Quick Answer
- There is no universal ad creative refresh frequency that works across every platform.
- In AdSpyder’s May 2026 sample of 24,650 ads, 54% were observed for less than 7 days.
- 22% lasted 30+ days, while only 9.4% remained observable for 90+ days.
- Meta had a 4-day median observed lifespan, YouTube 5.1 days and LinkedIn 43.3 days.
- Active advertisers also run multiple variants: Amazon Sponsored campaigns had a median of 5 variants, while Google Shopping had 16.
- Winning advertisers do not simply replace everything on a schedule. They test continuously, kill weak variations quickly and keep proven creatives running longer.
Ask a media buyer how often ads should be refreshed and you will hear familiar answers: every seven days, every two weeks, once frequency reaches three, or whenever CTR begins to decline.
AdSpyder’s archive suggests a more useful answer: creative refresh frequency is not one calendar rule. It is a portfolio behaviour. Advertisers test many ideas, remove most of them quickly and allow a smaller group of persistent creatives to continue running for weeks or months.
Important Research Distinction
AdSpyder can observe when competitor ads first and last appear, creative formats and variation volume. It cannot see a competitor’s private CTR, spend, CPA or ROAS. In this study, creative persistence is used as an observable market signal—not proof that every long-running ad is profitable.
Methodology and Dataset
This analysis combines two AdSpyder datasets from May 2026.
Dataset 1: Creative lifespan
AdSpyder sampled 24,650 ads with usable first-seen and last-seen date pairs across six advertising environments:
- 4,650 LinkedIn ads
- 4,000 Meta/Facebook/Instagram ads
- 5,000 YouTube ads
- 4,500 Bing/Microsoft ads
- 1,000 Google Search ads
- 500 Display ads
Dataset 2: Creative variation volume
AdSpyder separately analysed 20,000 real campaigns: 10,000 Amazon Sponsored campaigns and 10,000 Google Shopping campaigns, supplemented with advertiser-level samples from Meta and Google Search.
| Measurement | Definition |
|---|---|
| First seen | First date AdSpyder observed a creative. |
| Last seen | Most recent date AdSpyder observed the creative. |
| Observed lifespan | Difference between first-seen and last-seen dates. |
| Variant count | Number of ads or products associated with the campaign/advertiser unit available in the archive. |
Observed duration is a floor estimate. If an ad is still active today, its final lifetime will be longer than its current observed duration.
Executive Findings
54%
observed for under 7 days
22%
survived for 30+ days
9.4%
observed for 90+ days
5 / 16
median Amazon / Shopping variants
The strongest pattern is not simply “ads change fast.” It is that the creative portfolio splits into two groups:
- Test creatives: large numbers enter the market and disappear quickly.
- Persistent creatives: a smaller group survives long enough to become durable controls or evergreen assets.
This is also why AdSpyder’s creative lifespan study is more useful than a blanket recommendation to “replace ads every two weeks.”
Overall Creative Refresh Benchmark
Under 7 days — 54%
7–30 days — 24%
30+ days — 22%
90+ days — 9.4%
Source: AdSpyder platform data, May 2026. The 90+ day group is contained within the 30+ day cohort.
Creative Refresh Frequency Changes Dramatically by Platform
| Platform | Sample | Median Observed Days | 30+ Days | Practical Signal |
|---|---|---|---|---|
| 4,650 | 43.3 days | 70.7% | Long creative lifecycle | |
| YouTube | 5,000 | 5.1 days | 22.1% | Fast tests + durable winners |
| Meta | 4,000 | 4.0 days | 29.8% | Aggressive testing + long tail |
| Bing | 4,500 | 0* | 8.1% | Rapid observed turnover |
| Google Search | 1,000 | 0* | 4.9% | High message turnover |
| Display | 500 | 0* | 0.2% | Fastest observed replacement |
*A median of zero means at least half of sampled ads had matching first-seen and last-seen dates. It does not prove they literally ran for zero or one day.
Meta: Test Fast, but Do Not Automatically Kill Survivors
Meta’s four-day median might tempt teams to assume every creative should be replaced weekly. That would be the wrong conclusion.
Although 54.4% of sampled Meta ads were observed for less than seven days, 29.8% survived 30+ days and 15.5% survived 90+ days. The distribution suggests aggressive testing followed by long retention for a smaller subset.
LinkedIn: A Completely Different Creative Rhythm
LinkedIn had a 43.3-day median observed lifespan, and more than 70% of sampled ads reached 30 days. Applying a Meta-style weekly creative schedule to LinkedIn would therefore ignore how differently advertisers operate on the platform.
Search: Refresh Messaging, Not Necessarily Visual Assets
Search environments naturally produce more headline, description, keyword and offer combinations. Here, “creative refresh” is better thought of as message variation rather than replacing a design asset.
Format Changes Creative Lifespan Too
Platform averages hide another important variable: format.
| Platform / Format | Sample | Observed 30+ Days |
|---|---|---|
| LinkedIn — Image | 6,931 | 91.1% |
| Meta — Carousel | 5,954 | 39.9% |
| LinkedIn — Video | 1,956 | 39.8% |
| Meta — Video | 8,299 | 26.2% |
| Meta — Single Image | 5,271 | 23.3% |
| YouTube — Video | 5,000 | 22.1% |
On Meta, carousel creatives produced the strongest 30-day persistence signal in AdSpyder’s current feed: 39.9%, versus 26.2% for video and 23.3% for single-image ads.
LinkedIn behaves differently again: image ads reached 91.1% observed 30-day persistence in the sampled archive. The broader ad-format survival study shows why refresh planning should consider format as well as platform.
Winning Advertisers Do Not Wait for One Creative to Die Before Making the Next One
Creative lifespan tells us how long individual ads survive. Variant counts tell us how advertisers build the system around those ads.
| Platform | Measurement Unit | Median | Key Finding |
|---|---|---|---|
| Amazon Sponsored | Variants / campaign | 5 | 51% run 4–10 variants |
| Google Shopping | Products / campaign | 16 | 87% sit in 11–50 range |
| Meta | Recent ads / active page | 239 | Creative volume is ongoing, not one test batch |
| Google Search | Lifetime ads / advertiser | 3,147 | Testing volume compounds over time |
The detailed 20,000-campaign variant study found that single-variant campaigns were effectively nonexistent among the sampled active Amazon and Google Shopping campaigns.
The Real Refresh Pattern: Launch → Filter → Extend → Replace
The combined lifespan and variation data suggests a better operating model than “refresh every X days.”
Creative Refresh Frequency Is Not the Same as Ad Fatigue
This distinction matters.
| Creative Refresh | Ad Fatigue |
|---|---|
| Advertiser launches or replaces creative | Audience response deteriorates after repeated exposure |
| Observable externally | Must be confirmed with private performance data |
| Can happen because of seasonality, testing or new offers | Usually involves worsening response metrics |
For the performance-side framework, see AdSpyder’s ad fatigue benchmark study, which separates creative age from frequency, CTR, CPA and ROAS deterioration.
A Practical Creative Refresh Playbook
| Situation | Recommended Action |
|---|---|
| New creative launched | Avoid judging solely from age. Let enough account data accumulate for your objective and spend level. |
| Creative is weak early | Replace or pause based on account performance—not competitor lifespan alone. |
| Creative is 30+ days old and healthy | Keep it running while preparing new challengers. |
| Frequency rises and CTR falls | Introduce a materially different creative concept. |
| Several new creatives decline together | Investigate offer, audience, landing page and demand—not only the artwork. |
| Competitors suddenly rotate aggressively | Check for new offers, product launches, seasonal campaigns or format shifts. |
Does Industry Change Creative Refresh Frequency?
AdSpyder’s current public dataset supports strong industry-level differences in format mix, but it does not yet support a defensible industry-by-industry refresh-frequency benchmark.
For example, AdSpyder’s May 2026 Meta industry study found image-heavy mixes in categories such as retail, real estate, fashion and financial services. That matters because production effort and format choice can influence how easily teams create replacement assets.
But it would be misleading to turn those format percentages into claims such as “fashion refreshes every X days.” For industry-specific decisions, benchmark your direct competitors rather than inventing a category-wide calendar.
4 Patterns Worth Watching in Competitor Creative Timelines
AdSpyder’s winning creative research workflow recommends looking for repeated investment signals such as longevity, multiple variations, repeated hooks and cross-platform presence rather than simply choosing the newest ad.
How to Benchmark a Competitor’s Creative Refresh Frequency With AdSpyder
AdSpyder’s current Ad Library includes platform, country, format, first-seen date, estimated run duration, headline, CTA and destination URL among its structured ad fields.
Use Ad Analytics when you want to connect those individual creative changes with broader domain and campaign activity.
What Marketers Should Do Next
Build a pipeline, not a refresh deadline
If your next creative starts only after the current winner weakens, the production system is already late. Maintain a queue of new concepts while controls are healthy.
Refresh concepts before cosmetics
Changing the background color while keeping the same hook, offer and proof may not create meaningful audience novelty. Test different reasons to care, not only different ways to decorate the same message.
Keep proven controls alive
Long-running ads are not automatically stale. If your account data shows a creative continues meeting its target, let challengers compete with it rather than retiring it solely because it is old.
Benchmark direct competitors
The market-wide data tells you what is possible. Your closest competitors reveal the cadence your audience is actually exposed to.
Limitations and Methodology Notes
- Observed duration is not confirmed uninterrupted delivery. First-seen and last-seen indicate observation dates, not every day the ad served.
- Long-running does not guarantee profitable. AdSpyder cannot access competitor CTR, ROAS, CPA or conversion data.
- Removing an ad does not prove it failed. Campaigns can end because of inventory, seasonality, promotions or strategic changes.
- Platform datasets have different structures. Campaign-level variant counts are available for Amazon and Shopping, while Meta and Search require advertiser-level measures.
- Format survival is not conversion rate. Persistence shows continued advertiser use, not guaranteed business results.
- No universal refresh calendar is supported by this dataset. Platform, format, audience and campaign economics materially affect cadence.
Creative Refresh Checklist
Frequently Asked Questions
How often should ad creative be refreshed?
There is no universal schedule. AdSpyder’s May 2026 data shows median observed lifespan ranges from four days on Meta to 43.3 days on LinkedIn. Use platform context and your own campaign performance rather than a fixed calendar.
How often do Facebook and Instagram advertisers change creative?
The median observed Meta creative in AdSpyder’s sample lasted four days, but nearly 30% survived at least 30 days. That suggests frequent testing alongside a smaller group of long-running controls.
Should I replace an ad after seven days?
Not automatically. A seven-day-old creative may still be performing strongly. Use CTR, CPA, conversion rate, ROAS, frequency and campaign context before replacing it.
How many creative variants should I run?
The answer varies by platform. In AdSpyder’s 20,000-campaign study, Amazon Sponsored campaigns had a median of five variants and Google Shopping campaigns had 16. Meta’s active advertisers operate at much larger overall creative volumes.
Does a long-running competitor ad mean it is profitable?
No. Longevity is a useful persistence signal because the advertiser continued running the creative, but external tools cannot see its private profitability or ROAS.
Which Meta ad format survives longest?
In AdSpyder’s real-time Meta sample, 39.9% of carousel ads were observed for at least 30 days, compared with 26.2% of video and 23.3% of single-image ads.
How can AdSpyder help plan creative refreshes?
AdSpyder lets marketers inspect competitor ad history, first-seen dates, estimated run duration, formats, copy, CTAs and destination URLs across multiple platforms. Use those signals to benchmark competitor cadence and combine them with your own campaign data.
Stop refreshing creative because a calendar told you to.
Use AdSpyder to see when competitors launch new ads, which creatives disappear quickly and which hooks, formats and offers remain active for weeks or months.
