Meta advertisers have heard a lot about Advantage+ this year.
Advantage+ is not a new streaming service, despite the “plus.” And it’s not a healthcare plan, though it sounds like one.
It’s Meta’s version of a first-party data-based ad platform locked inside the blackest black box any Facebook advertiser has ever known, which it calls Advantage+ Shopping Campaigns (ASC).
If “the blackest black box” rings a bell, it’s because an AdExchanger report last week described Google’s Performance Max as the blackest black box in the history of Google ad products.
This isn’t self-plagiarism; it’s a trend.
Google PMax and Meta ASC are “kindred spirits,” according to Mike Ryan, head of ecommerce insights at Smarter Ecommerce, a retail tech company.
Their kinship is based on the unprecedented self-control Google and Meta have over those campaigns, and the lack of analytics or creative insights for advertisers.
Google PMax has been available to all advertisers for a little over a year, and was in a beta for a year before that. Meta ASC launched in a closed beta earlier this year with DTC and retail brands, and in September began rolling out to other advertiser verticals and accounts that request the product.
In talks with more than 20 agency and brand advertisers who have used ASC and unanimously approve, the clear conclusion is that Meta and Google are correct in betting that marketers will opt for ROAS over transparency and control. It’s only a matter of time before this approach becomes more widely adopted.
Machines are truly taking over the advertising world.
How it works
Meta ASC advertisers upload their creative assets, their conversion goals and acquisition rates, their daily budget cap and … that’s pretty much it. Meta returns the total amount spent and the revenue it attributes to the campaign. That’s pretty much it.
Advertisers bring in server-side connections to their sales data – another tactic that follows in Google’s footsteps. Meta can no longer rely on its pixel and SDK to connect impressions to conversions. Apple severed most of those links. But Apple can’t block a company from piping its CRM, CDP or data warehouse to Meta.
The most important thing PMax and ASC have in common: ad buyers opting for performance over granular targeting controls and analytics.
“ASC is one of the best features that Meta introduced to the advertising ecosystem in the last five years,” Rok Hladnik, managing partner of Flat Circle, a boutique social ad agency, told AdExchanger.
“This is the best product release that Facebook has had in a long, long time for advertisers,” said Ben Yahalom, president and cofounder of True Classic, a men’s clothing maker, and previously a Facebook client partner for advertising. The company added ASC a few months ago, and it’s already about 40% of total ad spend on Meta, he said.
The pros and cons
“The cons” are actually just one big negative for ASC. It returns next to no data or insights.
In the pre-existing campaign system for buying Facebook and Instagram ads, which were recently renamed “manual campaigns” (for an ad platform, that’s like saying “don’t use this”), advertisers had finer controls over interest-based targeting parameters. They could hone in closely on things like location data, creative optimization and lookalike audiences.
Advertisers couldn’t track individuals, but learned a great deal about their audience and which creative elements had high engagement or conversion rates.
An outdoor gear manufacturer might have different interest-based targeting for hunters, fishermen, hiking enthusiasts or fashion-minded customers, said Milo McMahon, founder and lead strategist of an agency called Outdoor Ecommerce. Even without exporting user-level data, the brand might know its campaign resonated for, say, fishing creative on Facebook and recreational hikers on Instagram.
That feedback is gone. With ASC, Meta does the pattern-matching based on its first-party identity graph. The platform reveals nothing about what creative is used and where it was served, or what interest-based signals led to conversions.
McMahon said Meta reps actively encourage advertisers to set campaign parameters as wide as possible.
As in, “The United States.”
Since brands don’t want to overspend on known existing customers at the expense of potential new customers, Meta does offer a control over who it targets. Advertisers can match a customer list to Meta, and then put a percent cap on the daily budget for that audience.
But any campaign restrictions, even what an advertiser might think of as a broad lookalike model of its customers, from ASC’s point of view is like a bird stuck in an airport.
“Audience targeting in the ASC world wants to go broad by default,” Yahalom said.
By channeling advertisers away from narrow campaigns with self-selected audiences or a specific location, he said Meta “is preventing advertisers from hurting themselves, unintentionally, by superimposing settings that are restricting the algorithm.”
And it is important to note that a tunnel-vision focus on “performance” is not the only selling point for ASC.
With manual campaigns, only a handful of creative assets can be effectively optimized. With so little signal data returning, campaigns can’t be sliced and diced for optimization like during the Facebook heyday. To build a statistically significant level of data to optimize creative, the campaign must focus on fewer creative elements.
ASC brings the creative elements in the optimization pool from four or five, perhaps, to 100-150 creative variations, according to Yahalom.
Brands also don’t necessarily want to go back to nitty-gritty control of Facebook ad buying being a core marketer competency, said Cody Plofker, CMO of the DTC cosmetics brand Jones Road Beauty. Brand-building efforts, creative and content, landing page testing and an understanding of profitable customer acquisition rates are the preferred core competencies, he said.
When it comes to the ad targeting nuances, the idea is to let the first-party data owner who knows best (that would be Meta) do its thing.
“[Platforms] have gone from manual media-buying to a more commoditized model, with the platform taking control,” Plofker said. The black box is a strike against ASC, he said, “but it pales in comparison to the advantages.”
Difference from PMax
Meta ASC has important distinctions from Google PMax. The biggest difference at the moment is maturity. PMax has been available for longer and to more advertisers. Its feature set has developed a bit more than ASC, which started a closed beta this year and has been generally available for only a few months.
“We pushed like crazy to get access,” Yahalom said.
Hladnik said that, for advertiser accounts with access for the past couple months and enough site conversion data for the optimization engine to run, “ASC is performing really well.”
And Hladnik said ASC campaigns scale more effectively than PMax.
That said, the easy scalability could be an early bird advantage. ASC campaigns could benefit from the fact that fewer advertisers use it, and as that total number doubles or triples, the advantage could go away.
ASC also doesn’t use cost cap bidding, which is a way of averaging out bids of different price ranges to meet an overall acquisition rate. That is available with Meta’s manual campaigns.
Cost cap bidding means that a brand with a daily budget cap of $10,000 that’s having a tough sales week may not fill anything like $10,000 per day. The cap is considered a ceiling, but the bidding strategy is focused on a profitable conversion metric, so if the company stops converting sales at the usual rates, a cost-capped campaign won’t just spend and spend.
With ASC, that same campaign is going to spend $10,000 every day, McMahon said.
For advertisers scaling from thousands of dollars spent on ASC ads per day to tens of thousands of dollars or more, the lack of a cost cap increases the risk that a bad sales stretch will be compounded by continuously high and unprofitable spending the brand can’t see or control.
PMax, a more mature product by the standards of these new obsidian boxes, does have cost caps. McMahon said it’s likely Meta rolls them out once it can rationalize data from Meta’s apps, Facebook and Instagram Shops (its on-platform payment service) and advertisers’ own sites and apps.
“When they do, it’s going to be a huge unlock for ASC,” he said.
Meta is only just starting to incorporate Facebook Shops and Instagram Shops data into ASC, according to advertisers. Though that observation is currently speculation, since the data baked into the ASC machine learning is, of course, undisclosed.
“Completely a hunch,” Yahalom said, “but one big piece that I think they’re starting to leverage now that they haven’t before is their own Shops data.”
ASC’s lower risk profile
ASC may be a black box akin to PMax, but it’s not the same kind of dangerous hot potato that Google is dealing with. That’s partly because ASC spans Facebook, Instagram, Messenger and the Meta Audience Network, while PMax requires advertisers to agree to blindly serve ads across email, mapping, search, sponsored product links, the entire Google Display Network and all YouTube varieties, including YouTube TV.
Facebook advertisers are not up in arms about possibly being forced into Instagram or Messenger. If ASC forced advertisers to the metaverse, it would be different.
Google has more sensitive access to online activity and spans a broader array of inventory types. PMax is also built on search, which is a more controversial data source for advertisers. PMax bids aggressively on branded search terms. But many marketers philosophically oppose buying their own terms, or eschew the tactic because they spent a lot of money testing it and found it unprofitable.
The machine learning trend
Google PMax and Meta ASC give the platforms wide license over targeting, bid prices and creative optimization.
But there’s widespread acceptance of these products by advertisers because they save time and improve results. Advertisers also see this as where the world is moving. And not just for Google and Meta.
Criteo, for instance, released a product in September called Commerce Max. Its distinction is that it relies on Criteo’s machine learning and serves campaigns across a wider swath of the open web and retailer sites in Criteo’s network than an advertiser would typically buy. (Sound familiar? It even borrowed part of PMax’s name.)
“Criteo’s predictive AI identifies the best path to conversion, removing guesswork and leading to better advertising outcomes,” according to the release.
And The Trade Desk has a product called Koa, an optimization engine that gives the DSP more control over targeting choices and data marketplace purchases. Koa also targets a conversion performance goal.
The Trade Desk and Criteo still produce log files that advertisers can audit and disclose creative used per impression. They also don’t own the media taking over media plans. But taking decisions away from advertisers and putting ad platform machine learning in charge of campaigns has achieved a level of acceptability, even popularity.
Agency implications
While the advertiser feedback to ASC has been enthusiastic, the tech puts agencies in an awkward spot.
“It is usurping the audience choices that we used to make,” McMahon said of ASC. “But, as a direct marketer, I don’t even need to know the fundamental mechanics behind it, if I understand whether it works or not.”
He said the role of social or performance agencies is shifting away from keen hands-on-keyboard knowledge of platforms like Facebook, YouTube or TikTok.
Before, agencies advised on targeting. How to hack the trends and tricks of particular ad platforms, dial up or down specific bids and select people to target were formerly huge markets for agency services, Yahalom said.
Now, agencies and marketers are ceding control over targeting to Google and Meta’s AI. But there’s a whole new services market in helping advertisers judge black box ad platforms like PMax or ASC as standalone channels.
Agencies continue to advise on creative strategy. And they can reconcile ad platforms on a channel level. Although no one can crack open PMax or ASC to understand what’s inside, agencies can measure the spend as a standalone channel, alongside OOH, TV or radio in a media mix modeling report.
For instance, True Classic retains a media agency and uses multiple platform measurement services startups, including Northbeam, Triple Whale and Measured, he said.
Even as machines take over the advertising world, agencies are persevering (at least for now), as marketers seek to understand the new AI-driven ad platform model.
