As cookies slowly phase out in response to digital privacy law and platform changes, publishers and advertisers must engage new models for revenue generation and reaching audiences. Transformations in digital media will require more creative ways to target ads, and in many cases, this era of digital advertising must build on old methods upgraded by new technologies.
Lineup’s Digital Product Manager, Tiffany Kelly, explains, “People are being more strategic in the way they are targeting – building out specific keywords, using time of day, emotions-based and weather-based targeting, among others to determine when users are more likely to engage their brands. They are becoming more thoughtful about relevancy and timing to make digital ads more personal. Everyone is getting super creative, but in many ways going back to basics.”
Advertisers and publishers collaborating in the next chapter of the digital media era will see success by integrating different aspects of these five ad targeting models to serve their readerships and audiences:
- Contextual Targeting
- Proximity and Weather Targeting
- Sentiment Targeting
- Multichannel Targeting and Retargeting
- Awareness and Attention-Based Targeting
Using a variety of these ad targeting approaches can help brands refine their digital approach to effectively engaging readers. Tiffany Kelly at Lineup suggests that the impact of a stronger reader and consumer digital experience will benefit both sides of the publisher-advertiser relationship in ways the industry is only beginning to see. “User experience is huge now–in the last 4-5 years, the industry is finally realizing that a bad ad experience will drive readers away forever. Even kids can recognize the difference between a good and poor ad experience, which impacts a brand. If the publisher pays attention and has strict rules with their ad experience, it makes a difference getting loyal readers.”
Contextual targeting with the support of artificial intelligence
Google estimates revenue could fall “for the top 500 publishers by 52% on average” when the cookie finally crumbles on Google Chrome in 2022; GlobeNewsWire forecasts ‘contextual targeting’ as a likely alternative solution: “Rather than targeting ads based on what consumers have viewed in the past, advertisers can target ads based on the content consumers are viewing in real time. Analytics will play a key role in matching ads with relevant, brand-safe, and viewable content.”
Those three components–relevance, brand safety, and viewability–have become growing concerns in a digital climate where readers remain inundated with information on a variety of platforms and advertisers must navigate a highly charged socio-political digital environment.
Contextual targeting isn’t new, but artificial intelligence and other technological advancements have changed the way it’s done, according to Digiday. “Advertisers moved away from contextual [and towards behavioural targeting] once audience tracking entered the scene, but the present-day resurgence can be attributed in part to consumers and a growing concern for how personal data is being used…With so much data across the Internet, especially for advertisers and brands, machine learning is crucial to tailoring a contextual approach to deliver not only positive contextual associations and brand suitability but also to create connections that convert.”
One example of this tech upgrade to contextual ad targeting is Oracle Contextual Intelligence (formerly Grapeshot). Tiffany Kelly says of the application, “Oracle Contextual Intelligence (Context) analyzes textual content on web pages at the massive scale and speeds required by automated advertising technology to determine the context, the central meaning, of the content on a page. The primary uses for Context are for brand safety, contextual targeting in advertising, and prediction of trends:
- Brand safety. You can use Context for brand safety purposes to help avoid having advertising messages placed on pages with content that is inappropriate or irrelevant to the advertiser. The decisioning can be based upon existing standards, such advertising industry groups’ approved lists of topics to avoid, or customized to match the particular needs of specific brands and products.
- Contextual targeting. Context finds pages whose content is relevant to advertising messages. The application can enhance other targeting applications, adding context to the mix to discover pages that might otherwise have been missed, such as in hard news sections of a website that might otherwise be avoided altogether. Both applications can also be applied to on-demand videos that include spoken word. The spoken language is converted to text, via a speech-to-text application, then analyzed for context using similar processes to text on a page.
- Trend prediction. The Predicts feature in Context offers a heads-up on trends that enables you to identify topics that are likely to provide growth in advertising impressions. Predicts combines analysis of web pages along with analysis of trending topics on social media to surface upcoming opportunities likely to generate advertising inventory. It monitors pages and social media, analyzing increase in use of certain language and trends to show topics that within hours or days are more likely to increase in volume of coverage. These topics may be around certain events or news, such as the doings of celebrities, holiday surprises, sporting events and so on.”
Oracle Contextual Intelligence is not the only technology of its kind changing the future of contextual targeting. George Manas, president and chief media officer at OMD USA told Beet.TV that other AI and machine learning programs are also “bringing a degree of precision, scale, and automation that is truly new and unique in the [media and ad] programmatic marketplace…”Some advertisers will choose to keep their content away from contexts driving sparking fear or anger, others away from controversy. With this sophisticated approach, advertisers will be able to refine their placement choices so their branded content is placed safely in view of engaged audiences in relevant ways.
Proximity and weather targeting for hyper-local relevance
Contextual relevance is not the only metric for gaining reader interest in advertising content–many advertisers are also turning to location-based targeting. While geo-fencing has been used for some time to serve digital ad specials in the same physical location as competitors’ brands or for other place-based angles, proximity targeting is also getting an upgrade.
The Drum celebrates the shift as “one of the most exciting advances in the context-based geo-targeting revolution.” Not only can ad technology determine fixed location for audiences based on census data and other insights, but can also leverage mobility: “[Audiences] are targeted in real-time based on where and when they go to various locations within a city. Imagine, as that sports shoe brand, being able to activate a marketing message on a digital billboard the moment a sports lover walks by and then retargeting that same person on their mobile device moments later.”
Of course, place and time aren’t the only considerations for what makes ad content relevant to readers. Weather can make a difference too, according to location ad technology company GroundTruth: “Want to take hyperlocal targeting one step further? Use location-based advertising and proximity targeting to serve ads to people based on where they are now, where they’ve been, or where they’re going. If someone’s just walked past a shoe store and sees an ad with the right offer, they might turn back to have a look. Combine this with location and weather-based targeting so you can show the right ads to the right people at the right time. It’s just like the corner stores that put out their umbrella display when the sky turns grey.”
This hyper-local ad strategy builds off the reality that proximity searches have jumped 150-500% on Google over the last 2 years; in tandem with publishers’ understanding of their readers’ location and interests based on first-party data, proximity and weather-based strategies are a localized approach that may continue to accelerate.
Sentiment targeting for effective engagement by large or small brands
The emotional context of content can be harmful to brands with “almost three-quarters of UK consumers (73%) [feeling] that the sentiment of an article impacts their perception of a brand that has advertised alongside it.” (What’s New In Publishing, 2020) But in the right circumstances, neutral or positive emotional context can also be a benefit.
This targeting model merges sentiment data points with any emotion-based reach. Brand safety is impacted by ad placement next to unwanted emotional content–perhaps something depressing, controversial, or fearful. But when the emotional tone, context, and content of an ad placement is intentionally selected, it can actually be a boost to a brand.
Tiffany Kelly, Digital Product Manager at Lineup Systems, explains, “Large media companies such as Hearst and News Corp have shared the advanced targeting strategies they’ve used to serve personalized ads to readers based on specific keywords within content they’re consuming or by their preferences, opinions and emotions – whether they’re in the purchasing mood or they’re feeling optimistic, for example. These targeting tactics reportedly increase ad engagement by up to 45%.”
Reader profiles built from first-party data collected through survey responses, context targeting, article history, and machine learning in real-time can be used to predict mood and target ads accordingly. This dynamic has elevated the relationship between publishers and advertisers to a whole new level. Says Kelly, “For the first time ever, advertisers, ad buyers, and publishers are working together. One major reason is to try to cut out walled gardens–Facebook, Google, and Amazon, for instance. Another reason is that the buy-side is realizing how powerful it is to have these relationships with publishers because they want their first-party data. They don’t see the high performance they see with publishers when they are buying programmatic, real-time bidding ads through random sites. Buyers may get really poor performance, even if they got a great deal on those impressions, so they are willing to pay higher rates to publishers to get the higher performance and impressions for their brands.”
The New York Times has been able to effectively leverage this sentiment targeting to its advantage at a parallel growth scale, but the approach isn’t accessible only to large publishers or ad buyers.
Businesses of any size can take advantage of sentiment targeting through free tech tools, suggests Kelly: “Small, medium, and large brands can use Google Analytics to build out audience segments from subscriptions like websites or newsletters, mood survey results tied to free articles based on interest, and other methods. If they’re using Google Ads Manager, they can import audiences they filled within Google Analytics and use those for audience targeting with key values and article-based targeting….all completely free emotions-based ad targeting and optimization.”
Read: More insights on emotions-based advertising.
Multichannel targeting and retargeting for stronger understanding of audiences
It’s clear that owning first-party data can benefit publishers and make them more valuable to advertisers. However, owning the platform to engage multichannel targeting and retargeting practices can also strengthen publishers’ insight offerings for ad buyers.
Digiday explains, “Most publishers already have a Facebook and Twitter presence, and perhaps even Instagram for sharing striking visuals that tell the story. The problem with these platforms is that they own the user data; because social media traffic is a referral source, publishers don’t get the benefit of being able to analyze that information to create a more personalized experience.
However, when publishers engage audiences over channels such as email and push notifications — channels they control and that generate the data they need to better understand audience interests, they can give audiences a more personalized experience. And, by building a unique user persona for each subscriber based on their email address, publishers can then recognize the same user across all those channels, customizing the content accordingly to provide a consistent experience.”
Of course, these insights can benefit publishers in refining their own content, engagement, and audience relationships, but it also gives them deeper knowledge of audience segments in ways that advertisers can benefit from too — such as which channels may be most effective for which audiences.
Alongside multichannel targeting, Hubspot also recommends retargeting, which can benefit both publishers and advertisers trying to grow their brands: “Retargeting — a form of advertising that targets your website’s bounced traffic on other platforms — is powerful when used in conjunction with multichannel marketing. By having multiple platforms from which your audience can find your website, you’ll end up increasing website traffic. Anyone who bounces away will see retargeting ads on other platforms, ones that you may have a presence on.”
Awareness and attention-based targeting builds off reader interest
Machine learning that has advanced contextual, proximity, and sentiment ad targeting can also point to the longevity of reader attention in an advertising space. Publishers can use data points like how long certain readers stay on a page–or how far they scroll through a newsletter or article–to offer alternative value metrics to advertisers. Advertisers can target based on attention and awareness, especially for subscriber audiences.
As pandemic and political pressures have compelled ad buyers to demonstrate their value and maximize shrunken marketing budgets, awareness and attention-based advertising offers another metric of success outside of more traditional measures like impressions.
Digiday explains how The Telegraph is already doing this effectively–and seeing it generate 40% higher ad memory by subscribers compared to non-subscribing readers:
“The Telegraph first introduced Metrics that Matter — its efforts to measure ad success beyond clicks and impressions — in November last year. First, it takes 11 metrics that it deems qualify user attention across its ads, like dwell time, interaction and time in view. The publisher ties how these attention metrics correlate to brand uplift metrics like awareness, consideration and action. Now, it’s in the process of tying how brand uplift leads to advertiser commercial goals through its partnership with tech company Infosum. The average campaign price has increased by 24% year-on-year since launching Metrics that Matter…After looking at 25 direct display and content-led ad campaigns that ran between March and June, The Telegraph found that the top five that performed highest in terms of attention metrics all scored high on driving higher action intent.”
Of course, with digital media and advertising at the center of mental health concerns for young people globally, setting intention and ethical practices is a must for using this ad targeting model as a publisher or ad buyer. Leveraging human attention in news media contexts for its market benefits–without boundaries–is potentially damaging to democratic political structures, human rights, and health.
MIT Technology Review explains some of these dangers from commodifying human attention without accountability or ethical consideration:
“For tech companies, pursuing the infinite growth of extracted human attention leads to a similar crisis of global consciousness and social well-being. We need to shift to a post-growth attention economy that places mental health and well-being at the center of our desired outcomes…recognize the massive asymmetric power that technology companies have over individuals and society. They know us better than we know ourselves. Any asymmetric power structure must follow the fiduciary or “duty of care” model exemplified by a good teacher, therapist, doctor, or care worker—that is, it must work in the service of those with less power. It must not operate with a business model based on extraction. Upgraded business models for technology need to be generative: they need to treat us as the customer and not the product, and align with our most deeply held values and humanity.