Five digital marketing lessons from comScore

Stephen Whiteside
Warc

Digital media promise to help brands reach consumers in uniquely personalised and engaging ways.

"But like anything in life," Gian Fulgoni – Chairman/Co-founder of comScore – suggested on a January 2016 Warc Webinar, "I think new technologies come with positives and negatives. And, as we've found in digital marketing, not everything that is claimed turns out to be 'as sold', if you will."

In helping his audience of marketers to "leverage digital marketing far beyond what you might have experienced to date", Fulgoni outlined five areas where current industry practices often fall short:

• prioritising clicks;
• balancing search and display;
• understanding mobile;
• targeting, viewabilty, brand safety and invalid traffic;
• cross-platform measurement.

Prioritising Clicks

Fulgoni's first attacked a long-standing measurement technique. "Clicks on display ads are at best an incomplete – and, at worst, a misleading – effectiveness metric," he said.

Putting this observation into context, he referenced a comScore poll that asked clients, agencies, publishers, and ad networks if they employed clicks to gauge the success of digital display ads. "To my utter astonishment," Fulgoni revealed, "approximately – on average – 40% of each sector said 'almost always'."

Although the clicks metric has the advantage of being "simple, fast and inexpensive", clickthrough rates for display ads – expressed per thousand impressions – stand at only 0.10% in the US. "If one believes clicks are relevant, I think one in a thousand would basically say these ads are having very, very little – if any – impact at all," Fulgoni said.

"Something is amiss here with these click rates. And what's amiss is that these click rates aren't reflecting the effectiveness of the ads."

As evidence, he flagged up comScore analysis – published in the Journal of Advertising Research – showing that display ads encourage desirable behaviours such as website visits, online searches and actual purchases. "What we found is that there's no relationship between the click rates and the impact of the campaign," he said.

Further investigations featuring around 140 case studies demonstrated how display ads boosted both online and offline sales. "All of this is, I should stress, is happening with click rates of 0.10%," added Fulgoni. "One of the lessons learned here: don't just use the click to evaluate the effectiveness of display ads. It's just not an accurate indicator."

Balancing Search and Display

Fulgoni asserted that foregrounding clicks can muddle attempts at attribution, too: "Search, I think, over the years has got a lot more credit than it deserves, because a lot folks have tended to look at the last click and attribute virtually all of the effectiveness of their marketing efforts to that last click.

"What that ignores is all of the other marketing effort that has been working at establishing the value of a brand, or the importance of a brand, in a consumer's mind."

And digital display ads have, once again, suffered disproportionately as a consequence. Drawing on data from online marketing by over 100 multichannel retailers, comScore determined that this channel possessed specific strengths when compared with, and deployed alongside, paid search.

"It shouldn't be surprising that you're getting a higher sales increase among people exposed to a search ad, because those people have basically said they are in the market for something," Fulgoni said. "But that's only half of the challenge, because you also have to take into account how many people you can reach with a display campaign versus a search campaign."

Paid search, of course, is restricted to a selection of keywords, whereas display ads connect with a wider audience – and thus yield a greater uptick in overall volume sales. "A display-only campaign is generally going to drive more sales lift than a search-only campaign. The reason is: while display has a lower lift, you can reach more people," said Fulgoni.

"If you put them both together, you see the synergy … That's basically the best route to go. So the lesson here is: don't forget the value of display campaigns in building awareness and credibility of your brand name so that you get more clicks when you're running your search campaign."

Understanding Mobile

The Chairman of comScore insisted that the need for granular understanding similarly applies to mobile – not least because smartphones and tablets are now responsible for 64% of the time US consumers are online.

Drilling down into this topic, Fulgoni reported that 55% of the time spent on the internet in America is now taken by apps: "I think that these apps are representing a big opportunity, but also a big challenge, for marketers. Sometimes it's easy to overestimate the ease with which you can get an app downloaded on somebody's phone.

"But the reality is, in terms of generating awareness … you are competing against, I think it is, somewhere around two million apps in the Apple App Store and then somewhere around the same number – maybe a little more – in the Android Play repository. So the challenge, first of all, is getting the app downloaded. And then there's another challenge, which is getting it used."

A potential solution might be mobile advertising itself, given this messaging can help achieve a range of commercial objectives. "One might, on the surface, think that mobile advertising is less effective than desktop," Fulgoni said. "That's turned out to not be the case."

So what are the main drivers behind mobile ad effectiveness? "I think there are a couple of key reasons," he proposed. "One is that while the screen is smaller on a phone, you can takeover than screen. And your share of voice is likely to be significantly higher … You've got to be careful about whether it's coming across as intrusive, but I think it's possible to address that, especially on social networks.

"The other reason why these mobile ads are working so well, we believe, is that you can deliver them in a nicely-targeted way closer to the point of purchase, and that makes it easier for the ad to work at generating incremental sales for you."

Targeting, Viewabilty, Brand Safety and Invalid Traffic

Advocates of digital advertising – be it on mobile or desktop – typically champion the targeting capabilities mentioned by Fulgoni, but usually omit his accompanying caveat: "The reality is that it's not 100% accurate."

Cookies – the tracking tools placed on desktop devices – are a particular source of difficulty in this area. Their shortcomings include:

• deletion by web users;
• an inability to sort between multiple users of a device;
• inaccurate estimates for unique visitors;
• overstating reach;
• understating frequency.;
• problems with ad servers, which cannot distinguish between when cookies have been deleted and when consumers have not previously been exposed to an ad.

As a consequence of such failings, comScore has found that a modest 44% of ads ultimately reach the intended demographic. "Don't take the cookie-based number as truth," Fulgoni said. "It's very important to know your agencies are aware of that and incorporating that reality into their media planning."

Equivalent transparency issues apply to other elements of the digital marketing ecosystem:

• Viewability: this covers "whether or not the ad was ever in view of the consumer", and has attracted attention from blue-chip marketers like Kellogg.

"You've got a lot of these ad impressions that are never seen by a consumer. And if you couple that with the fact that half of these ad impressions might never reach the right demographic segment, you can end up in a situation where only a quarter of your ads are working for you."

• Brand Safety: Fulgoni defined this subject as "whether the ad was delivered in the context of objectionable content – adult sites, or hate sites, etc."
• Invalid Traffic: marketers often face challenges in recognising whether an ad exposure "was fraud, and how that fraud was happening."

Brands like La Quinta Inns & Suites have moved to address fraudulent activity, and comScore also has useful advice for brands. "The premium sites do a pretty good job of cleaning up this fraudulent traffic. It's on the non-premium sites that the problem really is much larger," Fulgoni said.

Obtaining a clearer picture in each of these fields should, he concluded, be a priority. "You really need transparency in this world, especially in this world of programmatic buying, where the computer is off buying the ads – [and] especially if it's happening on the open exchanges," Fulgoni said.

Cross-Platform Measurement

As the number of marketing channels proliferates, measurement practices are simultaneously growing in complexity. "We believe that both audience measurement and the media planning that then uses that data must adopt what we call a unified, cross-platform view of the consumer," said Fulgoni.

"We've got to get a handle on how the consumer is behaving across all these devices, and I think that's what caused the big problems in the industry. And I think what the content owners – rightly so – have been complaining about: that when ads are run on their content that appears on a device that does not get measured, they don't get paid for it."

One crucial part of this endeavour involves de-duplication, and "making sure that you understand whether you're reaching incremental eyeballs with another platform versus just delivering more impressions against the same set of people."

ESPN and comScore, for instance, partnered on Project Blueprint with this goal in mind, and discovered the sports broadcaster achieved a 21% increase in unduplicated daily reach thanks to digital. "I think that generally speaking, it's more valuable to a marketer to reach incremental eyeballs than to deliver more impressions against the people that have already been reached," said Fulgoni.