Seven Ways to Find Better Creative References for Facebook and Instagram Ads
If you’re tired of guessing what will work in your Facebook and Instagram ads, you need a better way to spot what’s actually effective. That starts with understanding what strong creative looks like, then pulling real‑world examples you can reverse‑engineer. From Meta’s own tools to simple swipe files, you can build a reliable system for finding and organizing high‑performing ideas so you’re never starting from a blank screen again.
Know What Great Facebook and Instagram Ad Examples Look Like
Most effective Facebook and Instagram ads share a key characteristic: they communicate their main message within the first three seconds. The offer, event, or primary benefit appears immediately, and any on-screen text is placed in safe, readable areas avoiding roughly the top 14% and bottom 35% of the frame, where interface elements may obscure content.
Ads are typically formatted for 9:16 Reels and Stories, with concise copy centered on screen (around 125 characters or fewer for primary text, 40 characters or fewer for headlines, and 25 characters or fewer for descriptions). Branding remains consistent across visuals and text.
The creative usually features an actual product or service, and often includes a real person to provide context or demonstrate use.
Simple, low-production video is also a viable approach. Meta’s internal data indicates that such formats can improve click-through rates and ad recall, suggesting that high production value isn't always necessary for strong performance.
Use Meta’s Ad Library to Find Real Ad Examples
Open an advertiser’s results and sort by Started running. Ads that have been running for a longer period are more likely to be performing adequately, as ineffective ads are typically paused or replaced.
Click View more details to examine the headline, description, button text, and placements (for example, Reels or Feed).
Capture screenshots of the creatives and copy for reference in a swipe file, or use a tool such as the Foreplay Chrome extension to save and organize ads before they're removed or stop running.
The limitation of the native library is that it only surfaces what is currently active, so a paused campaign takes your reference with it. Third-party Meta ads library tools get around this by maintaining their own index GetHookd holds over 23 million ads with filters for placement, format, market and language, and anything you save stays in your collection after the campaign ends. That permanence matters more than it sounds the first time you try to compare a hook against one you noticed six months earlier.
Use Meta’s Best Small‑Business Ads as Blueprints
Instead of relying on assumptions, you can use Meta’s top-performing small‑business ads as reference models for your own campaigns. In the Ad Library, search for small‑business competitors by brand name or relevant product keywords, then sort by “started running” to identify ads that have been active for a long period, which can indicate satisfactory performance for the advertiser.
Open “View more details” to examine the headlines, descriptions, and call‑to‑action buttons. Note whether they emphasize benefits, customer reviews, or discounts, and consider how these approaches align with your own offer.
Focus on replicating the underlying structure such as a testimonial quote, a concise value statement, a clear product image, and a visible logo rather than copying exact layouts or wording.
You can save examples into organized Foreplay boards (or a similar system) to compare patterns across multiple ads. From there, select a few strong examples and turn them into A/B tests, changing only one variable at a time such as the headline, image, or call‑to‑action to isolate which elements contribute most to performance.
Spot Proven Winners by Ad Duration, Hooks, and CTAs
Once you have bookmarked a few promising examples, you can distinguish consistently effective ads from short-term tests by examining both their duration and structure. In the Meta Ad Library, use “View more details” to review the primary text, headline, description, and button copy, and note which ads have been running for the longest period.
Longer run times often indicate that an ad is meeting performance benchmarks well enough to justify continued spend.
Assess how the ad captures attention in the first three seconds whether through an offer, an event, or a new product and note where any on-screen text appears, giving preference to layouts that keep key elements in center-safe areas. Then identify recurring call-to-action (CTA) buttons and phrases, and check whether the same hook-and-CTA combinations are used consistently across different placements and over time.
This repetition can signal a refined, proven message rather than a one-off experiment.
Build a Simple Swipe File to Organize Ad Examples
With a few strong examples identified, it's useful to store them in a simple, organized swipe file.
You can use the Meta Ad Library to search competitors by brand name or product-related keywords, then save ads with long “started running” timestamps into a dedicated folder, as these are more likely to be proven performers.
Installing the Foreplay Chrome extension allows you to download and sort ads into boards, including ads that are no longer running.
For each example, record the primary text style, headline structure, call-to-action (CTA) wording, and what the creative highlights such as the product, logo, testimonials, or a specific offer.
Add brief labels and tag each ad by placement (e.g., Feed, Reels, Stories) so you can reference and brief new ads more efficiently.
Separate Visual Creative From Copy to See What Works
Although effective ads integrate visuals and copy, you can generate clearer performance insights by separating these elements during analysis. In the Ad Library, select “View more details” and categorize each ad into two components:
- Visual layout (e.g., framing, logo placement, product emphasis)
- Copy elements (primary text, headline, description)
When drafting copy for testing, stay within Meta’s recommended character ranges so that visual comparisons remain consistent: approximately 125 characters for primary text, 40 for headlines, and 25 for descriptions.
Check safe zones to ensure key elements such as faces, products, and on-image text don't appear in the top 14% or bottom 35% of the creative, where they may be obscured or cropped.
In Ads Manager, structure A/B tests so that you change only one variable at a time either the visual or the copy.
This approach makes it easier to attribute performance differences to a specific element and identify which visual treatments are most effective.
Compare Facebook vs Instagram Ad Placements and Formats
Treat Facebook and Instagram as related but distinct environments. They share the same targeting and buying tools, but their placements and formats differ, which affects how ads appear and how users engage with them.
Instagram is primarily oriented toward vertical content. Formats such as Reels and Stories, as well as many partner apps and sites, are optimized for a 9:16 aspect ratio.
Facebook offers more variation in formats, including 1:1 in Feed, in-stream, and Carousel placements, and 1.91:1 in right column or search results.
A practical approach is to capture video in a vertical format first and then use Ads Manager’s cropping tools to adjust for other placements, such as feeds.
Advantage+ creative can resize images automatically, but important visual elements and text should be kept central and limited to avoid truncation and readability issues.
Combining both video and image assets within the same campaign can help maintain consistent messaging while adapting to the strengths of each placement.
Turn Top Organic Posts and Phone‑Shot Videos Into Ad Concepts
Begin by reviewing existing content that has already performed well, rather than creating entirely new concepts. Analyze your top organic posts and phone‑shot Reels to identify patterns in effective hooks (for example, recurring behind‑the‑scenes timelapses) that can be adapted into ads while keeping the core message consistent.
Record in a 9:16 vertical format and position key text and visuals away from the top 14% and bottom 35% of the frame to maintain visibility across placements.
Convert high‑performing captions into ad copy that adheres to Meta’s recommended limits: primary text up to 125 characters, headline up to 40 characters, and description up to 25 characters.
Present the main message within the first three seconds, include identifiable human faces where relevant, and pair the video with a visually consistent image to support recognition and performance testing.
Recreate and A/B Test Your Best Facebook and Instagram Ad Ideas
Recreating and systematically A/B testing proven Facebook and Instagram ad concepts both your own and those of competitors is an effective way to identify new high-performing variations.
In Meta’s Ad Library, you can sort a brand’s ads by the oldest “started running” date to find creatives that have likely performed consistently over time.
Select “view more details” and document the key elements: hook, headline, description, and button text.
Use these as inputs to create your own variant rather than copying them directly.
Rebuild the underlying structure (for example: testimonial, review quote, and a concise benefit such as “upgrade my shoe game”) and adjust one variable at a time in Ads Manager.
Begin with a vertical 9:16 video and keep essential content out of the top 14% and bottom 35% of the frame to avoid interface overlaps.
Then add a corresponding image asset and enable Standard enhancements to allow the system to optimize delivery and formatting across placements.
Conclusion
You’ve got everything you need to find and use better creative references for Facebook and Instagram ads. Start by spotting what great ads do, then mine Meta’s tools, swipe proven structures, and separate visuals from copy. Organize your best examples, adapt them to each placement, and turn strong organic posts into testable ad concepts. Don’t chase “perfect” ideas recreate, A/B test, and keep iterating until the data tells you what truly works.