[{"data":1,"prerenderedAt":399},["ShallowReactive",2],{"guide-target-competitor-audiences-facebook":3},{"id":4,"title":5,"answer":6,"authorId":7,"body":8,"category":290,"ctaVariant":291,"dataset":290,"description":292,"examples":293,"extension":294,"faqs":295,"heroImage":320,"intro":321,"meta":322,"navigation":323,"path":324,"publishedAt":325,"seo":326,"sources":327,"stats":368,"stem":397,"updatedAt":325,"__hash__":398},"blog\u002Fblog\u002Ftarget-competitor-audiences-facebook.md","Target Competitor Audiences on Facebook (2027)","You cannot point Facebook ads at the fans of a competitor's Page. Meta only ever allowed fan targeting for Pages you manage, and the interest options that once served as a rough proxy for a rival's audience have been thinned: Meta began merging thousands of specific interests into broad groups from June 23, 2025, and campaigns still using discontinued ones stopped delivering on January 15, 2026. The legitimate ways to reach a competitor's buyers now are interest proxies (adjacent brands, publications, and behaviors that still map to their audience), a lookalike built from your own customer list (sized 1% to 10% of a country), engagement and video-view custom audiences you capture from your own content, and broad targeting paired with creative strong enough that Meta's delivery finds those buyers for you. Validate any proxy by testing it in an isolated ad set against a broad and a lookalike control, and keep it only if it beats them on cost per result.","xanny-lee",{"type":9,"value":10,"toc":278},"minimark",[11,16,20,23,26,30,33,36,88,91,95,98,101,104,107,110,113,117,120,123,132,135,143,215,218,221,224,228,236,239,242,246,249,252,255,259,262,265,268,272,275],[12,13,15],"h2",{"id":14},"what-targeting-competitor-audiences-actually-means-now","What \"targeting competitor audiences\" actually means now",[17,18,19],"p",{},"Start by clearing up a myth that half the advice online still repeats. Facebook never let you point ads at the fans of another brand's Page. The only fan targeting it ever offered was Connections targeting, which reached people who liked a Page, installed an app, or joined an event that you administer, plus their friends. It was for your own assets, not a rival's, and Meta retired it as a standalone option long ago. So the phrase \"target competitor followers\" describes something that was never a real button.",[17,21,22],{},"What did exist, and what people actually mean, is interest targeting. For years, large brands, magazines, and public figures showed up as selectable interests inside detailed targeting. Choosing the interest reached people who had signalled interest in that brand by liking its Page or engaging with related content. That was the closest thing to reaching a competitor's audience, and it was always a loose approximation, not a list of the rival's customers.",[17,24,25],{},"That approximation has been shrinking. Meta has spent years pruning detailed targeting, and 2025 was the biggest cut yet. The practical result: aiming at a competitor's buyers in 2026 is no longer about finding the right interest. It is about assembling a set of proxies and lookalikes, then proving which of them actually reaches the people who buy in your category, the ones your competitor also sells to. The rest of this guide is that assembly, and the test that keeps you honest.",[12,27,29],{"id":28},"what-no-longer-works-and-the-timeline-that-killed-it","What no longer works, and the timeline that killed it",[17,31,32],{},"Detailed targeting has been narrowing on a schedule, and knowing the dates helps you understand why an interest you relied on last year has vanished. Meta first removed options tied to sensitive topics on January 19, 2022, pulling categories that referenced health, religion, sexual orientation, and political affiliation (Search Engine Land, 2022). Ad sets already using them kept running for a grace period, until March 17, 2022, then stopped. A second sensitive-topic round landed in early 2024, removing more health, race, and ethnicity options, with impacted ad sets running until March 18, 2024 (Social Media Today, 2024).",[17,34,35],{},"The change that matters most for competitor targeting came in 2025. From June 23, 2025, Meta began consolidating thousands of specific detailed-targeting interests into broad groups, folding niche brands, music genres, car models, and food categories into wider buckets (Meta Business Help Center, 2026). Campaigns created before that date that still relied on a now-discontinued interest stopped delivering on January 15, 2026, unless the advertiser edited the ad set. In plain terms, a lot of the granular brand-name interests that used to serve as competitor proxies were merged away or diluted into broad categories.",[37,38,39,52],"table",{},[40,41,42],"thead",{},[43,44,45,49],"tr",{},[46,47,48],"th",{},"Date",[46,50,51],{},"What changed",[53,54,55,64,72,80],"tbody",{},[43,56,57,61],{},[58,59,60],"td",{},"January 19, 2022",[58,62,63],{},"Sensitive-topic options removed (health, religion, sexual orientation, politics); impacted ads ran until March 17, 2022",[43,65,66,69],{},[58,67,68],{},"Early 2024",[58,70,71],{},"Further sensitive options removed (health, race, ethnicity); impacted ad sets ran until March 18, 2024",[43,73,74,77],{},[58,75,76],{},"June 23, 2025",[58,78,79],{},"Thousands of specific interests merged into broad groups (niche brands, genres, car models, foods)",[43,81,82,85],{},[58,83,84],{},"January 15, 2026",[58,86,87],{},"Older campaigns still using discontinued interests stopped delivering",[17,89,90],{},"There is a second, quieter shift that changes how even surviving interests behave. Under Advantage+ audience, Meta now treats your detailed-targeting inputs as suggestions rather than fixed walls. Only location and minimum age are hard constraints; interests, custom audiences, and lookalikes are directional hints the system can expand past when it predicts a better result (Meta Business Help Center, 2026). So even if a competitor proxy interest still exists, picking it no longer fences delivery to those people. It nudges the algorithm and then lets it roam. That is not a bug to fight; it is the model you now design around.",[12,92,94],{"id":93},"interest-proxies-reaching-a-rivals-buyers-without-a-rivals-name","Interest proxies: reaching a rival's buyers without a rival's name",[17,96,97],{},"The interest that names your exact competitor may be gone, but the people who buy from that competitor still leave signals Meta can read. A proxy is any targetable interest or behavior that a rival's customers disproportionately share. Assemble a handful and you approximate the audience without depending on a single brand interest that could be retired tomorrow.",[17,99,100],{},"Three kinds of proxy tend to survive the consolidation and still map to a specific competitor's audience:",[17,102,103],{},"Adjacent and larger brands. The very biggest names in a category often remain as interests even when smaller ones are merged away. If your direct rival is a mid-sized player, the category leader it competes against is frequently still selectable, and its audience overlaps heavily with your rival's. You are aiming at the neighbourhood, not the exact house.",[17,105,106],{},"Media and publications the audience follows. Buyers in a niche read the same magazines, follow the same review sites, and watch the same shows. An industry publication or a well-known review outlet in your space is often a cleaner, more stable proxy than any brand, because media interests were culled less aggressively and describe a mindset rather than a single purchase.",[17,108,109],{},"Behaviors, not just interests. Behaviors describe actions Meta observes, and some are strong purchase signals that no privacy change touched, because they happen on-platform. The engaged-shoppers behavior, for example, reaches people who recently tapped a shopping call-to-action on Facebook or Instagram. Layer that on top of a category interest and you tilt toward people who actually buy, not just browse.",[17,111,112],{},"Combine two or three of these in one ad set rather than betting everything on one. A single proxy interest is noisy; a small stack (a category-leader brand, plus an industry publication, plus an engaged-shoppers behavior) describes the rival's buyer more tightly than any one line could. Keep the stack narrow enough to be a real hypothesis you can test, and remember that under Advantage+ audience the stack is a suggestion the system will expand from anyway.",[12,114,116],{"id":115},"how-to-validate-that-a-proxy-actually-maps-to-the-competitors-buyers","How to validate that a proxy actually maps to the competitor's buyers",[17,118,119],{},"This is the step almost everyone skips, and it is the difference between targeting a competitor's audience and burning money on a coincidence. A proxy interest is a guess until the numbers back it. Three checks, in order of effort, tell you whether a proxy is real.",[17,121,122],{},"First, the size sanity check, which costs nothing. Before you spend, look at the audience size Meta estimates for the interest. If the interest supposedly represents a mid-sized brand's followers but shows an estimated audience of many tens of millions, it is diluted: Meta has bundled it into something far broader than the brand, and most of those people never heard of your competitor. A proxy whose size is wildly larger than the brand's plausible following will mostly reach non-buyers. Prefer proxies whose scale is in the same order of magnitude as the audience you are actually trying to reach.",[17,124,125,126,131],{},"Second, the overlap check against your own buyers. Build a customer-list custom audience from people who have already bought in your category (your own purchasers are the best available stand-in for the rival's purchasers, because they buy the same kind of thing). Then use Meta's ",[127,128,130],"a",{"href":129},"\u002Fblog\u002Ffacebook-audience-overlap","audience-overlap tool"," to compare that buyer audience against the proxy. A proxy that genuinely reaches category buyers should show meaningful overlap with people who already purchase from you. Near-zero overlap is a warning that the interest points somewhere else entirely.",[17,133,134],{},"Third, and decisively, the isolated live test. The other two checks are pre-flight; this one is the truth. Put the proxy in its own ad set, with ad-set budget optimization so it gets a guaranteed spend, against a broad control and a lookalike control that use identical creative and identical budget. Run all three for at least two weeks so each can gather enough conversions to leave the learning phase, then read cost per result. Keep the proxy only if it beats both controls. If broad or the lookalike wins, the proxy is not finding the right people, and you have just saved yourself a season of paying more to reach worse buyers.",[17,136,137,138,142],{},"That third check is worth a worked example, because the logic is the whole point. Say you sell a mid-priced skincare product and you want the customers of a larger rival brand. You set up three ad sets under the Sales objective, the same way you would ",[127,139,141],{"href":140},"\u002Fblog\u002Fhow-to-run-a-facebook-ad","build any Facebook ad campaign",", each with its own budget of $30 per day, the same video creative, and the same optimization event (purchase). You let them run 14 days, so each spends $420 and the test costs $1,260 total.",[37,144,145,164],{},[40,146,147],{},[43,148,149,152,155,158,161],{},[46,150,151],{},"Ad set",[46,153,154],{},"Audience",[46,156,157],{},"Spend (14 days)",[46,159,160],{},"Purchases",[46,162,163],{},"Cost per purchase",[53,165,166,183,199],{},[43,167,168,171,174,177,180],{},[58,169,170],{},"A: Proxy interests",[58,172,173],{},"Category-leader brand + industry publication + engaged shoppers",[58,175,176],{},"$420",[58,178,179],{},"14",[58,181,182],{},"$30.00",[43,184,185,188,191,193,196],{},[58,186,187],{},"B: Value-based lookalike",[58,189,190],{},"1% lookalike off your top 25% of customers by revenue",[58,192,176],{},[58,194,195],{},"21",[58,197,198],{},"$20.00",[43,200,201,204,207,209,212],{},[58,202,203],{},"C: Broad",[58,205,206],{},"Location and age only, Advantage+ audience",[58,208,176],{},[58,210,211],{},"18",[58,213,214],{},"$23.33",[17,216,217],{},"Read the table across, not down a single column. The proxy stack (A) technically works: it produced sales. But it is the most expensive route at $30 per purchase, which tells you the interests are diluted or are reaching browsers rather than buyers. The value-based lookalike (B) wins clearly at $20, because it is built from the people who actually spend with you, and those people behave like the category's buyers, your rival's included. Broad (C) lands in between at $23.33 with no manual audience work at all, which is exactly what you would expect as Meta's delivery leans on creative and conversion signal.",[17,219,220],{},"The decision writes itself. Shift budget to the lookalike, keep broad as a scale lane, and either drop the proxy interests or hold them as a small experiment for the next creative. Notice what the test did not require: naming the competitor, scraping anything, or trusting a single interest. It let cost per result rank the routes for you. Rerun the same three-way comparison whenever you launch a genuinely new creative concept, because the winner can change with the ad, not just the audience.",[17,222,223],{},"The costs in that example are illustrative, but the reference points are real. WordStream's 2025 benchmarks put the all-industry cost per click at $1.92 on Leads campaigns, and Gupta Media's tracker put the blended Meta CPM near $8.19 for the year, so a $20 to $30 cost per purchase is a plausible band for a considered product. Your own numbers will differ; what matters is that the three routes are judged against each other on the same creative and budget.",[12,225,227],{"id":226},"lookalikes-off-your-own-list-the-most-dependable-route","Lookalikes off your own list: the most dependable route",[17,229,230,231,235],{},"The example already hinted at why the lookalike usually wins, and it deserves its own section because it is the sturdiest way to reach a competitor's kind of customer. A ",[127,232,234],{"href":233},"\u002Fblog\u002Ffacebook-lookalike-audiences","lookalike audience"," is new people Meta finds who resemble a source you give it. Seed it with your own best buyers and you are effectively saying, find more people who behave like the customers I already convert, which, in a shared category, is the same population your rival sells to.",[17,237,238],{},"The mechanics are simple and worth getting exactly right. The source needs at least 100 people from a single country, though Meta recommends 1,000 to 5,000, and seed quality beats raw count (Meta Business Help Center, 2026). The size slider runs from 1% to 10% of the target country's population, where 1% is the closest match to your source and 10% is the broadest reach. For competitor-grade precision, seed narrow and stay tight: build a value-based source from your highest-revenue customers rather than every buyer, then take a 1% to 2% lookalike off that. Value-based sourcing models who spends the most, not just who converted once, so the match points at the buyers worth stealing.",[17,240,241],{},"Lookalikes also sidestep the fragility that broke interest proxies. An interest can be merged away on Meta's schedule; a lookalike rebuilt off your live customer list refreshes with your business and never depends on Meta keeping a specific brand in its taxonomy. That durability matters more since Apple's App Tracking Transparency thinned the signal Meta receives: a University of Maryland study estimated ATT cut ad click-throughs by roughly 37% (University of Maryland, 2024), which raised the value of first-party sources you own outright. Your customer list is that first-party source, and it is the one input no privacy change can take from you.",[12,243,245],{"id":244},"engagement-and-video-view-custom-audiences-capture-the-interest-you-create","Engagement and video-view custom audiences: capture the interest you create",[17,247,248],{},"There is a route that turns competitor curiosity into a targetable audience you own, and it needs no pixel and no customer list. When you publish content that speaks to a rival's customer (an honest comparison, a switch offer, a demo that answers the objection your competitor cannot), the people who watch or interact are declaring themselves. An engagement custom audience captures them.",[17,250,251],{},"The strongest sources are video viewers segmented by watch depth and profile engagers. You can build an audience of people who watched a set percentage of your video, or who engaged with your Facebook Page or Instagram account, entirely from activity inside Meta. Because that signal lives on-platform, it is untouched by browser privacy changes, and the retention windows are generous: video, Page, and Instagram engagement audiences hold people for up to 365 days, while lead-form audiences hold up to 90 (Meta Business Help Center, 2026). A year-long window means a comparison video you run this quarter keeps feeding a retargetable pool for the next twelve months.",[17,253,254],{},"The competitor angle here is specific. Run a piece of creative built around the exact reason someone might leave your rival, then capture the 25% and 50% video viewers as an engagement audience and retarget them with a harder offer. You are not guessing at who follows the competitor; you are collecting the people who leaned in when you addressed that competitor's weakness. That warm audience will be small, so use it for the closing step of the funnel, and feed the same engagers into a lookalike seed once the pool is large enough to compound the effect.",[12,256,258],{"id":257},"let-broad-plus-strong-creative-find-them","Let broad plus strong creative find them",[17,260,261],{},"The uncomfortable truth running under all of this is that Meta would rather do the targeting than let you do it. Its delivery system now reads your creative and your conversion signal to decide who sees an ad, and it does this at a scale no manual interest stack can match. In November 2025 Meta detailed GEM, a ranking model that scores retrieved ads, and reported it lifted ad conversions by about 5% on Instagram and 3% on Facebook Feed (Engineering at Meta, 2025). That is the system getting better at finding your buyers on its own, including the ones who currently buy from your rival.",[17,263,264],{},"So the highest-leverage version of competitor targeting in 2026 is often the least literal. Run broad (location and age as the only real constraints), feed clean conversion data through the Pixel and Conversions API, and put all your effort into creative that a competitor's customer cannot scroll past. Broad does not mean vague. Write the hook to the exact person who buys from your rival and name their unmet need: the feature the competitor charges extra for, the wait time it makes them tolerate, the result it never quite delivers. The creative does the aiming that interests used to, and it does it in a way Meta rewards rather than deprecates.",[17,266,267],{},"Use the other routes as scaffolding around that broad core. Proxies and lookalikes are how you seed early delivery and how you run the validation test above; broad with sharp creative is usually how you scale the winner. The three are not rivals. They are a sequence: seed and test with proxies and lookalikes, prove which reaches the right people, then hand the winning creative to broad delivery and let the system compound it.",[12,269,271],{"id":270},"putting-it-together","Putting it together",[17,273,274],{},"The single button that promised a competitor's followers was always a myth, and the interest that stood in for it is fading on Meta's own schedule. That is not a loss so much as a nudge toward methods that were always sturdier. Reach a rival's buyers in 2026 by stacking a few honest proxies, by building lookalikes off the customers you already convert, by capturing the engagers your comparison content earns, and by trusting broad delivery when your creative is sharp enough to deserve it. Then let one rule govern all of them: never trust a route you have not tested against a control on the same creative and budget.",[17,276,277],{},"The work that never stops is the creative. Whether the winning audience turns out to be a lookalike, a proxy stack, or plain broad, the ad is what convinces a competitor's customer to switch, and the fastest lever on cost is a steady supply of fresh, competitor-aware creative rather than another audience tweak. A platform like AdPlay.ai keeps the research, the creative, the launch, and the read in one place, but the discipline holds with any workflow: study what the category's buyers respond to, make the ad that answers it, launch, read the result, and feed it into the next test. Audiences find themselves when the creative earns them.",{"title":279,"searchDepth":280,"depth":280,"links":281},"",2,[282,283,284,285,286,287,288,289],{"id":14,"depth":280,"text":15},{"id":28,"depth":280,"text":29},{"id":93,"depth":280,"text":94},{"id":115,"depth":280,"text":116},{"id":226,"depth":280,"text":227},{"id":244,"depth":280,"text":245},{"id":257,"depth":280,"text":258},{"id":270,"depth":280,"text":271},null,"neutral","Direct fan targeting is gone. How to reach a competitor's followers and buyers on Facebook in 2026 with interest proxies, lookalikes, and broad targeting.",[],"md",[296,299,302,305,308,311,314,317],{"question":297,"answer":298},"Can I target a competitor's Facebook followers directly?","No, and you never could. Meta's fan or Connections targeting only ever let you reach people who liked a Page, app, or event that you administer, not a competitor's. The closest thing was detailed targeting, where some large brands and publications appeared as selectable interests, so choosing that interest reached people who had shown interest in the brand. Even that proxy has narrowed since Meta began consolidating thousands of specific interests into broad groups on June 23, 2025, so the reliable routes today are interest proxies, lookalikes off your own customers, engagement audiences, and broad targeting.",{"question":300,"answer":301},"How do I reach people who buy from my competitors on Facebook ads?","Combine four legitimate methods. Use interest proxies (adjacent brands, industry publications, and behaviors like engaged shoppers that a rival's buyers also share). Build a lookalike from your own customer list, because your best buyers resemble the category's buyers, including your competitor's. Capture the people who engage with your own comparison or switch content as an engagement custom audience and retarget them. And run broad targeting with creative that speaks directly to a rival's customer, letting Meta's delivery system find them. Test the four against each other and keep what wins on cost per result.",{"question":303,"answer":304},"Why did my competitor interest disappear from Facebook detailed targeting?","Meta has been culling detailed targeting for years. It removed sensitive-topic options in January 2022 and again in early 2024, then from June 23, 2025 it began merging thousands of specific interests, including niche brands, music genres, car models, and foods, into broad groups. Campaigns created before that date that still used a discontinued interest stopped delivering on January 15, 2026. If your interest vanished, Meta usually suggests a broader replacement, but the sharper move is to switch to a lookalike off your customer list or to broad targeting with better creative.",{"question":306,"answer":307},"Do interest proxies still work for competitor targeting in 2026?","They can, but they are weaker and less literal than they used to be. Meta now treats detailed targeting inputs as suggestions rather than hard walls under Advantage+ audience, where only location and minimum age are true constraints, so the system may deliver beyond an interest you pick. Surviving proxies (a large adjacent brand, an industry magazine, an engaged-shoppers behavior) still nudge delivery in the right direction. Treat any proxy as a hypothesis and validate it in an isolated test before you trust it.",{"question":309,"answer":310},"What is the most reliable way to reach competitor customers now?","A value-based lookalike seeded from your own highest-value customers is usually the most dependable. Your best buyers behave like the category's buyers, so a tight 1% lookalike off them tends to reach the same kind of person your competitor sells to, without depending on a specific interest that Meta might retire. The source needs at least 100 people from one country, though 1,000 to 5,000 gives a cleaner signal, and value-based sources model revenue rather than raw count.",{"question":312,"answer":313},"How do I know if a proxy interest actually reaches a competitor's buyers?","Run three checks. First, a size sanity check, because an interest far larger than the brand's real following is diluted and mostly reaches non-buyers. Second, an overlap check against a custom audience of your own buyers, since a genuine proxy should overlap with people who already purchase in your category. Third, an isolated live test: put the proxy in its own ad set against a broad control and a lookalike control with identical creative and budget, run it two weeks, and keep it only if it beats both on cost per result. If it loses, the interest is not finding the right people.",{"question":315,"answer":316},"Is it against Facebook policy to target a competitor's brand as an interest?","Selecting an interest that Meta itself offers in detailed targeting is allowed, since you are choosing from Meta's own list, not scraping the competitor. What the rules restrict is the creative: you cannot imply a false affiliation with, or endorsement by, another brand, and you cannot misuse a trademark to deceive people about who is advertising. Aim your delivery at people the proxy reaches, but keep your ad clearly your own brand's message.",{"question":318,"answer":319},"Should I just use broad targeting instead of chasing competitor audiences?","Often, yes. Meta's delivery leans harder on creative and conversion signal every year, and its GEM ranking model raised Instagram ad conversions by about 5% in 2025, so broad targeting with a strong, competitor-aware creative frequently beats a hand-built interest stack. Broad does not mean aimless: write the ad to the exact person who buys from your rival, feed clean conversion data, and let the system find them. Use proxies and lookalikes to seed and to test, then scale whatever the numbers reward.","\u002Fimages\u002Fblog\u002Ftarget-competitor-audiences-facebook-hero.webp","You know exactly which brand is eating your lunch, and you want your ads in front of their customers. The old shortcut, picking a rival off an interest list and pointing spend at their fans, barely exists anymore, and the version most guides still describe was never really there in the first place. What is left is a set of legitimate, durable ways to reach the same people: proxies for their audience, lookalikes off your own buyers, the engagement you capture, and broad delivery aimed by strong creative. This guide walks each one, and how to prove a proxy actually maps to the right people before you spend on it.",{},true,"\u002Fblog\u002Ftarget-competitor-audiences-facebook","2027-06-16",{"title":5,"description":292},[328,332,335,339,343,346,349,352,355,358,362,365],{"label":329,"url":330,"year":331},"Meta Business Help Center, About Detailed Targeting","https:\u002F\u002Fwww.facebook.com\u002Fbusiness\u002Fhelp\u002F182371508761821","2026",{"label":333,"url":334,"year":331},"Meta Business Help Center, Updates to Detailed Targeting","https:\u002F\u002Fwww.facebook.com\u002Fbusiness\u002Fhelp\u002F458835214668072",{"label":336,"url":337,"year":338},"Search Engine Land, Meta will remove targeting options for sensitive topics on January 19","https:\u002F\u002Fsearchengineland.com\u002Fmeta-will-remove-targeting-options-for-sensitive-topics-on-january-19-378095","2022",{"label":340,"url":341,"year":342},"Social Media Today, Meta to Remove More Detailed Targeting Options for Ad Campaigns","https:\u002F\u002Fwww.socialmediatoday.com\u002Fnews\u002Fmetas-removing-detailed-targeting-options-ad-campaigns\u002F704375\u002F","2024",{"label":344,"url":345,"year":331},"Meta Business Help Center, About Lookalike Audiences","https:\u002F\u002Fwww.facebook.com\u002Fbusiness\u002Fhelp\u002F164749007013531",{"label":347,"url":348,"year":331},"Meta Business Help Center, About Custom Audiences","https:\u002F\u002Fwww.facebook.com\u002Fbusiness\u002Fhelp\u002F744354708981227",{"label":350,"url":351,"year":331},"Meta Business Help Center, How long people stay in a custom audience","https:\u002F\u002Fwww.facebook.com\u002Fbusiness\u002Fhelp\u002F419552341510847",{"label":353,"url":354,"year":331},"Meta Business Help Center, About Advantage+ Audience","https:\u002F\u002Fwww.facebook.com\u002Fbusiness\u002Fhelp\u002F273363992030035",{"label":356,"url":357,"year":342},"University of Maryland Smith School, Small Businesses Take Big Hit from Apple's Privacy Regulation","https:\u002F\u002Fwww.rhsmith.umd.edu\u002Fresearch\u002Fsmall-businesses-take-big-hit-apples-privacy-regulation",{"label":359,"url":360,"year":361},"Gupta Media, The True Cost of Social Media Ads (CPM Tracker)","https:\u002F\u002Fwww.guptamedia.com\u002Fsocial-media-ads-cost","2025",{"label":363,"url":364,"year":361},"WordStream \u002F LocaliQ, Facebook Ads Benchmarks 2025","https:\u002F\u002Fwww.wordstream.com\u002Fblog\u002Ffacebook-ads-benchmarks-2025",{"label":366,"url":367,"year":361},"Engineering at Meta, GEM Generative Ads Recommendation Model","https:\u002F\u002Fengineering.fb.com\u002F2025\u002F11\u002F10\u002Fml-applications\u002Fmetas-generative-ads-model-gem-the-central-brain-accelerating-ads-recommendation-ai-innovation\u002F",[369,373,376,379,381,385,389,393],{"label":370,"value":371,"source":372},"Lookalike audience size range (share of a country's population)","1%-10%","Meta Business Help Center, 2026",{"label":374,"value":375,"source":372},"Minimum source size to build a lookalike audience","100 people",{"label":377,"value":378,"source":372},"Engagement custom audience retention window (video, Page, Instagram)","up to 365 days",{"label":380,"value":76,"source":372},"Date Meta began merging specific interests into broad groups",{"label":382,"value":383,"source":384},"Estimated drop in ad click-throughs from Apple's ATT","~37%","University of Maryland, 2024",{"label":386,"value":387,"source":388},"Blended Meta (Facebook and Instagram) CPM, full year","$8.19","Gupta Media, 2025",{"label":390,"value":391,"source":392},"All-industry Facebook CPC, Leads objective","$1.92","WordStream, 2025",{"label":394,"value":395,"source":396},"Ad conversion lift from Meta's GEM ranking model on Instagram","+5%","Engineering at Meta, 2025","blog\u002Ftarget-competitor-audiences-facebook","WDqjxaNCoISAzZcWWZZ5chIqVrzL-pypr6GhcdtIcbw",1786093700573]