Targeting ads to specific ethnic groups can be sensitive and hard to get right. But advertisers should give it their best shot.
Marketers are zeroing in on Hispanic and Latinx audiences, in particular, because of their purchasing power, which could hit $2.6 trillion by 2025, according to Insider Intelligence. The demographic now makes up about 20% of the US population and 25% of American millennials and Gen Z.
Yet brands interested in targeting this audience run into challenges, from insufficient data that doesn’t accurately represent different segments to cookie-cutter creative that erases the complexity of a demographic with roots in more than 20 countries.
Without a strategy for tracking disparate data points on Hispanic users, “you’re gonna miss them,” said Ana Ruedaquintero, associate director of media strategy at digital marketing agency Wpromote. “They’re going to go over your head.”
Data problems
Bad data is one of the biggest culprits behind Hispanic and Latinx audience mistargeting.
On average, four in 10 US Hispanics are not accounted for in third-party data sets, according to data validation company Truthset.
Inaccurate data from third parties poses a “really big problem” for successful ad sales, said Brian Lin, TelevisaUnivision’s head of product management, speaking on stage at Advertising Week in New York City in October.
Lin referenced a national campaign it ran for a brand client last year meant for Hispanic viewers. Problem was, the campaign served roughly 70% of its impressions to non-Hispanic households.
One issue is that data brokers often categorize audiences as Hispanic based on surname. But many Spanish last names are also prevalent in other demographics, said Joe Camacho, co-founder and chief global expansion officer at CTV and video ad platform Sabio Holdings.
Marketers should not assume that what resonates with one Hispanic audience will resonate with all, said Adriana Waterston, an EVP at custom primary market research company Horowitz Research.
“It is a disservice to look at an average number, whether it’s the average Hispanic, Asian or Black person in the US, and think that they represent everyone,” she said.
First party first
Instead, buyers and programmers should lean on first-party data to accurately reach Hispanic consumers.
Putting first-party data in a clean room, such as Google’s Ads Data Hub or Amazon Marketing Cloud, is the first step, said Jess Nachtigall, EVP of analytics and optimization at Brandtech Group-owned performance agency Jellyfish.
“If you know the audience that you are already interacting with and how they identify, that’s powerful information” that can be used to find lookalike audiences, she said.
Brands that don’t have their own first-party data can negotiate private marketplace deals with quality publishers that do have a known and engaged Hispanic audience, Nachtigall said.
TelevisaUnivision, for instance, created a Hispanic household audience graph last year to address mistargeting caused by undercounting Hispanic TV viewership within legacy TV panels. The graph includes data such as account logins, IP addresses and purchase information.
Music video publisher Vevo is another example of a programmer that uses its own data to help buyers target Hispanic audiences based on who’s watching what and where.
Vevo categorizes audiences based on viewing data and designated market area because music preferences vary widely based on country of origin. For example, an advertiser looking to reach a Mexican audience should target ads against Mexican cumbia-style music videos in California because Mexican Americans are heavily concentrated on the West Coast, said Rob Vélez, Vevo’s VP of inclusive network and LATAM sales.
Third parties are still invited (for now)
But that’s not to say there isn’t a place for third-party data in reaching multicultural audiences.
When paired with first-party insights, third-party data can create a fuller picture for advertisers to show how their target audiences are consuming content, including in what language.
Broadly speaking, first-generation Latinx audiences are more likely to speak predominantly Spanish, whereas younger, second- or third-generation Latinx demos are more likely to speak English.
But many Latinx audiences consume media in Spanish for cultural reasons, even if their dominant language is actually English. Yet people tend to surf the web and use their mobile devices in their preferred language, which is a good indicator of whether they’re more likely to respond to an English or Spanish version of an ad, according to Juan Romo, a paid search manager at Brandtech Media Group.
This behavioral tendency is what Sabio has in mind when it accesses third-party mobile data through app SDKs, such as mobile language preference, geolocation and device ID. It then ties this information to IP addresses for household-based targeting.
Vevo, meanwhile, overlays its own first-party data with census-level demographic data to figure out the ages of viewers because age is a fairly good indicator of someone’s primary language, Vélez said.
Don’t force it
But there’s more involved in reaching a Hispanic audience than precise targeting. Advertising needs to pique a consumer’s interest.
Translating Spanish-language creative into English in a literal, word-for-word fashion, for example, doesn’t cut it. (Translation: For the love of all that’s holy, stay away from Google Translate.)
Instead, brands need to “transcreate,” said Wpromote’s Ruedaquintero. Meaning that they should capture the spirit, intention and context of the message, adapting rather than translating verbatim.
Ruedaquintero pointed to the McDonald’s “I’m Lovin’ It” campaign as a strong example of transcreation. McDonald’s could have gone with “Yo lo amo” or “Yo lo quiero,” which technically translates to “I love it” or “I want it,” respectively. But because “lo” can also translate to “him” and both phrases can be associated with people, using those taglines could sow confusion. That’s why McDonald’s chose to go with “Me encanta,” which better captures the feel of really loving something.
Consumers can also tell when advertisers are forcing inclusivity into their marketing.
For example, Sabio’s Camacho said he saw an ad for an automotive manufacturer that tried a little too hard to represent a Hispanic family with the surname Garcia. The ad kept explicitly naming the Garcias to show the brand’s Latinx inclusivity, but it came off as gimmicky, he said.
This ad wasn’t meant for you
Even with more reliable data and the best of intentions, however, ad-serving glitches and retargeting errors can mess with campaign targeting – with the result sometimes being that a non-Hispanic consumer sees a Spanish-language ad.
Today, programmatic is often the culprit behind mistargeting, said Roberto Ruiz, EVP of research, insights and analytics at TelevisaUnivision.
But mistargeting isn’t a new phenomenon. It wasn’t uncommon back in the day for buyers to mistakenly place an English-language ad with a Spanish-language radio station because they were looking for reach and that station happened to be the top one in a given market.
In the online context, consumers might get retargeted with Spanish-language ads because they’ve engaged, either on purpose or in error, with an ad meant for Spanish-speaking audiences.
A user who clicks on the Spanish version of a Bad Bunny ad will likely be lumped together with affinity or lookalike audiences tied to Spanish language media.
“You may have engaged one time,” Ruedaquintero said, “but one time could be enough for a retargeting campaign to go after you.”
Unfortunately, there is no perfect solution to mistargeting.
But with the right approach, Ruedaquintero said, brands can get pretty close.
“There’s always an opportunity for error, even if it’s minimal,” she said. “Targeting is never 100% fine-tuned unless I’m delivering the ads directly to you, which is not gonna happen.”
