. But more important: you also have to understand the context of the page.” 

That means looking beyond the metadata and other information bundled up in the image, and into the content surrounding it. Say you have a brand angling to be associated with a particular celebrity, and we know that person’s photo is on this page. But what if the article skews negative? The brand probably won’t want to be there. Understanding all of that data, image and page, together requires engineering chops. It’s taken a while for some of the players in the visual space to mature, engineering-wise.

There are also marketplace pressures leading the way to in-image advertising. The push to optimize for viewability has greatly cut down on publishers’ available display inventory and the corresponding increase in pricing has generally not made up for the lost revenue. Visual content lends itself intuitively to being viewable, and the 84% average viewability rates that NetSeer reports for in-image inventory beats the IAB’s benchmark of 70% for display campaigns. 

“That’s because you’re aligning the ad to the image, which is the focal point of the user’s attention,” says Bakhshaie. “Now that you have an ad that’s clearly visible, engagement tends to be anywhere from the order of 40% to 90% higher.”

So in-image could also be leveraged to minimize the display ad clutter hindering user experience. More and more—thanks in no small part to the rising ad blocking menace—publishers are looking for ways to continue monetizing while cutting back on ad placements and formats that users might consider intrusive. It’s a fine line to walk, especially when so many “intrusive” ads fetch high premiums. 

Take pre-roll—advertisers love it, it makes good money for publishers, and users typically will skip over it whenever they’re offered the option. That makes alternate means of monetizing video seem pretty appealing to publishers. 

A static ad overlaid onto a player (that the user can close out easily) might become a viable option for publishers on certain types of video content, or when the user hasn’t yet demonstrated a suitable level of engagement with the site. It’s also a good option for switching things up – monetizing video without bombarding users with pre-roll or mid-roll.

Programmatic Potential

Beyond contextual targeting, advertisers want to be able to employ their audience data in every type of advertising available. Publishers can make their visual ad inventory available in real-time exchanges and enable audience targeting on top of contextual.

But we all want to know how header bidding fits into this. Does it work? Presumably, the publisher has to share the appropriate image data, which sounds like extra legwork. And there’s the worrisome prospect of adding latency, when pubs are already dealing with enough concerns there: NetSeer’s survey found 79% of pubs want in-image advertising to have little to no impact on their site performance. 

Publishers want the highest yield, the most relevant ads, and close to no impact (maybe throw in a pony, too, if you’ve got one around). According to Bakhshaie, the ability to leverage header bidding with visual content “really comes down to the technical chops of the provider.”

Bakhshaie’s methodology involves processing image metadata and page content, then running the image through APIs to determine all options for placing and targeting the ads. 

“A lot of things have to be done in parallel,” he says. “You have to do that, and then be able to do header bidding. It becomes technically quite a feat.” 

To minimize latency and preserve user experience, Bakhshaie recommends (as do other people who have been proactive in onboarding header partners) limiting header partners to those with high yield and unique demand. 

“It means making batch calls as much as possible, so you’re limiting the number of calls you’re making on the client side,” he says. “It means pushing things to server-to server as much as possible. It means doing all these different things in conjunction, to make sure you’re having little to no impact on the publisher’s side.”

Avoiding User Experience Concerns

Page speed is a good place to start talking about user experience, but it goes beyond that to delivering relevant ads and avoiding overkill. Frequency capping is important, and Bakshaie advises that if you’re going to be placing ads over visual content, you want to make sure the revenue is high enough to justify it.

Also, the demand should be quality: Serving the user run-of-the-mill ads on top of content they’re directly engaged with can really muck up their loyalty to the site in question. Publishers need to determine how to cap frequency and how low they’re willing to go with CPMs in visual monetization. 

“Maybe one out of three times, when the revenue is interesting enough and the advertiser feels strongly about that specific user, they’re going to show a relevant ad,” Bakshaie says. “You’re not going to get run-of network advertising on top of images—they won’t pay enough. But you may get a national insurance provider that knows this user bought three other policies from them, and they have a fourth they want to recommend.”

There are times when a publisher might simply want to leave the image be. That’s something to work out with UX concerns, and with relationships between ops and editorial. NetSeer’s survey found 51% of pubs cited pushback from editorial as a reason for being hesitant to use in-image advertising. The argument: “We have this really nice picture; why would you want to put an ad anywhere near it?’” 

Bakhshaie recommends looking at all options to customize where an ad appears on visual content, and also choosing to show an ad only after the CPM surpasses a pre-set limit. 

He suggests you say to the editorial team: “I’m not going to show ads on your images 100% of the time. I’m going to show ads maybe 30% of the time, and by doing that you will still attain 80% of your revenue.”

Where Do We Go Next?

From NetSeer’s survey and other market indicators, it would seem visual monetization as an industry trend hasn’t peaked yet. Traditional newspapers have been monetizing more and more visual content on their digital properties, signifying a shift in monetization strategy away from older channels and toward new digital channels. 

And there’s rapid growth in stock images and UGC (user-generated content) images, which together constitute about 1.4 trillion new images per year. Companies like NetSeer have partnerships with major stock image companies, enabling simple integration of in-image advertising outlets. 

Even more interesting, some image distributors are developing plans that forgo (rather expensive) licensing fees in favor of revenue-share agreements that employ in-image advertising. In theory, with a lower cost barrier to entry, smaller pubs should be more willing to incorporate stock images, which means more ad inventory and opportunities – particularly to reach more niche audiences on mid-to-longer tail sites.

Furthermore, Bakhshaie points toward the alliance of Google and Shutterstock, and the unique position of companies that own all the content uploaded to their platforms. With an IPO waiting in the wings, Snapchat is particularly worth keeping an eye on. At the same time, the large platforms ought to pay attention to the more incremental, organic growth publishers have made in bringing the visual marketplace to a mature, data-driven position.

As the web becomes more visual overall, Bakhshaie sees text-based ads fading in prevalence, and growth of in-image demand at 50%-80% per quarter. It seems a fine time to hop on the trend as it’s still ramping up.