How to Tell If Competitor Ads Are Actually Scaling (Not Just Running)

An ad that has been live for 30 days and an ad that is being handed more budget today are different things. Here is where the difference actually shows up.

Why "days running" gets it wrong

Most operators check competitor ads one way. They sort for a high number of running days and treat 30+ days as proof of a winner.

That is a reasonable start and an incomplete one. Days running tell you an ad survived. They do not tell you whether it is being scaled now, because the two are not the same measurement: one is age, the other is movement. The same reading works one level up, on the store rather than the individual ad, and the signal there is the same shape.

The clearest way to see that is to plot both at once. Brandsearch charts every ad in a brand's account with its rank on one axis and how long it ran on the other, then colours each point by what its rank has been doing.

Read the horizontal axis on Loop Earplugs' account. There are ads climbing near the top of the rank scale at 30 days, at 90 days, at 150 days and out past 300 days. Age does not sort this chart. If days running were the signal, the points would form a slope, and they do not.

The legend is the other half of the answer. Every point here is blue: 20 of Loop's measured ads are labelled Scaling, none are Fading, and 10 sit in Not measured. This is what a healthy account looks like from the outside, and no single row of it could have told you that.

See what the same chart looks like for any brand you compete with

The signal is movement, and it has a definition

"Scaling" is not a vibe here. The app states the rule it used, and the rule is a rank change.

The header reads Scaling 20, gained 2+ rank positions. Every row then shows the move rather than the position: #12 to #3, #25 to #4, #13 to #11, #20 to #12.

Look at the first row against the fourth. The ad that climbed furthest, #25 to #4, has been running 141 days. The ad at #12 to #3 has been live 305 days. The shortest run in the four is 119 days. None of these are new ads, and all four are gaining. An operator filtering for "new ads only" to find what is being scaled would have missed the entire list.

One more thing worth noting: the threshold is written on the panel. Loop's says "gained 2+ rank positions". A brand with a busier account will show a different bar, so read the definition on the brand you are looking at rather than carrying Loop's number to it.

Duplicates are the second signal, and they are counted for you

When a brand finds a creative that works, it rarely just raises the budget on one ad. It duplicates the winner and runs the copies.

That count sits on the right of each row. The #25 to #4 ad carries 13 dup. The others carry 3, 3 and 4. So the ad that climbed 21 rank positions is also the one the brand has copied the most times, and you can see both facts in the same row without opening anything.

That pairing is the thing to look for. A rank climb on its own can be a competitor's rivals dropping out. A rank climb with a growing duplicate cluster is a brand deliberately pushing a creative it has decided is the winner.

Check the duplicate clusters in your own category

What the rank badge does and does not mean

Two things get misread constantly, so they are worth stating plainly.

The rank is worldwide. An ad's rank is computed across everything the brand runs in every market. When Loop's ad moves from #12 to #3, it moved against the brand's whole account, not against its European slice.

Disclosed spend is not. The euro figures in the product are the portion EU and UK rules force advertisers to publish. So an ad can rank first globally and still show a modest euro number, because the rest of its budget is never disclosed. Read a small euro figure as a disclosure boundary, not as weak performance.

This is also why rank movement is worth more than a spend snapshot when you are judging whether something is being scaled. Spend shows you one region's disclosed slice. Rank shows you the whole account.

Not every ad gets a verdict

The chart has a fourth bucket and it is the honest one. Loop's reads Not measured 10, and the app explains itself: ads seen ranked only once, or holding the same position, have nothing to report yet.

That matters when you build a habit around this. An ad with no verdict is not a failing ad. It is an ad without enough rank history to have a direction, and treating the absence of a signal as a negative signal is how people talk themselves out of watching a competitor who has simply just started.

Turn it into a weekly check

The workflow is short enough to actually repeat.

1. Open the brand you compete with and go to its Ad Rank tab.
2. Read the legend first. The split between Scaling, Evergreen, Fading and Not measured tells you the direction of the whole account before you look at a single ad.
3. Open the Scaling bucket and read the moves, not the positions. Note the definition on the panel, because it varies by brand.
4. Check the duplicate count on the biggest climbers. Movement plus duplication is the combination worth reacting to.

What you do next is not to copy the ad. It is to work out which message the brand has decided to put money behind, and to pressure-test whether the same demand is available to you. The mirror image of this check, reading the bucket where ads are losing rank instead of gaining it, is how to know when to kill an ad before your ROAS drops.

Open a competitor's Ad Rank tab and read this week's movers

Method

The figures and the screenshots on this page come from one brand's Ad Rank tab in Brandsearch, captured on 13 September 2026: loopearplugs.com, which sells earplugs for sleep, focus and live events. Rank movement, days running and duplicate counts are transcribed from the panels shown, not from a separate export, so everything quoted here is visible in the images above.

Brandsearch indexes 223M ads across Meta and TikTok and 13.7M stores (9M of them Shopify). Rank history accrues daily for tracked brands, which is why a movement can be stated at all: a snapshot can be copied from anywhere, a direction cannot.

Two limits worth stating. The Scaling threshold is set per brand and is printed on the panel. And ads in the Not measured bucket have no direction yet, so they are excluded from any claim above rather than counted as flat.

FAQ

Does a long-running ad mean it is still working?
Not by itself. Loop's chart has ads gaining rank at 300+ days and the same account has 10 ads with no measurable direction at all. Longevity says an ad was not switched off, which is a weaker claim than it sounds.

Is rank the same as spend?
No. Rank is computed worldwide across the brand's account. The euro spend figures are only the slice EU and UK disclosure rules require, so the two answer different questions and a low euro figure does not contradict a high rank.

What does the duplicate count tell me?
How many near-copies of that creative the brand is running. A high duplicate count next to a rank climb is the clearest outside evidence that a brand has picked its winner and is putting weight behind it.

Can I see this for a brand nobody is tracking?
The buckets and the chart render for any brand in the index. Daily rank history is what accumulates over time, so the longer a brand has been observed the more movement there is to read.