Why is my remarketing campaign not performing?

Alexandre Airvault
January 13, 2026

Start with the right diagnosis: is this a delivery problem or a conversion problem?

When someone tells me “remarketing isn’t performing,” I immediately clarify whether the campaign is not serving (few/no impressions, no spend), or it’s serving but not converting (impressions and clicks happen, but CPA/ROAS is poor). These two situations have completely different root causes—and if you treat a delivery problem like a performance problem (or vice versa), you can waste weeks “optimizing” the wrong thing.

Remarketing is also more sensitive than most campaign types to list eligibility, privacy/consent signals, and configuration details like “Targeting vs Observation.” That’s why it can look like it suddenly “broke” even when nothing changed in your ads.

  • If spend is near $0: prioritize audience eligibility, targeting mode, and consent/tagging.
  • If spend is normal but results are weak: prioritize segmentation, exclusions, conversion measurement quality, and frequency/creative strategy.

When remarketing isn’t spending: fix audience eligibility, setup, and privacy signals

1) Your audience is not eligible (or too small “where it counts”)

Remarketing can’t run if the platform can’t match enough recent, eligible users to actually serve ads. Audience size is not just “how many cookies touched your site in the last year.” Eligibility is based on active users in a recent window, and it varies by list type and where you’re trying to serve.

As a practical baseline, if your remarketing list has fewer than 100 active users in the last 30 days, expect limited or zero delivery in many remarketing surfaces. Even if your list shows “800 users,” the campaign may still struggle because not everyone is active/available at the moment ads would be served.

Also watch for list health and status issues. If a segment is set to Closed, it won’t keep growing. And if a segment hasn’t been used in targeting long enough, it can be automatically closed—meaning it will stop adding new users until you reopen it. On top of that, membership duration matters: if you’re using a very short duration (for example, 7 days) and you don’t have steady traffic, your list can “evaporate” below eligibility thresholds.

2) You built the campaign, but you’re not actually targeting the list

This is one of the most common (and most frustrating) setup mistakes: you add audiences, but you added them in Observation mode when you intended Targeting mode (or the reverse).

In Search (and commonly Shopping-style setups), new campaigns often default audience handling to Observation—meaning your ads can still show to everyone matching keywords, and the audience is just a reporting/bid-adjustment layer. If you intended a true “RLSA-only” style campaign (serve only to past visitors), you must ensure the audience setting is Targeting. Conversely, if you accidentally set Targeting in a keyword campaign with tiny lists, you can choke delivery to near zero because you’ve restricted the reach to an audience that may not meet minimum thresholds.

Bottom line: confirm whether the campaign/ad group is meant to be a remarketing-only campaign (Targeting) or a bid-layer / reporting campaign (Observation), and verify it matches your intent.

3) Your list isn’t being built because tagging and consent signals are blocking it

If your remarketing lists are staying at 0—or they were growing and then flatlined—this is usually not a bidding problem. It’s almost always one of these:

  • The site isn’t fully tagged: the base tag isn’t present across the site, or it’s firing inconsistently.
  • Consent settings disable ad personalization: if ad_personalization is denied, personalized advertising features like remarketing don’t receive data. If ad_storage is denied, advertising cookies can’t be read/written, which can drastically reduce list growth and matchability.
  • Your CMP/consent framework isn’t passing signals reliably: if the consent string isn’t available fast enough, tags can run in a restricted mode where remarketing features may be disabled for that traffic.

This is why remarketing performance can drop right after a consent banner change, a tag manager container update, a site redesign, or a move to a new domain/subdomain setup. The campaign didn’t “forget how to perform”—it simply stopped receiving eligible audience inputs at the same rate.

4) Customer Match and dynamic remarketing: “processing” and “freshness” bottlenecks

Two special cases deserve their own callout because they often look like “the campaign is dead” when the real issue is eligibility and processing.

Customer Match: if you’re using customer lists, remember that lists need sufficient active matched people to serve at the moment of auction. Too-small lists, overly narrow layering on top of the list, or insufficient bids/budget can all lead to “low or no volume.” Upload processing can also take time, and frequent re-uploads can keep lists in a near-constant processing state. From a maintenance perspective, Customer Match memberships also have a maximum retention window; if you don’t refresh/update the list, older memberships can age out and quietly reduce your usable match pool.

Dynamic remarketing (feed-based): if you rely on a product or business data source, your first processing cycle can take a few business days, and the campaign’s ability to serve can hinge on correct account linking and feed readiness. If the campaign is configured correctly but the data source is “under review,” mis-linked, or disallowed for your category settings, delivery can be extremely limited regardless of bids or creative quality.

When remarketing is spending but results are weak: raise intent, control repetition, and give bidding clean signals

1) Stop treating all visitors the same (segment by intent + recency)

Generic “All Visitors – 540 days” remarketing is easy to launch, but it’s rarely efficient. A visitor from yesterday who abandoned a checkout is not the same as someone who bounced from a blog post eight months ago. When you mix them, your messaging gets generic, your conversion rate drops, and your bidding system gets noisy signals.

I recommend structuring remarketing around a few clear intent tiers and short, meaningful membership windows that reflect your sales cycle. For example: product viewers (short window), cart/lead starters (very short window), and past converters (exclude or upsell with a different message). The goal is to make the audience’s “why” obvious, so your offer and creative can be specific instead of broad.

2) Fix conversion measurement first—then judge performance

Remarketing is often judged harshly because it’s closer to the bottom of the funnel, where tracking gaps are more visible. Before you declare “remarketing doesn’t work,” make sure you’re not undercounting conversions due to consent restrictions, missing tags, or configuration gaps.

Also account for conversion delay. If you’re optimizing or reporting too close to today, you can end up “optimizing away” campaigns that are actually working but have conversions that arrive days later. Depending on your conversion window settings, conversions can be reported long after the click, so recent date ranges often understate true performance.

If you’re using automated bidding, give the system a fair chance to stabilize. Learning and recalibration time is influenced by how many conversions you generate, how long your conversion cycle is, and which bid strategy you’re using. Big, frequent target changes (for example, repeatedly moving Target CPA/ROAS every couple of days) can create volatility that looks like poor remarketing performance, when it’s really just constant re-learning.

3) Manage ad fatigue with frequency discipline and creative rotation

Remarketing audiences are smaller by design, so fatigue happens fast. If people see the same message too many times, CTR and conversion rate drop, and you end up paying for repeated exposure that stops moving users forward.

For Display and Video campaigns, use frequency management to limit how often the same person sees your ads. This protects performance while you test messaging. Keep in mind that frequency capping isn’t available in every campaign type, so if you’re running a format that doesn’t support it, you need to be even more proactive with creative variety and tighter audience windows.

4) Eliminate “wasted remarketing” with smart exclusions

One of the fastest ROI wins in remarketing is excluding people who shouldn’t keep seeing the same ads. This is especially important if your “conversion” is a lead and you’re not pushing offline conversion quality back into the platform yet.

At minimum, build exclusions around recent converters (or qualified leads), employees/internal traffic where possible, and low-intent page visitors who never moved deeper into the funnel. This keeps your remarketing budget concentrated on the users most likely to convert, instead of re-buying impressions from people who already finished (or who were never a fit).

5) A practical recovery checklist you can run in under 30 minutes

  • Audience eligibility: confirm the list has at least ~100 active users in the last 30 days for the surfaces you’re targeting, and that the segment is Open (and not revoked/closed).
  • Membership duration: ensure durations match your sales cycle and aren’t so short that lists collapse below minimum thresholds.
  • Targeting mode: verify whether audiences are set to Targeting vs Observation at the right level (campaign vs ad group) based on your intent.
  • Consent signals: confirm ad_storage and ad_personalization are implemented correctly for eligible traffic; investigate any CMP changes that may restrict remarketing.
  • List sources: for Customer Match, confirm uploads are processed and refreshed; for dynamic remarketing, confirm data sources are processed and accounts are properly linked.
  • Performance controls: add converter exclusions, tighten recency tiers, and apply frequency management where supported before you change bids or targets.

Let AI handle
the Google Ads grunt work

Try now for free
Issue / Question What you see in the account Likely root cause Key checks & fixes Relevant Google Ads docs
Delivery vs. conversion problem Delivery issue: Very low impressions and spend.
Performance issue: Normal spend and traffic, but poor CPA/ROAS.
Mixing these two problems leads to the wrong “fix” (e.g., changing bids/creatives when the campaign actually can’t serve, or hunting for delivery issues when tracking/intent is the problem).
  • First split your diagnosis: is the problem spend/delivery or conversion performance?
  • If spend ≈ $0, focus on eligibility, targeting mode, tagging, and consent.
  • If spend is healthy but results are weak, focus on segmentation, exclusions, measurement quality, and frequency/creative.
How your data segments work, About conversion windows
Audience too small or ineligible Campaign has very low/no impressions and can’t scale, even with competitive bids and budgets. Lists show some users but campaigns still don’t serve consistently. Remarketing lists don’t have enough active users in the last 30 days for the surface you’re targeting, or the segment is Closed / not refreshing. For example, Google Display Network generally needs ≥100 active users in the past 30 days; Search and YouTube remarketing often need ≥1,000 active users. ([support.google.com](https://support.google.com/google-ads/answer/2472738?utm_source=openai))
  • Check each data segment’s 30‑day active users and status (Open vs Closed).
  • Aim for at least ~100 recent active users as a baseline; more for Search/YouTube.
  • Increase membership duration if lists are “evaporating” (too short windows and low traffic).
  • Avoid overly long windows like “All visitors – 540 days” for performance; use them only if you truly need broader reach.
How your data segments work
Targeting vs Observation misconfiguration
  • Campaign serves to everyone instead of only past visitors.
  • Or, delivery collapses to near zero when you “turn on remarketing.”
Audience segments were added in the wrong mode:
  • Observation when you intended a remarketing‑only setup.
  • Targeting on small lists, unintentionally restricting reach.
New Search/Shopping campaigns often default to Observation. ([support.google.com](https://support.google.com/google-ads/answer/7067580/use-google-ads-editor-to-set-up-remarketing-campaigns?utm_source=openai))
  • At campaign/ad‑group level, confirm whether you need:
    • Targeting → serve only to people in your data segment (pure remarketing/RLSA).
    • Observation → serve based on keywords but layer the audience for reporting/bid adjustments.
  • Adjust the setting to match your strategy and list size.
About Targeting and Observation settings, Use Google Ads Editor to set up campaigns with your data segments
Tags and consent blocking list growth
  • Remarketing lists stuck at 0 users or flatlined after previously growing.
  • Performance drop after a consent banner, tag/container, or site/domain change.
  • Global site tag / GTM container not firing on all relevant pages.
  • Consent mode signals such as ad_storage or ad_personalization are denied, which disables or heavily restricts remarketing. ([support.google.com](https://support.google.com/tagmanager/answer/10718549?utm_source=openai))
  • Consent Management Platform not passing signals reliably or early enough, so tags run in restricted mode.
  • Verify global tag / GTM is installed on all pages and firing consistently.
  • Audit consent mode:
    • Ensure ad_storage and ad_personalization are granted where legally appropriate.
    • Make sure consent status is available before your remarketing tags fire.
  • Re‑test after any CMP or tag changes.
Tag Manager consent mode support, Consent mode reference
Customer Match list and “no/low volume” issues
  • Customer Match campaigns show “no or low volume” despite uploads.
  • Lists appear to be constantly “In progress,” or shrink over time.
  • Too few active matched users at auction time (list too small or fragmented).
  • Additional targeting layers (geo, devices, segments) overly restrict reach.
  • Memberships older than the 540‑day maximum retention are expiring, and lists aren’t refreshed. ([support.google.com](https://support.google.com/google-ads/answer/6276125?hl=en&utm_source=openai))
  • Frequent uploads keep the list in a near‑constant processing state.
  • Confirm list has at least 100+ recent active users; aim higher for stable volume.
  • Refresh uploads regularly so users don’t age out of the 540‑day window.
  • Relax unnecessary targeting layers while testing list health.
  • Allow up to ~48 hours for new uploads to finish processing.
Create a customer list, Fix Customer Match issues with list upload or low volume, Customer Match Best Practices
Dynamic remarketing / feed problems
  • Dynamic remarketing campaign created but shows very limited impressions.
  • Feed or business data source shows “under review” or disapproved items.
  • First processing cycle not completed or feed under review.
  • Merchant Center or business data not properly linked to the campaign.
  • Feed or business type not allowed for your category.
  • Confirm the account is correctly tagged for dynamic remarketing and that the appropriate business type/parameters are implemented.
  • Check data feed status and policies; resolve any disapprovals.
  • Verify Google Ads–Merchant Center or business data linking and that the correct feed is attached at campaign level.
Set up a dynamic remarketing campaign
Poor segmentation (“All visitors – 540 days”)
  • Remarketing spends but CPA/ROAS are weak.
  • Audiences are built on a single broad segment (e.g., All visitors).
High‑intent and low‑intent users are mixed together:
  • Cart abandoners and product viewers are treated like bounced blog readers.
  • Bidding and creative can’t differentiate between strong and weak intent, creating noisy signals for Smart Bidding.
  • Build separate segments by intent and recency:
    • Recent product viewers (short windows).
    • Cart/lead starters (very short windows).
    • Past converters (exclude or upsell with different creative).
  • Align membership durations with your real sales cycle.
  • Tailor messaging and offers per segment instead of using generic ads.
How your data segments work
Conversion tracking gaps and conversion delay
  • Remarketing “looks bad,” especially in recent date ranges.
  • Fewer conversions than expected vs. backend/CRM.
  • Under‑counting due to missing tags, consent restrictions, or misconfigured conversion actions.
  • Not accounting for conversion lag and conversion windows — recent days under‑report true performance. ([support.google.com](https://support.google.com/google-ads/answer/9119707?utm_source=openai))
  • Automated bidding constantly reset by frequent Target CPA/ROAS changes, so it never stabilizes.
  • Audit conversion setup and ensure the Google tag or Google Analytics integration is working for all final actions.
  • Check conversion windows and attribution; avoid judging performance too close to “today.”
  • Allow enough time and volume for Smart Bidding to learn before making big target changes.
Set up your web conversions, About conversion windows, About data freshness
Ad fatigue & frequency problems
  • Display/Video remarketing shows falling CTR and CVR over time.
  • Users complain about “seeing your ads everywhere.”
Small audiences are exposed to the same creatives too many times, causing fatigue and declining incremental impact per impression.
  • Use frequency management in Display and Video campaigns to cap impressions per user where supported. ([support.google.com](https://support.google.com/google-ads/answer/6034106/use-ad-rotation-and-frequency-capping?hl=en-GB&utm_source=openai))
  • Shorten membership durations on lower‑intent segments to reduce over‑exposure.
  • Rotate creatives and test different messages/offers for the same segment.
Use ad rotation and frequency capping, Frequency capping: Definition
“Wasted remarketing” and weak exclusions
  • Remarketing budget spends heavily but includes people who already converted or were never qualified.
  • Leads of mixed quality, with no feedback loop into Google Ads yet.
Too many low‑value users remain eligible:
  • Recent converters aren’t excluded.
  • Internal traffic and very low‑intent page visitors keep getting ads.
  • Build and apply exclusions for:
    • Recent purchasers or qualified leads.
    • Employees/internal IPs where feasible.
    • Low‑intent pages (e.g., bounced blog visits) that never progressed.
  • Concentrate remarketing on users who reached deeper funnel stages.
How your data segments work
30‑minute recovery checklist A remarketing campaign that “suddenly stopped working” or has drifted into poor performance. Often a combination of:
  • Shrinking/ineligible lists.
  • Changed consent/tagging behavior.
  • Misaligned targeting mode, durations, or exclusions.
  • No controls on frequency or intent tiers.
In order:
  • Audience eligibility: Check 30‑day active users ≥ ~100 and segment status = Open.
  • Membership duration: Align with your sales cycle; avoid windows so short that lists fall under thresholds.
  • Targeting mode: Confirm Targeting vs Observation at the correct (campaign/ad‑group) level.
  • Consent signals: Validate ad_storage and ad_personalization for eligible traffic.
  • List sources: For Customer Match, ensure uploads are processed and refreshed; for dynamic remarketing, confirm feed linking and policy approval.
  • Performance controls: Add converter exclusions, refine recency tiers, and apply frequency management before changing bids or bid targets.
How your data segments work, About Targeting and Observation settings, Consent mode reference, Create a customer list, Set up a dynamic remarketing campaign, Use ad rotation and frequency capping

If your remarketing campaign isn’t performing, the first step is to separate a delivery problem (low impressions/spend because the campaign can’t serve) from a conversion-performance problem (normal traffic but weak CPA/ROAS). From there, the usual culprits tend to be ineligible or shrinking audiences (not enough active users), “Targeting” vs “Observation” misconfigurations that either over-restrict or over-broaden reach, tagging and consent mode changes that quietly stop lists from growing, overly broad segmentation (e.g., lumping all visitors together), missing exclusions for recent converters, ad fatigue from excessive frequency, and measurement issues like conversion lag or tracking gaps. If you want a faster way to spot which of these is happening in your account, Blobr connects to Google Ads and continuously checks for these common failure points, then surfaces clear, prioritized actions—supported by specialized AI agents (including landing-page alignment agents like the Campaign Landing Page Optimizer and Keyword Landing Optimizer) so you can fix performance blockers without spending hours digging through settings.

Start with the right diagnosis: is this a delivery problem or a conversion problem?

When someone tells me “remarketing isn’t performing,” I immediately clarify whether the campaign is not serving (few/no impressions, no spend), or it’s serving but not converting (impressions and clicks happen, but CPA/ROAS is poor). These two situations have completely different root causes—and if you treat a delivery problem like a performance problem (or vice versa), you can waste weeks “optimizing” the wrong thing.

Remarketing is also more sensitive than most campaign types to list eligibility, privacy/consent signals, and configuration details like “Targeting vs Observation.” That’s why it can look like it suddenly “broke” even when nothing changed in your ads.

  • If spend is near $0: prioritize audience eligibility, targeting mode, and consent/tagging.
  • If spend is normal but results are weak: prioritize segmentation, exclusions, conversion measurement quality, and frequency/creative strategy.

When remarketing isn’t spending: fix audience eligibility, setup, and privacy signals

1) Your audience is not eligible (or too small “where it counts”)

Remarketing can’t run if the platform can’t match enough recent, eligible users to actually serve ads. Audience size is not just “how many cookies touched your site in the last year.” Eligibility is based on active users in a recent window, and it varies by list type and where you’re trying to serve.

As a practical baseline, if your remarketing list has fewer than 100 active users in the last 30 days, expect limited or zero delivery in many remarketing surfaces. Even if your list shows “800 users,” the campaign may still struggle because not everyone is active/available at the moment ads would be served.

Also watch for list health and status issues. If a segment is set to Closed, it won’t keep growing. And if a segment hasn’t been used in targeting long enough, it can be automatically closed—meaning it will stop adding new users until you reopen it. On top of that, membership duration matters: if you’re using a very short duration (for example, 7 days) and you don’t have steady traffic, your list can “evaporate” below eligibility thresholds.

2) You built the campaign, but you’re not actually targeting the list

This is one of the most common (and most frustrating) setup mistakes: you add audiences, but you added them in Observation mode when you intended Targeting mode (or the reverse).

In Search (and commonly Shopping-style setups), new campaigns often default audience handling to Observation—meaning your ads can still show to everyone matching keywords, and the audience is just a reporting/bid-adjustment layer. If you intended a true “RLSA-only” style campaign (serve only to past visitors), you must ensure the audience setting is Targeting. Conversely, if you accidentally set Targeting in a keyword campaign with tiny lists, you can choke delivery to near zero because you’ve restricted the reach to an audience that may not meet minimum thresholds.

Bottom line: confirm whether the campaign/ad group is meant to be a remarketing-only campaign (Targeting) or a bid-layer / reporting campaign (Observation), and verify it matches your intent.

3) Your list isn’t being built because tagging and consent signals are blocking it

If your remarketing lists are staying at 0—or they were growing and then flatlined—this is usually not a bidding problem. It’s almost always one of these:

  • The site isn’t fully tagged: the base tag isn’t present across the site, or it’s firing inconsistently.
  • Consent settings disable ad personalization: if ad_personalization is denied, personalized advertising features like remarketing don’t receive data. If ad_storage is denied, advertising cookies can’t be read/written, which can drastically reduce list growth and matchability.
  • Your CMP/consent framework isn’t passing signals reliably: if the consent string isn’t available fast enough, tags can run in a restricted mode where remarketing features may be disabled for that traffic.

This is why remarketing performance can drop right after a consent banner change, a tag manager container update, a site redesign, or a move to a new domain/subdomain setup. The campaign didn’t “forget how to perform”—it simply stopped receiving eligible audience inputs at the same rate.

4) Customer Match and dynamic remarketing: “processing” and “freshness” bottlenecks

Two special cases deserve their own callout because they often look like “the campaign is dead” when the real issue is eligibility and processing.

Customer Match: if you’re using customer lists, remember that lists need sufficient active matched people to serve at the moment of auction. Too-small lists, overly narrow layering on top of the list, or insufficient bids/budget can all lead to “low or no volume.” Upload processing can also take time, and frequent re-uploads can keep lists in a near-constant processing state. From a maintenance perspective, Customer Match memberships also have a maximum retention window; if you don’t refresh/update the list, older memberships can age out and quietly reduce your usable match pool.

Dynamic remarketing (feed-based): if you rely on a product or business data source, your first processing cycle can take a few business days, and the campaign’s ability to serve can hinge on correct account linking and feed readiness. If the campaign is configured correctly but the data source is “under review,” mis-linked, or disallowed for your category settings, delivery can be extremely limited regardless of bids or creative quality.

When remarketing is spending but results are weak: raise intent, control repetition, and give bidding clean signals

1) Stop treating all visitors the same (segment by intent + recency)

Generic “All Visitors – 540 days” remarketing is easy to launch, but it’s rarely efficient. A visitor from yesterday who abandoned a checkout is not the same as someone who bounced from a blog post eight months ago. When you mix them, your messaging gets generic, your conversion rate drops, and your bidding system gets noisy signals.

I recommend structuring remarketing around a few clear intent tiers and short, meaningful membership windows that reflect your sales cycle. For example: product viewers (short window), cart/lead starters (very short window), and past converters (exclude or upsell with a different message). The goal is to make the audience’s “why” obvious, so your offer and creative can be specific instead of broad.

2) Fix conversion measurement first—then judge performance

Remarketing is often judged harshly because it’s closer to the bottom of the funnel, where tracking gaps are more visible. Before you declare “remarketing doesn’t work,” make sure you’re not undercounting conversions due to consent restrictions, missing tags, or configuration gaps.

Also account for conversion delay. If you’re optimizing or reporting too close to today, you can end up “optimizing away” campaigns that are actually working but have conversions that arrive days later. Depending on your conversion window settings, conversions can be reported long after the click, so recent date ranges often understate true performance.

If you’re using automated bidding, give the system a fair chance to stabilize. Learning and recalibration time is influenced by how many conversions you generate, how long your conversion cycle is, and which bid strategy you’re using. Big, frequent target changes (for example, repeatedly moving Target CPA/ROAS every couple of days) can create volatility that looks like poor remarketing performance, when it’s really just constant re-learning.

3) Manage ad fatigue with frequency discipline and creative rotation

Remarketing audiences are smaller by design, so fatigue happens fast. If people see the same message too many times, CTR and conversion rate drop, and you end up paying for repeated exposure that stops moving users forward.

For Display and Video campaigns, use frequency management to limit how often the same person sees your ads. This protects performance while you test messaging. Keep in mind that frequency capping isn’t available in every campaign type, so if you’re running a format that doesn’t support it, you need to be even more proactive with creative variety and tighter audience windows.

4) Eliminate “wasted remarketing” with smart exclusions

One of the fastest ROI wins in remarketing is excluding people who shouldn’t keep seeing the same ads. This is especially important if your “conversion” is a lead and you’re not pushing offline conversion quality back into the platform yet.

At minimum, build exclusions around recent converters (or qualified leads), employees/internal traffic where possible, and low-intent page visitors who never moved deeper into the funnel. This keeps your remarketing budget concentrated on the users most likely to convert, instead of re-buying impressions from people who already finished (or who were never a fit).

5) A practical recovery checklist you can run in under 30 minutes

  • Audience eligibility: confirm the list has at least ~100 active users in the last 30 days for the surfaces you’re targeting, and that the segment is Open (and not revoked/closed).
  • Membership duration: ensure durations match your sales cycle and aren’t so short that lists collapse below minimum thresholds.
  • Targeting mode: verify whether audiences are set to Targeting vs Observation at the right level (campaign vs ad group) based on your intent.
  • Consent signals: confirm ad_storage and ad_personalization are implemented correctly for eligible traffic; investigate any CMP changes that may restrict remarketing.
  • List sources: for Customer Match, confirm uploads are processed and refreshed; for dynamic remarketing, confirm data sources are processed and accounts are properly linked.
  • Performance controls: add converter exclusions, tighten recency tiers, and apply frequency management where supported before you change bids or targets.