Non-skippable ads now run beside Shorts. The frequency ceiling moved to 7

Non-skippable in-stream ads can now be mixed with In-feed and Shorts in the same Target frequency campaign, and the weekly target range for that mix tops out at 7 instead of 4. Both changes are in Google's help page, both are unannounced, and the second one is the one that costs money if you misread it.

The reason is simple: Target frequency is not a cap. It is a distribution goal. Asking for 7 does not mean people see your ad seven times — it means Google concentrates your budget on fewer people until the average hits seven. Same money, less reach, and a higher cost per person actually reached.

That trade is worth making in some accounts and clearly wrong in others. The arithmetic below separates the two, and it takes about ten minutes with numbers you already have.

What the page said in October, and what it says now

Google made no announcement. The Target frequency help page is undated, so the only honest way to describe the change is to compare the current text against an archived capture.

Read the middle row carefully, because it is where most commentary will get this wrong. In October a campaign that included non-skippable in-stream sat inside the 2-4 weekly / 4-8 monthly bucket. There is now one range for everything, so the highest weekly target a non-skippable-inclusive campaign can request moved from 4 to 7. What did not change is the format's duration — the October page simply never stated it, so there is no length change to report, only a ceiling and a mix.

You can check it yourself: web.archive.org/web/20261014065357/https://support.google.com/google-ads/answer/12400225.

The setting is a target, not a cap — Google says so in the FAQ

The FAQ on that same page kills the most common misreading:

"If I set my weekly or monthly Target frequency as X, is that also a frequency cap? No, this is not a frequency cap. To achieve a weekly or monthly Target frequency, a subset of users may view the ad more times in a week than the target, which will help offset users who view it less than the target in a week."

So a target of 7 produces a distribution, not a wall: some people see it four times, some see it eleven. The average lands near seven. If your concern is a single person being hit eleven times by a 20-second unskippable, this setting never solved that — Cap impression frequency is a different control.

The arithmetic: fixing the target fixes your reach

Three formulas do all the work here.

• Weekly impressions capacity = (weekly budget ÷ CPM) × 1,000
• Reach at target ≈ weekly impressions ÷ target frequency
• Cost per 1,000 people reached = CPM × target frequency

That third line is the one to write on a whiteboard. Every step up the target multiplies your cost per reached person by the same factor, because you are paying the same CPM for impressions that increasingly land on people you already paid to reach.

Work it end to end with round numbers — none of these come from a client account, they are there so you can substitute your own:

• Weekly budget: $4,500
• CPM: $9.00, so 500,000 impressions a week
• Weekly reachable users in the geo and audience: 250,000

At a target of 7, that budget reaches 71,400 people instead of 250,000. To hold the same 250,000 people at 7, the weekly budget has to go to $15,750 — a 250% increase to buy the identical audience, each of them seeing the ad five more times.

That is the whole decision in one table. Moving from 4 to 7 is not an upgrade you enable. It is a purchase you make, and the price is reach.

What the mix does to the 20 seconds you paid for

Here is the part the format comparison tables skip. The target applies to the campaign, across all five formats, and the formats do not cost the same per impression. In-feed and Shorts inventory is cheap; a 20-second unskippable is not. When you put them in one campaign with one frequency target, the cheap formats absorb most of the delivery, because that is how the system hits the target efficiently.

Which means a target of 7 does not buy seven 20-second unskippables. It buys seven exposures of which your unskippable might be one, with the other six showing up as a thumbnail or a swipeable clip. If your creative only makes sense when someone watches the full 20 seconds — a demonstration, a claim that needs the setup, a story with a turn — mixing degrades exactly the property you chose the format for.

Google's page still says so out loud about the alternative: in-stream-only is described as a "legacy single ad format," and you "won't be able to add in-feed or Shorts to this campaign, which may result in less reach or a higher cost-per-thousand impressions (CPM)." That is not a warning to avoid — it is the price of message integrity, stated plainly. For a brand whose creative needs the full length, paying it is the rational choice, and no amount of frequency room changes that.

The creative count sits in the same paragraph as the frequency

One line from the best-practices block deserves more attention than the ranges:

"Try to add at least as many creatives as your Target frequency number, if not more, to avoid creative fatigue."

Read as an operating rule, a target of 7 requires seven videos. Most accounts we look at are running two or three. At three creatives and seven impressions, each video plays about 2.3 times to the same person inside a week — that is not a frequency strategy, it is the same three films on rotation, and the person who sees you seven times is the person most likely to remember they saw you seven times.

Two smaller constraints from the same page that decide whether the number is even reachable: one ad group per Target frequency campaign, and reserve buying only for non-mixed formats — so the five-format mix is bought at auction. Reach and frequency reporting also lags up to 72 hours, which means the first two days of a flight tell you nothing about achieved frequency.

When the higher target is a tax, not reach

Four conditions. If two or more are true, stay at 2-4 and spend the difference on more creative.

• The budget and the audience are both fixed. Then target and reach move in opposite directions by definition. Your only question is which one you are buying.
• You have fewer creatives than the target. The extra impressions repeat a film the person has already watched in full.
• The flight is shorter than a week. Google's own guidance is at least 7 days, with 2-4 weeks working best. A 5-day flight at 7 has to front-load, and the last two days of the curve are the ones you never see.
• The message needs the full 20 seconds. Mixing trades message completeness for cheap touches, one impression at a time.

And the reverse — when 5-7 is genuinely right:

The audit before you touch the slider

Pull the last four weeks of any Target frequency campaign and write down five numbers: weekly budget, CPM, unique users reached, achieved avg. impr. freq. per user (7 days), and the count of distinct creatives actually delivering. Then divide impressions by reached users and compare the result to the target you set.

If achieved frequency is well below target, the budget never supported the number and raising it to 7 changes nothing except your expectations. If achieved frequency is at or above target while reach is shrinking week over week, you are already paying the tax — the target is binding and the audience is recycling. If both look healthy and you have seven creatives in rotation, the new ceiling is genuinely available to you.

The same discipline applies to every target Google lets you set. We covered the same pattern with bid targets that stopped being suggestions: a number you type into Google Ads is an instruction about how to allocate, not a promise about outcome. This is that lesson in the format dropdown rather than the bid field.

Frequency decisions also sit on top of the same inventory question we covered in Meta's placement-exclusion change: who decides which surfaces your money touches, and what you give up when you stop deciding. One difference here — Google lets you choose, and prices both choices, which is more than most platforms offer. And because frequency is only as honest as the numbers behind it, the reporting discipline in what to demand a buyer prove about invalid traffic applies to the reach figure you are dividing by.

If you want to run this arithmetic on a live campaign — one account, one audience estimate, the reach you give up at each target — bring it to us and we will do it with your numbers rather than an example. Our digital ads and PPC team does this before every YouTube budget review, because the ceiling moving is not a reason to use it.