What is Frequency Capping?
Also known as: Frequency cap, Impression cap
What is frequency capping?
Frequency capping is the rule that limits how often a single user or household sees the same ad inside a defined window. The cap sets a ceiling. Once the user hits it, the platform stops serving the ad until the window resets.
A cap is written as two numbers. Impressions over time. "3 per 7 days" means each user sees the ad three times per week, then stops. The window can be a day, a week, or the full flight.
Caps live at different levels. Some platforms let buyers cap at the ad, ad group, campaign, or account level. A cap at the ad level still lets a user see five different ads from the same campaign. A campaign-level cap limits total exposure across every ad in it.
Why frequency capping exists
Frequency capping exists because attention is finite and budgets are not. Without a cap, the auction will happily serve the same user the same ad ten times in a week. The first three impressions teach. The next seven annoy. Per Meta's frequency guidance, weekly frequency above 3.5 on cold audiences correlates with rising CPM and falling CTR.
Three problems show up when caps are missing.
Ad fatigue. CTR falls 30 to 60 percent. CPM rises 15 to 40 percent. The same creative that worked at frequency 1.4 stops working at frequency 4.6.
Brand backlash. "Stop showing me this ad" replies become a leading indicator of fatigue, often 48 hours before CPA breaks. Sentiment drops on the audience the campaign was meant to convert.
Wasted spend. Money that should reach new prospects gets burned on the same 5,000 people. A campaign that could touch 800,000 unique users at frequency 2 reaches 200,000 at frequency 8 for the same budget.
Setting frequency caps
Caps can be written against three different events. Each unit suits a different goal.
| Cap unit | What it counts | Best fit |
|---|---|---|
| Per impression | Times the ad rendered to the user | Awareness, reach, brand campaigns |
| Per click | Times the user clicked the ad | Performance campaigns where each click costs money |
| Per conversion | Stops serving once the user converts | Retargeting and CRM exclusion lists |
Most platforms default to per-impression caps. They are the easiest to count and the most generic. Per-click caps are common on programmatic display and CTV. Per-conversion caps are usually built outside the platform via audience exclusion lists rather than a true cap field.
The window matters as much as the number. A cap of 5 per day is far looser than a cap of 5 per week. Match the window to the buying cycle. A 24-hour DTC purchase decision wants tight daily caps. A 90-day B2B consideration cycle wants weekly or monthly windows.
Frequency capping by platform
Each platform exposes caps differently. Some give granular control. Some hide the lever inside auto-optimization.
| Platform | Cap level | Default behavior | Notes |
|---|---|---|---|
| Meta (Facebook, Instagram) | Ad set, reach objective only | Auto-optimized on most objectives | Reach campaigns expose the field. Conversion campaigns hide it. Per Meta frequency guidance |
| Google Ads (Display, Video, Demand Gen) | Campaign, ad group, ad | Manual cap field exposed | Per Google Ads frequency cap docs, Search has no cap |
| TikTok Ads | Ad group | Auto-optimized for prospecting | Manual cap in Reach and Frequency objective |
| LinkedIn Ads | Account-level | Default 1 per 48 hours per campaign | Conservative defaults. Most B2B accounts loosen them |
| Programmatic DSPs | Line item, advertiser | Fully exposed, multi-window | The Trade Desk, DV360, Amazon DSP support stacked caps |
| Connected TV (CTV) | Household-level | Set in the DSP | Per IAB Tech Lab CTV buyers guide, the unit is household, not user |
Search is the gap. Search ads do not run frequency caps because the user triggers each impression by typing a query. On every other channel, the platform decides when to serve. The cap is the buyer's lever.
Optimal frequency caps by goal
The right cap depends on the job. Awareness wants more reach per dollar. Conversion wants more impressions per converter. Retargeting wants persistence without harassment. Use these as starting bands. Verify against your own CPA curve.
| Goal | Audience | Suggested cap | Reasoning |
|---|---|---|---|
| Brand awareness | Cold, broad | 3 per 7 days | Maximize unique reach. Frequency above 3 burns budget on repeat viewers |
| Consideration | Cold prospecting | 4 to 5 per 7 days | Need repeat exposure to register the message |
| Conversion | Cold + warm blend | 5 to 7 per 7 days | More touches drive purchase intent on lower-cost items |
| Retargeting | Warm site visitors | 7 to 10 per 7 days | Smaller pool, higher purchase intent, tolerates more touches |
| Cart abandoners | Hot, high intent | 10 to 14 per 7 days | Tight 7-day window. Aggressive cap before the lead goes cold |
| CTV brand campaign | Household | 3 per household per 7 days | Big screen, sound on. Three impressions land harder than seven on social |
The pattern: tighter audience, higher cap. A cold 5-million-person prospecting pool fatigues at frequency 3. A 25,000-person retargeting pool can hold frequency 8 because the buying intent is already there.
Set the cap below the fatigue band, not at it. A cap at 3 still lets some users see the ad 4 or 5 times when the auction over-delivers. A cap at 2 keeps the real-world max under 3 for most users.
Real-world example with numbers
A DTC cookware brand spends 60,000 dollars per month on Meta. No frequency cap. The conversion objective handles delivery on its own.
Month 1: Frequency hits 5.8 by week 3. CTR drops from 1.9 to 0.9 percent. CPA climbs from 38 to 71 dollars. Reach plateaus at 340,000 unique users. Comments turn negative.
The team adds a cap of 4 per 7 days at the ad-set level. They expand the audience from 1.1 million to 4.2 million. Same creative library.
Month 2: Frequency settles at 2.6. CTR recovers to 1.7 percent. CPA drops to 41 dollars. Reach climbs to 1.1 million unique users. Same spend. Three times the unique audience.
The cap did not fix the campaign alone. The audience expansion did the heavy lift. The cap forced the auction to spread spend instead of pounding the same 50,000 people. Cap plus audience expansion delivered a 42 percent CPA improvement at constant budget.
Frequency capping in a cookieless 2026
Frequency caps used to run on third-party cookies. The cookie identified the user across sites. The DSP counted impressions per cookie. Safari and Firefox blocked that identifier years ago. Chrome's privacy sandbox finished the job in 2024. Per the IAB Tech Lab privacy sandbox guidance, buyers now run frequency control across three identifier paths.
First-party logged-in IDs. Meta, Google, TikTok, Amazon, and Netflix count frequency against their own logged-in user. Inside a walled garden, the cap is precise. Across walled gardens, it does not exist.
Universal IDs. UID 2.0, RampID, and ID5 stitch identifiers across the open programmatic web for users who consent. Coverage is partial, often 40 to 60 percent of impressions per The Trade Desk's UID 2.0 adoption reporting.
Household and IP-based caps. On CTV and OTT, the IP address anchors the household. Frequency runs per household. Less granular, more durable, since IP does not need a cookie consent prompt.
Buyers can no longer set one cap across every channel and trust the math. Each channel runs its own counter. Cross-channel frequency control now belongs in mixed media modeling and post-campaign measurement, not in a real-time cap field. The total exposure budget gets managed at the plan level. Frequency capping in 2026 is a portfolio problem.
Related terms
Frequently asked questions
What is frequency capping in simple terms?
Frequency capping is the rule that limits how many times one user sees your ad in a set window. A cap of 3 per 7 days means each user sees the ad up to three times per week. Cross that limit and the platform stops serving them. The goal is to spread spend across more unique people.
What is a good frequency cap on Meta?
For cold prospecting, 3 impressions per 7 days holds CTR steady on most accounts. Warm retargeting can run 5 to 7 per 7 days before fatigue hits. Per Meta's own delivery guidance, weekly frequency above 3.5 on cold audiences correlates with rising CPM and falling click rate.
Does Google Ads support frequency capping?
Yes. Google Ads supports frequency capping on Display, Video, and Demand Gen campaigns at the campaign, ad group, or ad level, per Google's frequency capping documentation. Search campaigns do not use frequency caps because the user triggers the impression by typing a query.
How do you set a frequency cap on connected TV?
CTV frequency caps run at the household level, not the user level, because a TV is shared. DSPs like The Trade Desk let buyers set caps such as 3 per household per 7 days across all inventory. Per IAB Tech Lab guidance, household caps are the standard unit of measurement on CTV.
What happens if you set the cap too low?
Too-low caps starve the auction. The platform cannot find enough impressions to deliver the budget, CPM rises, and the algorithm leaves spend on the table. A cap of 1 per week on a cold prospecting audience usually under-delivers by 30 to 50 percent. Caps should be set just below the fatigue band, not far below it.