[{"data":1,"prerenderedAt":452},["ShallowReactive",2],{"guide-facebook-ad-image-text-overlay":3},{"id":4,"title":5,"answer":6,"authorId":7,"body":8,"category":351,"ctaVariant":352,"dataset":351,"description":353,"examples":354,"extension":355,"faqs":356,"heroImage":381,"intro":382,"meta":383,"navigation":384,"path":385,"publishedAt":386,"seo":387,"sources":388,"stats":418,"stem":450,"updatedAt":386,"__hash__":451},"blog\u002Fblog\u002Ffacebook-ad-image-text-overlay.md","Facebook Ad Image Text: Is the 20% Rule Gone?","Facebook retired the 20% text rule and its Text Overlay grid tool in September 2020, so ads are no longer rejected or capped for how much text sits on the image. Heavy on-image text is still a bad bet: it tends to earn lower engagement, which quietly raises your cost against a blended Meta CPM of about $8.19 (Gupta Media, 2025), and it is hard to read on the mobile feed where 98.5% of Facebook users are (Statista, 2022). The practical rule now is to keep on-image text short and legible, move the detail into the primary text and headline fields, and test a text-light version against a text-heavy one so the auction, not a folk memory, decides.","xanny-lee",{"type":9,"value":10,"toc":339},"minimark",[11,16,20,23,26,30,33,36,39,42,45,49,58,61,115,118,121,125,128,131,207,210,213,217,220,223,226,229,233,241,302,305,308,312,315,318,321,329,333,336],[12,13,15],"h2",{"id":14},"the-short-answer-the-rule-is-gone-the-preference-is-not","The short answer: the rule is gone, the preference is not",[17,18,19],"p",{},"Here is the myth, stated plainly so it can be put down. Facebook used to reject or throttle any ad image where text covered more than 20% of the picture, and it made you run every image through a grid tool to prove you were under the line. That rule is dead. Meta retired it in September 2020, and the company told advertisers directly that it would \"no longer penalize ads with higher amounts of image text in auctions and delivery.\" The Text Overlay tool that policed the limit was pulled at the same time, and its page now redirects to nothing. If a designer or a well-meaning blog still tells you your headline is too big for the grid, they are quoting a rule that has not existed for years.",[17,21,22],{},"So you can put as much text on an ad image as you like, and Facebook will not stop you. That is the whole of the good news, and it is genuine: bold headlines, price stamps, and callouts that used to fail the checker now sail through review untouched.",[17,24,25],{},"The catch is that \"allowed\" and \"advisable\" are different things, and the gap between them is where budgets leak. Meta itself did not stop believing that clean creative wins. Its Business Help Center still tells advertisers that images with less than 20% text tend to perform better, and it still nudges you to use fewer words, shrink the font, and move most of your copy into the body text rather than the image. The 20% figure survived the rule that killed it, demoted from a gate into a guideline. The rest of this page is about that guideline: how much text is genuinely safe now, why a wall of words quietly costs you more even when nothing is rejected, and how to prove the right answer for your own audience instead of trusting either the old rule or a hunch.",[12,27,29],{"id":28},"what-actually-changed-in-2020-and-what-did-not","What actually changed in 2020, and what did not",[17,31,32],{},"It helps to separate two things that the old rule bundled together, because only one of them went away.",[17,34,35],{},"The first was a hard eligibility check. Before September 2020, an image that failed the 20% test could be disapproved outright or have its reach severely limited, regardless of how good the ad was. You would upload a perfectly reasonable promo graphic, watch it get flagged, and spend an afternoon shrinking a headline to squeeze under an arbitrary line. That check is what Meta removed. There is no measured text ratio behind the scenes anymore, no square-counting, no automatic penalty tied to how much of the image is words. An ad that is 60% text is as eligible to run as one that is 5% text.",[17,37,38],{},"The second thing is the ordinary machinery of the auction, and that never changed. Facebook decides which ads to show, and what to charge, partly by how people respond to them. An ad that stops the scroll and earns taps is cheap to deliver; an ad that people swipe past is expensive, because the system has to show it to more people to get the same result. Text density feeds into that mechanism not as a rule but as a symptom: cluttered, hard-to-read images tend to earn weaker engagement, and weaker engagement is what actually raises your cost. Meta did not add a text penalty back through a side door. It simply stopped pretending it needed a rule when the auction already prices bad creative on its own.",[17,40,41],{},"That distinction matters when you read advice online. A lot of pages describe a phantom \"text penalty\" as if Facebook secretly still counts your words and taxes you for them. It does not. What it counts is the click-through rate, the video hold, the comments and shares, the things a real person does when a real ad is worth their attention. A text-heavy image usually scores worse on those, so it usually costs more, but the cause is human, not algorithmic bookkeeping. Fix the engagement and the cost follows, whether the image has three words or thirty.",[17,43,44],{},"One more piece of the old workflow is worth burying properly, because people still look for it. The Text Overlay tool, the grid checker at the old ads tools address, is gone. You do not need to validate an image before you upload it, there is no \"OK \u002F Low \u002F Medium \u002F High\" verdict to chase, and no reason to keep a copy of the tool bookmarked. If you want to know whether an image carries too much text, the only test that means anything now is a live one, covered at the end of this guide.",[12,46,48],{"id":47},"why-text-heavy-images-still-cost-you","Why text-heavy images still cost you",[17,50,51,52,57],{},"The clearest way to see the cost is to follow the arithmetic, because it is not mysterious. Cost per click is roughly the CPM (the price of a thousand impressions) divided by the ",[53,54,56],"a",{"href":55},"\u002Fblog\u002Fgood-ctr-facebook-ads","click-through rate",", then scaled. When the CPM holds steady and the click-through rate falls, the cost per click rises. Text density does not touch the CPM directly, but it moves the CTR, and that is enough.",[17,59,60],{},"Work a concrete example on real numbers. Gupta Media's tracker put the blended Meta CPM at about $8.19 across 2025, so imagine two versions of the same offer running at that price of reach. The text-light version, a clean product shot with a four-word hook, earns a 2.5% click-through rate. The text-heavy version, the same product buried under a headline, three feature callouts, and a starburst price, earns 1.2%, because the eye has nowhere to land and the thumb keeps moving. At an $8.19 CPM, a thousand impressions costs $8.19 either way. The clean ad turns that into 25 clicks, about $0.33 a click. The busy ad turns it into 12 clicks, about $0.68 a click. Same offer, same spend on reach, and the text-heavy creative is paying roughly twice as much for every visitor, purely because fewer people stopped. Push that across a real budget and the difference is the campaign.",[62,63,64,83],"table",{},[65,66,67],"thead",{},[68,69,70,74,77,80],"tr",{},[71,72,73],"th",{},"Version at $8.19 CPM",[71,75,76],{},"Click-through rate",[71,78,79],{},"Clicks per 1,000 impressions",[71,81,82],{},"Effective cost per click",[84,85,86,101],"tbody",{},[68,87,88,92,95,98],{},[89,90,91],"td",{},"Text-light (clean visual, short hook)",[89,93,94],{},"2.5%",[89,96,97],{},"25",[89,99,100],{},"about $0.33",[68,102,103,106,109,112],{},[89,104,105],{},"Text-heavy (headline plus callouts baked in)",[89,107,108],{},"1.2%",[89,110,111],{},"12",[89,113,114],{},"about $0.68",[17,116,117],{},"The click-through rates above are illustrative, chosen to show the mechanic, not benchmarks to hold yourself to. The point is the shape of it: nothing in that table is a penalty Meta imposed. It is just what happens when creative earns fewer taps. And the pressure runs one way over time. Meta reported that its average price per ad rose about 9% across full-year 2025, so the price of reach drifts upward every year regardless of what you do. When the baseline keeps climbing, the engagement you win back with cleaner creative is the lever you actually control. Out-bidding a busy auction is expensive and temporary; making an ad worth stopping for is the durable fix.",[17,119,120],{},"There is a delivery angle on top of the cost angle. Because the system spends against ads that perform, a text-heavy creative that underperforms does not just cost more per click, it tends to get less of your budget over time as Meta shifts spend toward whatever is working in the same campaign. So the practical effect people describe as \"deprioritized delivery\" is real, but it is an outcome of weak engagement, not a text-ratio rule reinstated. Either way, the instruction to the advertiser is identical: earn the attention and the delivery and the cost both improve.",[12,122,124],{"id":123},"how-much-on-image-text-is-actually-safe-now","How much on-image text is actually safe now",[17,126,127],{},"Since there is no limit to obey, the useful question is not \"how much am I allowed\" but \"how much helps.\" The honest answer scales with the job the image is doing. A few words that name the hook almost always help. A paragraph almost never does, because nobody reads a paragraph at scroll speed, and the words that matter get lost in the ones that do not.",[17,129,130],{},"A rough way to think about it, replacing the dead grid tool with judgement:",[62,132,133,149],{},[65,134,135],{},[68,136,137,140,143,146],{},[71,138,139],{},"Text approach",[71,141,142],{},"What it looks like",[71,144,145],{},"Cost and delivery risk",[71,147,148],{},"Best for",[84,150,151,165,179,193],{},[68,152,153,156,159,162],{},[89,154,155],{},"None",[89,157,158],{},"Pure visual, all copy in the text fields",[89,160,161],{},"Lowest, but the image must carry the idea alone",[89,163,164],{},"Strong product shots, lifestyle, brand imagery",[68,166,167,170,173,176],{},[89,168,169],{},"Light (a few words)",[89,171,172],{},"A short hook, a price, or one benefit",[89,174,175],{},"Low; usually the sweet spot",[89,177,178],{},"Most feed ads, offers, single-benefit hooks",[68,180,181,184,187,190],{},[89,182,183],{},"Moderate (a short headline plus one callout)",[89,185,186],{},"A headline and a supporting line or badge",[89,188,189],{},"Medium; fine if legible and hierarchical",[89,191,192],{},"Sales, launches, comparison hooks",[68,194,195,198,201,204],{},[89,196,197],{},"Heavy (multiple lines, several callouts)",[89,199,200],{},"A mini-poster of stacked claims",[89,202,203],{},"Higher; engagement usually drops",[89,205,206],{},"Rarely worth it in-feed; occasionally infographics",[17,208,209],{},"The old 20% figure still earns its keep as a gut check inside that spread. Not because Facebook measures it, but because a fifth of the image is about as much as one glance can absorb before the thumb wins. If your text is creeping past that, the question to ask is not \"will this pass\" but \"which of these words actually need to be on the picture.\" Usually the answer is one line, and the rest belongs in the copy fields.",[17,211,212],{},"Two exceptions are worth naming so the rule does not become dogma. Text that is part of the scene, a sign in a photograph, a label on a product, a logo on a storefront, is not \"text overlay\" and does not count against readability the way a stamped headline does; the eye reads it as part of the world, not as a claim. And a genuine infographic or a comparison chart, where the information is the creative, can carry more text on purpose. The guideline is about legibility and engagement, not a vendetta against letters. If the words help a real person understand the offer faster, keep them. If they are there because the designer had space to fill, cut them.",[12,214,216],{"id":215},"mobile-legibility-designing-text-people-can-actually-read","Mobile legibility: designing text people can actually read",[17,218,219],{},"The reason text discipline matters so much is not the rule and not even the auction. It is the screen. Facebook is a phone product: 98.5% of its users reach it on a mobile device, and 81.8% use nothing but a phone, per Statista's device data. Every headline you bake into an image is going to be read, if it is read at all, on a rectangle a few inches tall, at arm's length, in a moving feed, often in bright light or with the sound off. Design for that reality or the text is decoration, not communication.",[17,221,222],{},"A handful of rules cover almost all of it. Keep the message to a few words, because a phone screen and a scrolling thumb give you roughly a second. Set the type large and heavy; if you have to zoom to read it on your own laptop, it is invisible on a handset. Hold a strong contrast between the text and whatever sits behind it. The web accessibility standard, WCAG, sets 4.5:1 as the minimum contrast ratio between normal text and its background, and while that is an accessibility floor rather than an ad rule, it is an excellent target for ad copy, because the same low-vision and bright-sunlight conditions it protects against are exactly the conditions your feed ad lives in. Light text on a busy photo, or a thin grey caption on white, fails that floor and disappears on a phone.",[17,224,225],{},"Placement matters as much as size. Meta's own guidance for full-screen formats is to keep important text and logos out of the top and bottom of Stories and Reels, because the interface (the profile name up top, the call-to-action and progress bar down below) overlays those zones and will cover anything you put there. Treat the center of the frame as the only safe place for words in vertical creative. In the feed, keep text clear of the very edges too, since cropping across placements can clip a headline that sat too close to a corner.",[17,227,228],{},"Then there is sound. A large share of the feed scrolls on mute, which is why captions and on-screen text do so much work in video ads: the words on the screen often are the message, spoken audio or not. That cuts both ways for this topic. It is an argument for having some legible text in a video (a caption, a key line) and an argument against a static image that is already a wall of words, because the reader has no patience for two jobs at once. Give the eye one thing to read and make it readable.",[12,230,232],{"id":231},"where-the-text-belongs-image-versus-the-copy-fields","Where the text belongs: image versus the copy fields",[17,234,235,236,240],{},"The single most useful reframe on this whole subject is to stop asking how much text goes on the image and start asking which text goes where. When you ",[53,237,239],{"href":238},"\u002Fblog\u002Fhow-to-run-a-facebook-ad","set up a Facebook ad",", it has more than one place for words, and each is built for a different kind of copy. Putting the right words in the right slot solves the density problem almost by itself.",[62,242,243,256],{},[65,244,245],{},[68,246,247,250,253],{},[71,248,249],{},"Where the words go",[71,251,252],{},"What it is good for",[71,254,255],{},"Watch-outs",[84,257,258,269,280,291],{},[68,259,260,263,266],{},[89,261,262],{},"On the image",[89,264,265],{},"One hook, a price, or a single benefit the eye grabs in a glance",[89,267,268],{},"Keep it short and high-contrast; more than a line usually hurts",[68,270,271,274,277],{},[89,272,273],{},"Primary text (above the image)",[89,275,276],{},"The story, the detail, the proof, the offer terms",[89,278,279],{},"The reader taps \"more\" to expand it, so front-load the hook",[68,281,282,285,288],{},[89,283,284],{},"Headline (below the image)",[89,286,287],{},"The single clearest statement of the offer or value",[89,289,290],{},"Short and specific; it renders next to the button",[68,292,293,296,299],{},[89,294,295],{},"Call-to-action button",[89,297,298],{},"The action itself, chosen from Meta's preset labels",[89,300,301],{},"Match it to the destination so the promise is kept",[17,303,304],{},"The advantage of the copy fields over baked-in image text is not only clutter. Text in the primary text and headline is real text: it is selectable, it can be read by screen readers, it does not fight the visual for space, and you can rewrite it without re-exporting a graphic. Text stamped into an image is a picture of words, invisible to those systems and frozen until you open a design tool again. Meta's advice to keep most of your words in the body text rather than the image is really this point in one sentence. Let the image do the one thing an image does best, stop the scroll, and let the fields carry the argument.",[17,306,307],{},"This also quietly kills the temptation to overload the picture. Once you have somewhere proper to put the features, the guarantee, and the fine print, the image no longer has to be a poster. It gets to be a hook. That is the version that reads on a phone and wins the auction.",[12,309,311],{"id":310},"how-to-test-text-light-against-text-heavy","How to test text-light against text-heavy",[17,313,314],{},"You do not have to take any of this on faith, and you should not, because your audience and your offer are their own experiment. The clean way to settle it is a concept test, the same discipline that decides any creative question: build the two honest versions, run them in the same conditions, and read the result on the metrics that cost you money.",[17,316,317],{},"Make two ads for one offer. The first is text-light: a strong visual, the product or the person, with a short overlay of a few words at most. The second is text-heavy: the same offer, same product, but with a fuller headline, a couple of feature callouts, and a price baked into the image, the kind of design the old 20% rule would have failed. Keep everything else identical, the audience, the placements, the primary text, the headline, the button, so the only difference the test can attribute is the text treatment on the image itself. Put both in the same ad set so they compete in the same auction against the same people, and give each enough budget and enough days to gather real data rather than a morning of noise.",[17,319,320],{},"Then read the right numbers. Click-through rate tells you which creative stopped more thumbs. Cost per result (the purchase or lead, not a like or a reach figure) tells you which one actually earned its keep. If the text-light version wins on both, you have your answer for this offer, and you have it as evidence, not opinion. If the text-heavy version wins, believe it: some offers, some infographic-style creatives, and some audiences genuinely reward more information, and the whole point of testing is that the auction, not a retired rule, gets the final word. Most accounts find the lighter version cheaper, which is exactly what Meta's \"performs better\" guidance predicts, but the value of running the test is that you stop guessing.",[17,322,323,324,328],{},"The reason this beats the old grid tool is that the tool only ever answered a question that no longer matters (will this pass) and never answered the one that always did (will this work). A ",[53,325,327],{"href":326},"\u002Fblog\u002Ffacebook-ad-ab-testing","split test"," answers the real one, and it keeps answering as your audience and your offers change. Running that research-generate-test-read loop quickly is the difference between an account that improves and one that just spends, and keeping the whole cycle in one place, as a platform like AdPlay.ai does, makes it faster, but the discipline is what matters, whatever tools you use to run it.",[12,330,332],{"id":331},"a-short-checklist-before-you-publish-the-image","A short checklist before you publish the image",[17,334,335],{},"Put the myth down for good and the working rules take its place. There is no 20% rule, no grid tool, and no rejection for text density; an ad can be as wordy as you like and still run. What has not changed is that clean creative is cheaper, because the auction prices engagement, and text-heavy images tend to earn less of it, which raises your cost per result even when nothing is flagged. And the screen is unforgiving: nearly everyone is on a phone, much of the feed is on mute, and a headline that is illegible or buried does no work at all.",[17,337,338],{},"So before an image goes live, run it past a few plain questions. Can a stranger read the words in one glance on a phone, at a strong contrast? Is there one idea on the image, with the detail moved into the primary text and headline where it belongs? Is the text clear of the edges and, for Stories and Reels, out of the top and bottom zones the interface covers? And do you actually know this treatment works, or are you assuming, when a split test against a text-light version would tell you? Answer those honestly and the amount of text on the image stops being a rule to fear and becomes a lever you control: less to read, easier to see, cheaper to deliver, and proven by your own numbers rather than a memory of a rule Facebook itself deleted years ago.",{"title":340,"searchDepth":341,"depth":341,"links":342},"",2,[343,344,345,346,347,348,349,350],{"id":14,"depth":341,"text":15},{"id":28,"depth":341,"text":29},{"id":47,"depth":341,"text":48},{"id":123,"depth":341,"text":124},{"id":215,"depth":341,"text":216},{"id":231,"depth":341,"text":232},{"id":310,"depth":341,"text":311},{"id":331,"depth":341,"text":332},null,"neutral","Facebook's 20% text rule was retired in 2020. Here is how much text on an ad image is safe now, why heavy text still costs more, and where copy belongs.",[],"md",[357,360,363,366,369,372,375,378],{"question":358,"answer":359},"Is the Facebook 20% text rule still a thing in 2026?","No. Meta retired the 20% text rule in September 2020 and told advertisers it would no longer penalize ads with higher amounts of image text in auctions and delivery. There is no longer a hard cap, no automatic rejection for text density, and no grid tool that scores your image. What remains is a preference, not a rule: Meta's Business Help Center still states that images with less than 20% text tend to perform better, so treat 20% as a design guideline rather than a gate.",{"question":361,"answer":362},"How much text can I put on a Facebook ad image now?","As much as you want, technically, because there is no enforced limit anymore. Practically, keep it minimal. A short headline of a few words, a price, or a single benefit reads well and stays legible on a phone. Once the image becomes a paragraph, engagement usually drops and your cost per result climbs, because the auction rewards ads people stop for. The old 20% figure is still a useful sanity check: if text covers more than a fifth of the image, ask whether the words belong in the primary text field instead.",{"question":364,"answer":365},"Will Facebook reject my ad for having too much text?","Not for the amount of text alone. Since 2020, text density is no longer a rejection reason. Your ad can still be rejected for what the text says (prohibited claims, restricted categories such as health or finance, personal-attribute language, or misleading offers) under Meta's Advertising Standards, but never simply for covering the image in words. If a text-heavy ad is disapproved, read the specific policy in the notice; it will point at the message, not the pixel count.",{"question":367,"answer":368},"Does text-heavy creative cost more on Facebook?","Usually, yes, but indirectly. There is no hidden text tax added to your bid. The cost comes through engagement: a busy, text-heavy image is easier to scroll past, so it earns a lower click-through rate, and because cost per click is roughly CPM divided by click-through rate, a weaker CTR pushes your effective cost up even when the CPM is unchanged. Meta's own guidance that text-light images perform better is really a statement about this engagement effect.",{"question":370,"answer":371},"Where should the text go: on the image or in the primary text and headline?","Split it. The image should carry one idea a scrolling thumb can read in a second: a hook, a price, or a single benefit. The detail (the features, the proof, the terms of the offer) belongs in the primary text above the image and the headline below it, which are selectable, translatable, and do not clutter the visual. Meta explicitly recommends putting most of your words in the body text rather than baking them into the image.",{"question":373,"answer":374},"What are the mobile readability rules for text on a Facebook ad image?","Design for a thumb-sized screen held at arm's length. Use a large, heavy font, keep the message to a few words, and hold a strong contrast between text and background. The WCAG accessibility standard sets 4.5:1 as the minimum contrast for normal text, which is a good floor for ad copy too. Keep text away from the edges, especially the top and bottom of Stories and Reels where Meta's interface overlays sit, and remember most video is watched on mute, so the on-screen words often carry the whole message.",{"question":376,"answer":377},"Where is the Facebook text overlay tool or grid checker now?","It is gone. The Text Overlay tool, which divided your image into a 5-by-5 grid and counted how many squares contained text, was retired alongside the rule in 2020, and its page now redirects. You no longer need to check an image before uploading. If you want to gauge whether an image is too text-heavy, the honest test is not a tool but a split test: run a text-light version against a text-heavy one and compare cost per result.",{"question":379,"answer":380},"How do I test text-light versus text-heavy creative?","Build the same offer two ways: one image with a clean visual and a short overlay, one with a fuller headline or callouts baked in. Put both in the same ad set so they compete in the same auction, give each enough budget and time to gather results, then compare click-through rate and cost per result, not likes or reach. Change only the text treatment between them so the difference you read is real. Most accounts find a text-light version wins on cost, but let your own numbers, not a rule of thumb, make the call.","\u002Fimages\u002Fblog\u002Ffacebook-ad-image-text-overlay-hero.webp","You baked a bold headline into the image, and now you are second-guessing it: will Facebook throttle the reach, or was that rule scrapped years ago? It was scrapped. What replaced it is quieter and, for your budget, more important. This guide settles the 20% myth, then shows how much text is actually safe, how to keep it readable on a phone, and how to test text-light against text-heavy so the numbers decide, not a rumour.",{},true,"\u002Fblog\u002Ffacebook-ad-image-text-overlay","2027-06-20",{"title":5,"description":353},[389,393,396,400,404,408,412,415],{"label":390,"url":391,"year":392},"Search Engine Journal, Facebook Removes the 20% Text Limit on Ad Images","https:\u002F\u002Fwww.searchenginejournal.com\u002Ffacebook-removes-the-20-text-limit-on-ad-images\u002F381844\u002F","2020",{"label":394,"url":395,"year":392},"Social Media Today, Facebook's Removing Its Restrictions on Text Content in Ad Images","https:\u002F\u002Fwww.socialmediatoday.com\u002Fnews\u002Ffacebooks-removing-its-restrictions-on-text-content-in-facebook-ad-images\u002F585705\u002F",{"label":397,"url":398,"year":399},"Meta Business Help Center, Creative Best Practices for Text in Ads","https:\u002F\u002Fwww.facebook.com\u002Fbusiness\u002Fhelp\u002F223409425500940","2026",{"label":401,"url":402,"year":403},"W3C WAI, Understanding SC 1.4.3 Contrast (Minimum)","https:\u002F\u002Fwww.w3.org\u002FWAI\u002FWCAG21\u002FUnderstanding\u002Fcontrast-minimum.html","2023",{"label":405,"url":406,"year":407},"Gupta Media, The True Cost of Social Media Ads (CPM Tracker)","https:\u002F\u002Fwww.guptamedia.com\u002Fsocial-media-ads-cost","2025",{"label":409,"url":410,"year":411},"Statista, Distribution of Facebook Users by Device","https:\u002F\u002Fwww.statista.com\u002Fstatistics\u002F377808\u002Fdistribution-of-facebook-users-by-device\u002F","2022",{"label":413,"url":414,"year":407},"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":416,"url":417,"year":407},"DataReportal, Essential Facebook Statistics and Trends","https:\u002F\u002Fdatareportal.com\u002Fessential-facebook-stats",[419,423,427,431,435,439,442,446],{"label":420,"value":421,"source":422},"When Meta retired the 20% image-text rule and Text Overlay tool","September 2020","Search Engine Journal, 2020",{"label":424,"value":425,"source":426},"Meta's current guidance on how much text an ad image should carry","under 20%","Meta Business Help Center, 2026",{"label":428,"value":429,"source":430},"Minimum text-to-background contrast for legibility (WCAG AA, normal text)","4.5:1","W3C WAI, 2023",{"label":432,"value":433,"source":434},"Blended Meta (Facebook and Instagram) CPM, full year","$8.19","Gupta Media, 2025",{"label":436,"value":437,"source":438},"Facebook users who access the platform via mobile phone","98.5%","Statista, 2022",{"label":440,"value":441,"source":438},"Facebook users who access via mobile only (no desktop)","81.8%",{"label":443,"value":444,"source":445},"Meta average price per ad, full-year 2025","+9%","Meta, 2025",{"label":447,"value":448,"source":449},"People Facebook ads could reach, January 2025","2.28 billion","DataReportal, 2025","blog\u002Ffacebook-ad-image-text-overlay","Oi-0I2CH1hqSfgzQzKfcnHUlwoVgiWmffXJCs48_qSE",1786093700608]