How to Save Ads From an Ad Library (Before the Links Die)

There are three reliable ways to save ads from an ad library: screenshot the creative with its metadata visible, record the ad's details in a structured swipe file, or use a library with built-in boards that archive the creative permanently. What does not work is the thing everyone tries first — bookmarking the ad's URL. Most ad-library links are tied to active campaigns, and when the campaign stops, the link dies. This guide covers the failure mode first, then the saving methods that survive it, platform by platform.

Why ad-library links expire#

Ad libraries are transparency surfaces, not archives. Each platform keeps ads visible on roughly this basis:

• Meta Ad Library: shows ads that are currently active. When a campaign is paused or ends, ordinary commercial ads drop out of the library — only political and social-issue ads are retained long-term (Meta keeps those for years). That shared library link from last month's winning ad? Once the advertiser rotates creatives, it 404s.
• Google Ads Transparency Center: shows ads from verified advertisers with limited retention — longer for ads shown in the EU (a Digital Services Act requirement) and for political ads, shorter for everything else. Check Google's current documentation for exact windows; the practical rule is the same: don't trust the link to be there next quarter.
• TikTok: the Commercial Content Library covers ads shown in the EEA with DSA-driven retention, while Creative Center "Top Ads" is a curated, rotating showcase that advertisers can remove themselves from at any time.
• Native networks (Taboola, Outbrain, MGID…): no official libraries exist at all, so there is no link to save in the first place — only independent captures.

The consequence: a swipe file built out of URLs rots silently. You open it in three months and half the references are gone, usually including the winners — because winning ads eventually get rotated too.

Platform-by-platform: what survives#

The common thread: no major library offers an export or download button, because these surfaces were built for accountability, not for research convenience. Whatever saving method you choose has to be your own.

Method 1: screenshots, done properly#

Screenshots work on every platform and can't expire. They scale badly, but for a personal swipe file they are fine — if you capture more than the image:

• Include the advertiser name and the date context in the frame (or the filename), not just the creative. A creative without attribution and dates is inspiration; with them, it's evidence.
• Use a naming convention you can search later: brand_platform_angle_date beats Screenshot (417).png by an order of magnitude.
• For video ads, screenshot the hook frame plus the first three seconds' worth of key frames rather than screen-recording everything — you'll actually review stills; you'll never rewatch 200 recordings.
• Store by vertical and angle, not by platform. When you need "weight-loss testimonial angles" next year, a platform-based folder tree is useless.

Method 2: save the metadata, not just the image#

The pixels are the least valuable part of a saved ad. What makes a swipe file compound in value is the structured data around each creative. A minimal record worth keeping:

Two of these deserve emphasis. First, dates: an ad you can prove ran for six weeks is a fundamentally different reference than an ad that ran for two days, and run duration is the signal that separates them. Second, the angle field: tagging by hook, angle and claim — the three layers of creative anatomy — is what turns a pile of screenshots into a searchable pattern library.

Method 3: use a library with permanent boards#

For open-web native ads, the archive problem is solved at the source. OpenAdLibrary captures creatives at the moment they are observed running and keeps them — with advertiser, network, geo, first/last-seen dates and the traced landing page — regardless of whether the campaign later stops. Saving is one click: add any creative to a board, and the board holds the full archived record permanently. As of June 2026 the index holds 725,000+ native creatives across 49 networks with 1.3 million landing captures, so most native ads you'd want to save are already in it, dates and all.

That inverts the usual workflow: instead of racing to screenshot ads before they vanish, you search the archive after the fact — including for ads that already stopped running, which are often the most instructive ones. Boards are shareable, which is what makes this work for teams: strategist curates, studio consumes. Try it from the native ad spy tool; boards are available on the free tier, with limits detailed on the pricing page.

Organizing a swipe file people actually use#

However you capture, organization decides whether the file gets opened twice. What works, distilled from how buying teams actually operate:

1. Tag by angle and hook, not by "cool". "Us-vs-them comparison", "fear-of-ingredient", "local-pricing curiosity" are retrievable categories. "Great ad 👍" is not. The library of proven ecommerce angles is a reasonable starter taxonomy.
2. Separate proven from interesting. Two folders minimum: ads with evidence (long run times, many placements) and ads that merely caught your eye. Only the first folder should inform briefs.
3. Review on a schedule. A monthly 30-minute pass — what's new, what died, what patterns repeat — is what converts storage into learning. Scoring saved creatives systematically gives that pass a structure.
4. Prune. Creative styles fatigue at the market level, not just the campaign level; a hook that saturated last year is a warning, not a template. When a pattern stops appearing in live captures, note it — that's creative fatigue operating at vertical scale.

For teams, put the capture step on a schedule rather than trusting opportunistic saving: a 15-minute weekly pass over each watched competitor — new creatives into the board, dead creatives annotated with their final run length — produces a longitudinal record that no amount of retroactive screenshotting can reconstruct. It is the difference between owning a photo album and owning a time series.

Copyright and usage ground rules#

Saving ads for research, analysis and internal reference is standard competitive-intelligence practice — the ads were served publicly, and archiving public advertising has decades of precedent. The line sits at reuse: re-running someone's creative, cloning their landing page, or lifting their copy wholesale moves you from research into copyright and trademark infringement. The working rule for briefs: learn the pattern, rebuild the execution. If the legal boundary matters to your team (it should), the legality of competitor ad research covers where observation ends and infringement begins.

The meta-lesson of all of this: ad libraries answer "what is running now?", but a swipe file answers "what has worked over time?" — and only the second question compounds. Pick a capture method that preserves dates and attribution, tag by mechanism, and your file becomes the research asset most buyers wish they had started three years ago.