[{"data":1,"prerenderedAt":496},["ShallowReactive",2],{"guide-blended-roas-vs-platform-roas":3},{"id":4,"title":5,"answer":6,"authorId":7,"body":8,"category":407,"ctaVariant":408,"dataset":407,"description":409,"examples":410,"extension":411,"faqs":412,"heroImage":437,"intro":438,"meta":439,"navigation":440,"path":441,"publishedAt":442,"seo":443,"sources":444,"stats":467,"stem":494,"updatedAt":442,"__hash__":495},"blog\u002Fblog\u002Fblended-roas-vs-platform-roas.md","Blended ROAS vs Platform ROAS (2027)","Platform ROAS is the return a single ad platform like Meta reports inside its own dashboard, counting only the sales its attribution model credits to its own ads. Blended ROAS is total revenue divided by total ad spend across everything, with no attribution model involved. The two diverge because platforms self-report, count view-through and last-click credit, and often double-count the same order, a gap that widened after Apple's App Tracking Transparency cut ad click-throughs by an estimated 37.1% and forced Meta to model conversions it can no longer see. Use platform ROAS as the per-creative signal for what to scale or cut, and blended ROAS as the honest read on whether the business is actually growing.","likit-sae-lee",{"type":9,"value":10,"toc":395},"minimark",[11,16,20,23,26,30,33,42,45,48,51,119,122,125,129,132,135,138,141,144,147,151,154,157,160,168,171,175,178,261,264,268,271,274,277,281,284,287,290,293,374,377,380,384,387],[12,13,15],"h2",{"id":14},"two-roas-numbers-that-answer-two-different-questions","Two ROAS numbers that answer two different questions",[17,18,19],"p",{},"Return on ad spend looks like one metric, but the moment you run more than one channel there are really two of it, and they are not the same number. Platform ROAS is what an ad platform reports inside its own account. Open Meta Ads Manager and the ROAS column shows Meta-attributed purchase value divided by Meta spend, calculated with Meta's attribution model. Blended ROAS is total revenue from every source divided by total ad spend across every channel, pulled from your own books rather than any dashboard.",[17,21,22],{},"The distinction matters because the two are measured on different bases. Platform ROAS is narrow and self-reported: the platform decides which sales it caused, then grades itself on them. Blended ROAS is wide and independent: it counts every dollar that came into the business against every dollar you spent to acquire it, and it never asks a platform what it thinks it earned. One is a per-channel signal. The other is the closest thing you have to the truth of whether marketing is growing the company.",[17,24,25],{},"Most teams get into trouble by treating platform ROAS as if it were the second thing. They sum up the revenue each platform claims, feel good about the total, then wonder why the bank balance disagrees. The reconciliation is not a spreadsheet error. It is the predictable result of asking a self-interested reporter to mark its own homework, and it got materially worse after 2021. Understanding exactly where the gap comes from is what lets you use both numbers instead of being misled by either.",[12,27,29],{"id":28},"the-formulas-side-by-side","The formulas, side by side",[17,31,32],{},"Both formulas are simple. The difference is entirely in what goes into each one.",[17,34,35,36,41],{},"Platform ROAS, for a single channel, is the standard ",[37,38,40],"a",{"href":39},"\u002Fblog\u002Fhow-to-calculate-roas","ROAS calculation"," applied to one platform: the revenue that channel's attribution model credits to its ads divided by what you spent on that channel. In Meta Ads Manager the default attribution setting is 7-day click and 1-day view, meaning Meta credits a purchase to your ad if the person clicked within seven days before buying, or merely saw the ad within one day before buying without clicking. Change that window and the reported revenue changes with it, even though not a single real sale moved.",[17,43,44],{},"Blended ROAS is total revenue divided by total ad spend. Total revenue is every sale in the period, from paid, organic, email, direct, and referral, taken from your store or accounting system. Total ad spend is everything you paid across all platforms in the same window. There is no attribution model, no window setting, and nothing for a platform to inflate. Many operators call the same idea MER, the marketing efficiency ratio.",[17,46,47],{},"A worked example makes the divergence concrete. The numbers below are illustrative, chosen to show the mechanism rather than any real account.",[17,49,50],{},"Say a store books $200,000 in total revenue for the month and spends $50,000 on ads: $35,000 on Meta and $15,000 on Google. Now read what each dashboard claims.",[52,53,54,67],"table",{},[55,56,57],"thead",{},[58,59,60,64],"tr",{},[61,62,63],"th",{},"Line",[61,65,66],{},"Figure",[68,69,70,79,87,95,103,111],"tbody",{},[58,71,72,76],{},[73,74,75],"td",{},"Total store revenue (all channels)",[73,77,78],{},"$200,000",[58,80,81,84],{},[73,82,83],{},"Total ad spend",[73,85,86],{},"$50,000",[58,88,89,92],{},[73,90,91],{},"Meta Ads Manager reports 4.0x on its $35,000",[73,93,94],{},"claims $140,000 revenue",[58,96,97,100],{},[73,98,99],{},"Google Ads reports 5.0x on its $15,000",[73,101,102],{},"claims $75,000 revenue",[58,104,105,108],{},[73,106,107],{},"Sum of platform-claimed revenue",[73,109,110],{},"$215,000",[58,112,113,116],{},[73,114,115],{},"Blended ROAS ($200,000 \u002F $50,000)",[73,117,118],{},"4.0x",[17,120,121],{},"Look at the two claimed-revenue figures. The platforms together claim $215,000 of revenue in a month when the store only booked $200,000. Two dashboards have credited themselves with more sales than actually exist, and neither is wrong by its own rules. The blended ROAS, 4.0x, is the only figure the profit and loss statement can live with, because it is the one number no platform got to touch.",[17,123,124],{},"One refinement is worth knowing, because it catches a common objection. Blended ROAS counts all revenue, including repeat orders from existing customers who would have come back without any ad. For a mature brand with a large returning base, that can flatter the blended figure and hide a weak acquisition engine. The fix teams reach for is a new-customer version, sometimes written as new-customer blended ROAS or naMER, which puts first-order revenue from new customers over total ad spend. It answers a sharper question: is paid marketing actually buying growth, or just being credited for loyalty you already had? Track the standard blended figure for overall efficiency and the new-customer version when you specifically want to know whether acquisition is paying off. Neither one relies on a platform's attribution model, which is the whole point.",[12,126,128],{"id":127},"why-the-two-numbers-diverge-attribution-not-arithmetic","Why the two numbers diverge: attribution, not arithmetic",[17,130,131],{},"The gap is not a rounding problem. It is baked into how platform attribution works, and there are four causes worth naming.",[17,133,134],{},"The first is the walled garden. Each platform can only see the touchpoints that happen inside it. Meta knows a buyer clicked a Meta ad; it has no idea the same person later searched your brand on Google, opened two emails, and asked a friend before checking out. So it takes full credit for a journey it only witnessed one step of. Every platform does this, which is why their claims overlap.",[17,136,137],{},"The second is view-through attribution. With a 1-day view window still active, Meta credits itself when someone was served an ad, did not click, and bought within a day for reasons that may have nothing to do with the impression. A loyal customer who was going to reorder anyway scrolls past your ad in the morning and buys that afternoon, and the platform books a conversion it did not cause. View-through inflates platform ROAS the most for brands with strong repeat demand.",[17,139,140],{},"The third is double counting across channels. Because two platforms can each claim the same order, summing platform-attributed revenue overstates true paid-driven revenue. That is exactly what the worked example above shows: $215,000 claimed against $200,000 real. The more channels you run, the wider the overlap, and the more flattering the summed picture looks.",[17,142,143],{},"The fourth is that the reported number is increasingly modeled rather than observed, which is where iOS 14.5 comes in.",[17,145,146],{},"Underneath all four is an incentive worth saying plainly. A platform's default settings are set by the platform, and the generous ones make its ads look more effective. A 7-day click window credits more sales than a 1-day window; leaving view-through on credits more than turning it off. None of that is dishonest, and Meta documents every setting openly, but the defaults tilt toward the platform's own story about how much value it created. Blended ROAS has no such incentive because you own the inputs. That is the deeper reason to keep a self-reported number and an independent number side by side: not because one is a lie, but because only one of them is measured by a party with nothing to gain from the answer.",[12,148,150],{"id":149},"what-ios-145-and-att-actually-broke","What iOS 14.5 and ATT actually broke",[17,152,153],{},"Before April 2021, Meta could follow a large share of iPhone users deterministically from ad click to purchase using the device advertising identifier. Apple's App Tracking Transparency, launched with iOS 14.5, replaced that with an explicit prompt asking users to allow tracking, and users overwhelmingly declined. Flurry, sampling millions of daily active devices, reported that only about 4% of US users and roughly 11% worldwide opted in during the first weeks. The signal Meta had built its optimization and reporting on largely evaporated.",[17,155,156],{},"The measured damage was real. A University of Maryland study estimated that ATT cut ad click-throughs by 37.1%, because people were served less relevant ads, and that Meta ad spending fell 6.8% as advertisers shifted budget elsewhere. The same research found small direct-to-consumer firms that leaned heavily on Facebook ads suffered a roughly 60% revenue drop relative to firms less exposed to digital advertising, with new-customer acquisition hit harder than repeat orders.",[17,158,159],{},"To keep reporting anything at all for opted-out users, Meta moved to privacy-preserving aggregated measurement and statistical modeling. Instead of observing each conversion, the system now estimates a portion of them from aggregate patterns, with reporting delays and no demographic breakdowns for those events. The practical consequence for this guide is blunt: a growing share of the ROAS in your dashboard is modeled, not counted. Modeled conversions are useful for steering delivery, but they are an estimate, and estimates drift from ground truth in ways a bank statement never does.",[17,161,162,163,167],{},"The ",[37,164,166],{"href":165},"\u002Fblog\u002Ffacebook-ads-attribution-window","attribution windows"," themselves have also been shrinking, which changes reported ROAS without changing a single sale. On January 12, 2026, Meta stopped returning the 7-day-view and 28-day-view attribution windows in its reporting, leaving one day as the longest view-through window still available. A benchmark or an internal target built on the old longer windows will read differently against today's data, so any historical ROAS comparison has to account for the window it was measured on.",[17,169,170],{},"None of this makes platform ROAS useless. It makes it a noisier, self-reported, partly-estimated signal that systematically leans optimistic. That is a fine thing to steer creative with and a dangerous thing to run a P&L on.",[12,172,174],{"id":173},"what-each-number-can-and-cannot-tell-you","What each number can and cannot tell you",[17,176,177],{},"The cleanest way to hold both metrics in your head is to line up what each one is actually good for. They are complements, not rivals.",[52,179,180,193],{},[55,181,182],{},[58,183,184,187,190],{},[61,185,186],{},"Dimension",[61,188,189],{},"Platform ROAS (e.g. Meta Ads Manager)",[61,191,192],{},"Blended ROAS (MER)",[68,194,195,206,217,228,239,250],{},[58,196,197,200,203],{},[73,198,199],{},"What it measures",[73,201,202],{},"Revenue the platform's model credits to its own ads, over that platform's spend",[73,204,205],{},"All revenue from every source, over all ad spend",[58,207,208,211,214],{},[73,209,210],{},"Where the number comes from",[73,212,213],{},"The ad platform, self-reported",[73,215,216],{},"Your own store or accounting books",[58,218,219,222,225],{},[73,220,221],{},"Attribution involved",[73,223,224],{},"Last-touch-style within its walls, view-through, modeled for opted-out users",[73,226,227],{},"None; it ignores attribution entirely",[58,229,230,233,236],{},[73,231,232],{},"Granularity",[73,234,235],{},"Per campaign, ad set, and creative",[73,237,238],{},"Whole account or whole business",[58,240,241,244,247],{},[73,242,243],{},"Main blind spot",[73,245,246],{},"Cannot see other channels; double-counts; noisier and more modeled since ATT",[73,248,249],{},"Cannot tell you which ad or channel drove a sale",[58,251,252,255,258],{},[73,253,254],{},"Best used for",[73,256,257],{},"Relative decisions: which creative or campaign to scale or cut",[73,259,260],{},"Truth check: is the business growing efficiently overall",[17,262,263],{},"Read across the table and the division of labor is obvious. Platform ROAS is the only metric granular enough to compare creative A against creative B, because blended ROAS has no idea which ad did what. Blended ROAS is the only metric honest enough to tell you whether the account as a whole is paying for itself, because platform ROAS is structurally biased toward claiming it is.",[12,265,267],{"id":266},"how-to-use-each-without-fooling-yourself","How to use each without fooling yourself",[17,269,270],{},"Use platform ROAS as a relative signal, never as an absolute truth. Inside a single ad account, on a single attribution setting, the platform applies the same generous rules to every ad. That consistency is the point: if creative A shows a 5x and creative B shows a 2x under identical measurement, A really is outperforming B for the platform's optimization, even if neither absolute number is trustworthy. So lean on platform ROAS to rank creative, spot fatigue, and decide what to scale or kill. Just stop adding those numbers up and reporting the sum as revenue.",[17,272,273],{},"Use blended ROAS as the scoreboard for the business. Because it comes from your books and ignores attribution, it cannot be gamed by a window change or a modeled conversion. The most useful move any operator can make is to watch blended ROAS as spend changes. If you push an extra $10,000 into Meta and blended ROAS holds or rises, the spend is genuinely incremental. If Meta's dashboard shows a healthy 4x on the new spend but blended ROAS drops, the platform is harvesting sales that would have happened anyway, and the incremental return is far below what the dashboard claims. That divergence is the single most important thing platform ROAS hides and blended ROAS reveals.",[17,275,276],{},"This also protects you from a rising cost base that has nothing to do with your creative. Media keeps getting more expensive: Gupta Media's tracker put the blended Meta CPM near $8.19 across 2025, and Meta reported its average price per ad rose about 9% over full-year 2025. When the cost of reach climbs every year, a platform ROAS that looks stable can quietly mask an efficiency problem, because it is measured against inflated, partly-modeled revenue. Blended ROAS catches it, because the rising spend lands in the denominator and the real revenue lands in the numerator, with nothing in between.",[12,278,280],{"id":279},"reconciling-the-two-a-monthly-rhythm","Reconciling the two: a monthly rhythm",[17,282,283],{},"The goal is not to pick a winner between the two numbers. It is to run them together so each covers the other's blind spot. A simple monthly rhythm does the job.",[17,285,286],{},"Start with blended ROAS from your books, because that is the number the business is actually being paid. Then record what each platform reports for the same period. Now compute the relationship between them: how does blended ROAS compare to the summed platform-claimed revenue over total spend? That ratio is your over-attribution tell. If platforms collectively claim far more revenue than your store booked, they are double-counting heavily, and you should discount their absolute ROAS accordingly when you plan budgets.",[17,288,289],{},"Watch the ratio over time rather than obsessing over any single month. A stable relationship means your platform numbers, though inflated, are inflated consistently, so they remain a reliable relative signal. A ratio that suddenly widens, where platform ROAS holds up but blended ROAS falls, is an early warning that spend is buying less incremental revenue than the dashboards suggest, often a sign of audience saturation or seasonal repeat demand the platform is claiming credit for.",[17,291,292],{},"A short illustrative log shows what to look for. Again, the numbers are made up to demonstrate the read, not real account data.",[52,294,295,316],{},[55,296,297],{},[58,298,299,302,305,307,310,313],{},[61,300,301],{},"Month",[61,303,304],{},"Total revenue",[61,306,83],{},[61,308,309],{},"Summed platform-claimed revenue",[61,311,312],{},"Blended ROAS",[61,314,315],{},"Platform vs blended gap",[68,317,318,334,354],{},[58,319,320,323,325,327,329,331],{},[73,321,322],{},"Month 1",[73,324,78],{},[73,326,86],{},[73,328,110],{},[73,330,118],{},[73,332,333],{},"+8% over-claim",[58,335,336,339,342,345,348,351],{},[73,337,338],{},"Month 2",[73,340,341],{},"$230,000",[73,343,344],{},"$60,000",[73,346,347],{},"$250,000",[73,349,350],{},"3.8x",[73,352,353],{},"+9% over-claim",[58,355,356,359,362,365,368,371],{},[73,357,358],{},"Month 3",[73,360,361],{},"$235,000",[73,363,364],{},"$80,000",[73,366,367],{},"$320,000",[73,369,370],{},"2.9x",[73,372,373],{},"+36% over-claim",[17,375,376],{},"In months 1 and 2 the platforms over-claim by a steady 8 to 9%, so their dashboards, though inflated, track reality closely enough to trust as a relative signal. In month 3 spend jumps to $80,000, the dashboards still report strong per-channel ROAS, and yet blended ROAS collapses to 2.9x while the over-claim balloons to 36%. That is the pattern to act on: the extra spend added far less real revenue than the platforms took credit for. Without the blended column, month 3 would look like a good month. With it, you can see the account hit diminishing returns and cut back before the next cycle.",[17,378,379],{},"When you need to know a channel's true contribution, run a holdout. Pause or meaningfully cut one channel for a defined window and watch what happens to total revenue, not to the paused channel's dashboard. If total revenue barely moves, that channel was claiming more credit than it deserved. If it drops in step with the cut, the channel is genuinely incremental. A holdout is tool-agnostic, costs nothing but a little short-term revenue risk, and answers the one question attribution models argue about endlessly. Longer term, that same reasoning is what media-mix and incrementality methods formalize, but the monthly blended-versus-platform read plus the occasional holdout gets a small team most of the way there without any of the complexity.",[12,381,383],{"id":382},"the-number-that-pays-the-bills","The number that pays the bills",[17,385,386],{},"Platform ROAS and blended ROAS are not competing for your trust. They are two instruments on the same dashboard, and a pilot who reads only one flies blind. Platform ROAS is your altimeter for creative: precise enough to tell one ad from another, useful for every scale-or-cut decision, and biased optimistic in a way you learn to discount. Blended ROAS is your fuel gauge for the business: it does not care which engine burned the fuel, only whether you are still climbing.",[17,388,389,390,394],{},"The teams that grow steadily are the ones who let each number do its own job. They rank creative on platform ROAS, judge the business on blended ROAS, and treat a widening gap between the two as the signal to investigate rather than a rounding quirk to ignore. That habit matters more each year, as attribution windows keep shrinking, more of the reported number is modeled rather than observed, and the cost of reach keeps drifting upward. The dashboard will always tell you a flattering story about the sales it thinks it caused. Your books tell you the one that pays the bills. The discipline is reading the loop the same way every month, the same loop behind learning ",[37,391,393],{"href":392},"\u002Fblog\u002Fhow-to-run-a-facebook-ad","how to run a Facebook ad"," in the first place: research what is working, ship the next test, then check both numbers before you decide what it really did. A platform like AdPlay.ai keeps that loop in one place, but the discipline holds with any workflow, because the metric that grows a company is the one no platform gets to inflate.",{"title":396,"searchDepth":397,"depth":397,"links":398},"",2,[399,400,401,402,403,404,405,406],{"id":14,"depth":397,"text":15},{"id":28,"depth":397,"text":29},{"id":127,"depth":397,"text":128},{"id":149,"depth":397,"text":150},{"id":173,"depth":397,"text":174},{"id":266,"depth":397,"text":267},{"id":279,"depth":397,"text":280},{"id":382,"depth":397,"text":383},null,"neutral","Platform ROAS is what Meta claims; blended ROAS is total revenue over total ad spend. Why they diverge after iOS 14.5, and how to use each.",[],"md",[413,416,419,422,425,428,431,434],{"question":414,"answer":415},"What is the difference between blended ROAS and platform ROAS?","Platform ROAS is the return a single ad platform reports inside its own dashboard: Meta-attributed revenue divided by your Meta spend, using Meta's attribution model. Blended ROAS is total revenue from every source divided by total ad spend across every channel, with no attribution model involved. Platform ROAS is a per-channel, self-reported figure; blended ROAS is a whole-business figure that comes from your own books. They answer different questions, so they are not supposed to match.",{"question":417,"answer":418},"What is the blended ROAS formula?","Blended ROAS equals total revenue divided by total ad spend. Total revenue is every sale in the period from every channel (paid, organic, email, direct, referral), taken from your store or accounting system, not from any ad dashboard. Total ad spend is everything you paid across all platforms in the same period. Because it never asks a platform what it thinks it earned, no single channel can inflate it. Many operators call the same idea MER, the marketing efficiency ratio.",{"question":420,"answer":421},"Why does Meta report a higher ROAS than my store's actual revenue shows?","There are three structural reasons. Meta counts view-through conversions, crediting itself when someone saw an ad and bought later without clicking. It uses last-touch-style credit inside its own walls, so it cannot see that a buyer also came through Google, email, or a friend's recommendation. And since Apple's App Tracking Transparency, Meta models a share of conversions it can no longer observe directly. Add two platforms together and the same order often gets counted twice, so summed platform-claimed revenue can exceed the total your store actually booked.",{"question":423,"answer":424},"Did iOS 14.5 and ATT make platform-reported ROAS less accurate?","Yes. When iOS 14.5 launched Apple's App Tracking Transparency in April 2021, only about 4% of US users and roughly 11% worldwide opted in to tracking, per Flurry data. That cut the deterministic signal Meta relied on. A University of Maryland study estimated ATT reduced ad click-throughs by 37.1% and Meta ad spending by 6.8%. Meta now fills much of the gap with modeled and aggregated measurement, so a growing share of reported ROAS is estimated rather than directly observed.",{"question":426,"answer":427},"Should I optimize campaigns to platform ROAS or blended ROAS?","Use both, for different decisions. Platform ROAS is the relative signal for which creative, ad set, or campaign to scale or cut, because it is the only number granular enough to compare two ads against each other. Blended ROAS is the truth check on whether the whole account is actually growing efficiently. If platform ROAS climbs while blended ROAS falls, the platform is claiming credit it did not earn, and you are scaling a mirage.",{"question":429,"answer":430},"Is blended ROAS the same as MER (marketing efficiency ratio)?","Effectively yes. MER, the marketing efficiency ratio, is total revenue divided by total marketing spend, which is the same math as blended ROAS. Some teams draw a fine distinction, where MER may fold in non-ad marketing costs while blended ROAS counts strictly ad spend, but both exist to sidestep attribution and measure the business as a whole. Treat them as the same discipline: judge the total, not one platform's self-report.",{"question":432,"answer":433},"Is blended ROAS always lower than platform ROAS?","Not necessarily. Blended ROAS puts all revenue over all ad spend, so a brand with heavy organic, email, or repeat revenue can show a blended ROAS higher than any single platform reports. A store that depends almost entirely on paid acquisition, where platforms over-credit and double-count, will usually see summed platform ROAS look better than the blended figure. What is consistent is that the two measure different bases, so the ratio between them, tracked over time, is more useful than either number alone.",{"question":435,"answer":436},"How do I reconcile platform ROAS with blended ROAS each month?","Track the gap, not just the levels. Each month, note blended ROAS from your books and the platform ROAS each channel reports, then watch how the ratio between them moves. When you increase spend, the honest test is whether blended ROAS holds: if platform ROAS stays high but blended slips, the incremental spend is not adding incremental revenue. A short holdout, where you pause or cut a channel and watch total revenue, tells you more about true contribution than any dashboard.","\u002Fimages\u002Fblog\u002Fblended-roas-vs-platform-roas-hero.webp","You export Meta's numbers and the dashboard says a 4x return, then you open your store's books and the money does not add up. The ad platform is not lying, exactly. It is answering a narrower question than the one your profit and loss statement asks. Platform ROAS tells you what Meta's attribution model credits to its own ads; blended ROAS tells you whether every dollar of ad spend, across every channel, actually grew the business. This guide explains why the two numbers pull apart, what iOS 14.5 did to the gap, and how to use each without fooling yourself.",{},true,"\u002Fblog\u002Fblended-roas-vs-platform-roas","2027-03-19",{"title":5,"description":409},[445,449,453,457,460,463],{"label":446,"url":447,"year":448},"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","2024",{"label":450,"url":451,"year":452},"AppleInsider, Only 4% of iOS users in US opting in to ad tracking (Flurry data)","https:\u002F\u002Fappleinsider.com\u002Farticles\u002F21\u002F05\u002F07\u002Fonly-4-of-ios-users-in-us-are-opting-in-to-ad-tracking-report-says","2021",{"label":454,"url":455,"year":456},"Meta for Developers, Ads Insights API metric availability updates (attribution window changes effective January 12, 2026)","https:\u002F\u002Fdevelopers.facebook.com\u002Fblog\u002Fpost\u002F2025\u002F10\u002F16\u002Fads-insights-api-metric-availability-updates\u002F","2025",{"label":458,"url":459,"year":456},"Gupta Media, The True Cost of Social Media Ads (CPM tracker)","https:\u002F\u002Fwww.guptamedia.com\u002Fsocial-media-ads-cost",{"label":461,"url":462,"year":456},"Meta, Fourth Quarter and Full Year 2025 Results","https:\u002F\u002Finvestor.atmeta.com\u002Finvestor-news\u002Fpress-release-details\u002F2026\u002FMeta-Reports-Fourth-Quarter-and-Full-Year-2025-Results\u002Fdefault.aspx",{"label":464,"url":465,"year":466},"Meta Business Help Center, About Attribution Models and Attribution Settings","https:\u002F\u002Fwww.facebook.com\u002Fbusiness\u002Fhelp\u002F460276478298895","2026",[468,472,475,479,482,486,490],{"label":469,"value":470,"source":471},"US iOS users who opted in to app tracking after iOS 14.5","~4%","Flurry via AppleInsider, 2021",{"label":473,"value":474,"source":471},"Worldwide iOS users who opted in to app tracking after iOS 14.5","~11%",{"label":476,"value":477,"source":478},"Estimated drop in ad click-throughs caused by Apple's ATT","37.1%","University of Maryland, 2024",{"label":480,"value":481,"source":478},"Decline in Meta ad spending after the ATT rollout","6.8%",{"label":483,"value":484,"source":485},"Longest view-through attribution window after the Jan 2026 change","1 day","Meta for Developers, 2025",{"label":487,"value":488,"source":489},"Blended Meta (Facebook and Instagram) CPM, full-year 2025","$8.19","Gupta Media, 2025",{"label":491,"value":492,"source":493},"Meta average price per ad change, full-year 2025","+9%","Meta, 2025","blog\u002Fblended-roas-vs-platform-roas","xsvRcpGZik3JHqU2ub7ohsuv68iyK5-xvl3B_goWPj4",1786093699807]