Google Ads Customer Match Best Practices

Last updated: September 7, 2026

1. Executive Summary

Customer Match is Google Ads’ first-party data activation layer. When implemented correctly, it lets you reach known customers, suppress them from acquisition spend, and feed high-quality signals into Smart Bidding and Performance Max. Five principles govern every decision in this guide.

• First-party data quality determines everything. Match rates of 29%–62% are typical across advertisers.^[1] The gap between the low and high ends is almost always explained by data normalization, identifier completeness, and list recency—not by audience size alone.
• Upload all available identifiers, never just email. Google explicitly recommends including email, phone, mobile device ID, and postal address for each contact in the same CSV row. Multiple identifiers for the same person in the same row improve match accuracy.^[1][2]
• Refresh on a schedule you can defend to your compliance team. A Customer Match list membership expires after 540 days; a list drops below eligibility when it has fewer than 100 members updated within that window.^[1][3][4] Automate daily refreshes via the Data Manager API or a certified upload partner.
• Use Observation first, Targeting deliberately. All policy-compliant advertisers can use Customer Match in Observation and Exclusions. The Targeting mode—which restricts delivery exclusively to matched users—requires the account to have 90 days of Google Ads history and more than $50,000 USD in total lifetime spend.^[1][5]
• Consent and compliance are upstream problems, not ad-platform afterthoughts. Consent must be recorded at collection time, mapped in your CRM, and passed through the upload workflow. Google requires the consent field in API uploads, and its EU User Consent Policy adds additional obligations for EEA audiences.^[4][6]

2. Benchmarks and Numbers at a Glance

Metric	Typical range or threshold	Applies when	Source
Customer Match match rate (typical range)	29%–62% of uploaded records	All advertisers; Google Help page; no sample size stated	[1]
Minimum list size for ongoing eligibility	100 members added or updated within 540 days	All Customer Match lists; Google policy requirement	[3][4]
Maximum list membership duration	540 days from add/refresh date	All Customer Match lists; entries older than 540 days stop being eligible	[1][4]
Lifetime spend threshold for Targeting mode	More than $50,000 USD total lifetime spend	Required to use Customer Match in Targeting, Observation, and manual bid adjustments	[1][7]
Account history threshold for Targeting mode	90 days of Google Ads history	Required alongside the $50,000 lifetime spend threshold	[1][7]
API bulk upload: identifiers per AddOperations call	Up to 10,000 identifiers per call (optimal)	Google Ads API OfflineUserDataJob; vendor guidance from Google developer docs	[6]
API bulk upload: maximum operations per job	No more than 1,000,000 operations per job (optimal)	Google Ads API OfflineUserDataJob; vendor guidance from Google developer docs	[6]
Recommended API refresh frequency	At least once per day (append to lists)	API-based upload workflows; Google best-practices page	[1]
Third-party claim: Search campaign list minimum (post-2025 change)	100 users (reduced from 1,000)	Search campaigns only; third-party reporting, not confirmed in Google’s own help docs as of article date	[8]
Third-party claim: minimum active members before any surface serves	1,000 active members	All surfaces; third-party agency blog; not confirmed in Google’s official help docs—treat as conservative floor	[9]
API access cutoff for inactive Customer Match adopters	April 1, 2026 (requests from tokens with no CM uploads Oct 1, 2025–Mar 31, 2026 blocked)	Developer tokens using OfflineUserDataJobService or UserDataService for Customer Match	[10][11]

Note on conflicting minimums: Google’s official help documentation cites 100 members as the eligibility threshold.^[3][4] A third-party source claims 1,000 active members are required before any surface serves ads.^[9] Another third-party source reports the Search minimum was reduced from 1,000 to 100.^[8] Because these figures conflict, this guide recommends the most conservative approach: build lists to at least 1,000 active members before expecting consistent delivery, while treating 100 as the hard eligibility floor below which lists are disabled.

3. What Customer Match Is and Why It Matters

Customer Match is Google Ads’ mechanism for activating the customer data you already own. You upload hashed identifiers—email addresses, phone numbers, postal addresses, or mobile device IDs—directly to Google, which then matches those identifiers to signed-in Google users. The matched audience becomes available for targeting, bid adjustment, exclusion, or as a signal for automated bidding systems across Search, Shopping, Performance Max, Demand Gen, YouTube, and Gmail.^[3][4]

The strategic importance of Customer Match has grown sharply since 2024 as third-party cookie deprecation has accelerated across browsers. Advertisers who built strong Customer Match programs before that shift retained durable audience signals that cookie-based remarketing could no longer supply. In 2026, Customer Match is the most reliable way to bring deterministic, identity-based audience data into Google Ads without depending on cookie or pixel continuity.^[1]

Customer Match data also serves as the highest-quality seed signal available to Google’s AI systems. Performance Max and Smart Bidding treat Customer Match lists as direct inputs into their optimization models. A high-quality, frequently refreshed Customer Match list of your best customers gives these systems a concrete signal of what a valuable user looks like, which can improve targeting for users who are not yet in the list.^[1][3]

Three broad use cases justify most Customer Match investment:

• Retention and upsell: Reach current customers with relevant offers differentiated from acquisition messaging.
• Suppression: Exclude current customers or recent converters from prospecting campaigns to prevent wasted spend and messaging mismatches.
• Acquisition signal: Use your best-customer list as a seed for Smart Bidding and optimized targeting to find similar high-value users.^[1][3]

4. Eligibility and Policy Requirements

Customer Match is not available to every Google Ads account by default in all modes. Google’s policy distinguishes between two tiers of access, and the requirements differ meaningfully.^[1][5]

Tier 1: Observation and Exclusions (all eligible advertisers)

Any advertiser with a good history of policy compliance and a good payment history may use Customer Match in Observation mode and as an Exclusion. These modes do not restrict delivery to the matched audience; they layer the audience on top of existing targeting for reporting or suppression purposes. There is no minimum lifetime spend and no minimum account age requirement for these modes.^[5]

Tier 2: Targeting, Observation with bid adjustments (higher threshold)

To use Customer Match in Targeting mode—which restricts campaign delivery exclusively to the matched audience—and to apply manual bid adjustments to Customer Match audiences, an account must meet two criteria simultaneously:^[1][7]

• 90 days of Google Ads account history
• More than $50,000 USD in total lifetime Google Ads spend

Accounts that do not meet these thresholds can still use Customer Match; they are simply limited to Observation and Exclusion uses until the thresholds are met.

First-party data requirement

Google’s Customer Match policy is explicit: you may only upload customer information you collected directly from users in a first-party context—from your website, app, physical store, or another situation where the customer shared their information directly with you. Purchasing third-party lists, appending data from data brokers, or uploading data you did not collect directly is a policy violation.^[4]

Privacy policy disclosure requirement

Your privacy policy must disclose that you share customer data with third parties to perform services on your behalf. This disclosure must be in place before you upload any Customer Match data. Google’s EU User Consent Policy adds a separate consent requirement for EEA users; outside the US, confirm whether local privacy law (such as GDPR in the EU or PIPEDA in Canada) imposes additional consent obligations before uploading.^[4]

Approved upload methods

Google requires the use of its approved upload methods: the Google Ads UI, the Google Ads API (via OfflineUserDataJob for accounts active before April 1, 2026), the Data Manager API (the primary path for new API adopters after April 1, 2026), or a certified Customer Match upload partner.^[4][10][11]

5. Data Sources and List Building

The quality of a Customer Match program is determined almost entirely by the quality of the underlying list. Google allows you to upload data from any first-party source where customers shared their information directly with you. In practice, US advertisers draw from four primary sources.^[4]

Accepted first-party data sources

• Website sign-ups, account registrations, and checkout forms
• Mobile app onboarding and in-app data collection
• Physical store transactions and loyalty program enrollments
• Phone or in-person interactions where data was captured with consent^[4]

Identifiers you can upload

Google supports four types of identifiers in Customer Match uploads. Uploading all identifiers you have for each contact—in the same CSV row—is the single most impactful step you can take to improve match rate.^[1][2]

Identifier type	Required fields	Normalization before hashing
Email address	Email	Lowercase; remove leading/trailing spaces; remove periods before gmail.com or googlemail.com domain
Phone number	Phone	E.164 format (e.g., +12025551234); remove spaces, dashes, parentheses
Postal address	First Name, Last Name, Country, Zip	Lowercase; remove leading/trailing spaces; use ISO 3166-1 alpha-2 country codes
Mobile device ID	Mobile Device ID (IDFA or AAID)	Lowercase; no hashing required for device IDs

Source: Google Ads upload specification.^[2]

Segmentation strategy

Do not upload a single undifferentiated CRM export. Segmenting your list by customer lifecycle stage before upload lets you apply different bidding, messaging, and exclusion logic to each segment. Recommended segments for most US advertisers include:

• High-LTV active customers (purchased within 90 days, top 20% by revenue): use as a positive signal in Performance Max and Smart Bidding; apply bid uplift in Search Observation.
• Lapsed customers (no purchase in 180–540 days): use in win-back campaigns with dedicated messaging and offers.
• Recent converters (purchased within 30 days): exclude from prospecting campaigns immediately post-conversion.
• Leads or trial users who never converted: use in nurture campaigns with separate creative from customer retention campaigns.

6. Uploading, Hashing and Automation

Getting data into Google Ads correctly is a technical requirement, not just a best practice. Errors in file format, hashing, or normalization are the most common cause of low match rates and upload failures.^[2]

File format requirements

• File type: CSV only
• Encoding: ASCII or UTF-8; UTF-16 is not supported and will cause upload errors^[2]
• Column headers: use exact English headers Google specifies: Email, Phone, First Name, Last Name, Country, Zip
• Missing or misspelled headers cause silent field-level failures; the row may upload but the identifier will not match^[2]

Hashing

Google requires identifiers to be hashed with SHA-256 before upload, using hex encoding. You may hash the data yourself before uploading, or you may allow the Google Ads UI to hash the data on your behalf using the same standard. Either approach is acceptable; hashing yourself gives you more control over the normalization step and is strongly recommended for API-based workflows where you can verify the output before transmission.^[2]

The normalization sequence before hashing is strict. For email addresses: remove leading and trailing spaces, convert to lowercase, remove periods that appear before the domain in gmail.com and googlemail.com addresses (e.g., [email protected] becomes [email protected]). For phone numbers: convert to E.164 format (+1XXXXXXXXXX for US numbers). For names and addresses: lowercase, remove leading/trailing spaces.^[2]

Upload methods compared

Method	Best for	Recommended refresh cadence	Key constraint
Google Ads UI (manual upload)	Small lists, one-time or infrequent uploads, teams without API access	Weekly minimum; set a calendar reminder	Manual process; prone to drift if reminders are missed
Google Ads API (OfflineUserDataJob)	Accounts active in Customer Match before April 1, 2026	At least once per day[1]	Blocked for new adopters after April 1, 2026[10][11]
Data Manager API	New API adopters after April 1, 2026; all accounts going forward	At least once per day	Requires API credentials and developer setup
Certified upload partner / third-party automation	Teams using CRM platforms (e.g., Salesforce, HubSpot) with native connectors	Near real-time or daily, depending on platform	Partner must be Google-certified; consent signals must pass through

API upload workflow (OfflineUserDataJob)

For accounts using the Google Ads API, Google’s recommended workflow is: (1) create the customer list, (2) create an OfflineUserDataJob, (3) add user data in batches of up to 10,000 identifiers per call, (4) run the job, (5) poll for success, (6) verify match rate, (7) target the list. Keep individual jobs to no more than 1,000,000 operations for optimal processing. Use one account to modify a list and avoid mixing create and remove operations in the same job.^[6]

7. Improving Match Rates and Data Hygiene

Match rates of 29%–62% are the published typical range from Google.^[1] A match rate at the low end almost always indicates a solvable data quality or normalization problem, not an inherent ceiling. The diagnostic process is straightforward: if your match rate is below 40% and you are uploading only email addresses, the first corrective action is to add phone numbers and postal addresses to the same rows.

The five most common match rate killers

• Single-identifier uploads. Email-only files lose all the users whose email address has changed or whose Google account is registered to a different address. Adding phone and address data significantly expands coverage.^[1][2]
• Incorrect normalization before hashing. An email address hashed without lowercasing or without removing the gmail.com period variant will not match even if the user exists in Google’s system. The SHA-256 hash of [email protected] and [email protected] are completely different strings.^[2]
• Stale data. Email addresses or phone numbers that users changed 18 months ago generate no match. Regular CRM hygiene—removing bounced emails, updating phone numbers, deduplicating records—directly improves match rate.^[1]
• UTF-16 encoding. Files saved in UTF-16 from some CRM exports will fail to upload or will produce silent field-level errors. Always specify UTF-8 in your export settings.^[2]
• Splitting identifiers across rows. Uploading a separate row for the same person’s email and another row for their phone number does not combine into a stronger match. Both identifiers must be in the same row to be associated with the same user record.^[1]

CRM hygiene steps that improve match rate

• Remove duplicate records before each export
• Standardize phone numbers to E.164 at CRM entry, not at export time
• Flag and remove role-based email addresses (e.g., info@, support@) that no individual owns
• Remove records with invalid email syntax
• Confirm country code mapping for any non-US contacts before including them
• Maintain a suppression file for contacts who have explicitly opted out of ad targeting^[1][2]

Using match rate as a diagnostic

Google states explicitly that match rate should be used as a benchmark to diagnose formatting and data-quality issues, not as a performance KPI in itself.^[1] The correct monitoring cadence is: check match rate after every upload; compare it to the prior upload’s rate; investigate any drop of more than 5 percentage points as a potential pipeline or normalization failure.

8. Targeting vs. Observation and Exclusions

The single most consequential configuration decision in Customer Match is choosing between Targeting and Observation. The distinction affects delivery volume, audience eligibility requirements, and campaign structure.^[5]

Observation mode

In Observation mode, the Customer Match audience is layered onto the campaign without restricting delivery. The campaign continues to reach all users who match the existing keyword, placement, or signal targeting, and it also collects performance data segmented by whether a user is in the Customer Match list. This data can then be used to apply bid adjustments—for example, increasing bids by 25% for users on the high-LTV customer list—without limiting the campaign’s total reach.^[5]

Observation is the correct default for:

• Search campaigns where volume matters and you want bidding signals, not audience restriction
• Any account that has not yet crossed the $50,000 lifetime spend threshold^[1]
• Testing and measurement phases where you want to understand how Customer Match segments perform before making structural changes

Targeting mode

In Targeting mode, the campaign delivers exclusively to users who are matched in the Customer Match list. This is appropriate for dedicated retention campaigns, loyalty offers, or win-back campaigns where reaching non-customers would waste budget and misalign messaging. Targeting mode requires the account to have 90 days of history and more than $50,000 in total lifetime spend.^[1][5]

Exclusions

Exclusions remove a Customer Match audience from a campaign’s delivery. Exclusions are available to all policy-compliant advertisers regardless of account age or spend. They are the most universally applicable use of Customer Match and produce measurable budget efficiency gains in acquisition-focused campaigns by preventing spend on users who are already customers.^[5]

Removing lists from Smart Bidding

Google’s documentation notes that you can remove specific Customer Match lists from Smart Bidding and optimized targeting if a list is stale, too broad, or contains segments that would distort automation. This is a distinct setting from campaign-level exclusion and applies specifically to the signals fed to Google’s bidding models.^[3]

9. Customer Match Across Campaign Types

Customer Match behavior, recommended configuration, and strategic value differ by campaign type. The table below maps the recommended use for each surface.

Campaign type	Recommended Customer Match use	Configuration	Key note
Search	Observation with bid adjustments; Exclusions for current customers in acquisition campaigns	Add list at ad group or campaign level; set bid adjustment (+15% to +30% for high-LTV segments)	Targeting mode appropriate only when campaign is intentionally restricted to known audience
Performance Max	Audience signal (not a hard filter); feed high-quality, segmented lists	Add as audience signal in asset group; include high-LTV customer list	PMax treats Customer Match as a signal to Smart Bidding, not a delivery restriction[1][3]
Demand Gen	Audience signal for lookalike expansion; Exclusions for existing customers in acquisition flows	Add exclusion list at campaign level; use best-customer list as signal	Optimized targeting uses Customer Match signals to expand beyond the uploaded list
YouTube	Remarketing, suppression, re-engagement, upsell, churn reduction	Targeting mode for dedicated retention campaigns; Observation for general campaigns	Especially effective for sequential messaging to known high-intent users[4][5]
Gmail / Display	Retention messaging, loyalty offers, lapsed-customer win-back	Targeting mode for dedicated campaigns; Observation for prospecting with bid adjustment	Match rate applies to Gmail directly when email is the uploaded identifier
Shopping	Bid adjustments for high-LTV customers; Exclusions for recent purchasers	Observation with positive bid adjustment on high-LTV list	Customer Match cannot be used as a hard audience filter in standard Shopping

Performance Max and Customer Match

Performance Max does not support hard audience targeting via Customer Match in the same way that Search does. Instead, Customer Match lists function as audience signals—inputs that tell Smart Bidding what a high-value user looks like. The quality of the signal matters more than the size of the list. A 5,000-record list of verified high-LTV customers is a stronger signal than a 50,000-record unfiltered CRM dump.^[1][3]

You can also use Customer Match as an exclusion within Performance Max to prevent existing customers from being served acquisition-focused creative. This is implemented at the campaign level and is one of the highest-ROI configuration decisions available in PMax.^[3]

10. Privacy, Consent and Compliance

Customer Match is a first-party data product, but first-party does not mean unconstrained. Google’s policy, US state privacy law, and federal regulations each impose requirements that must be met before a single record is uploaded. As of September 2026, US advertisers face a fragmented state-level privacy landscape in the absence of a comprehensive federal consumer privacy law.

Google’s policy requirements

• Your privacy policy must disclose that you share customer data with third parties to perform services on your behalf.^[4]
• You may only upload data you collected in a first-party context where the customer shared it directly with you.^[4]
• For EEA users, Google’s EU User Consent Policy requires valid consent signals to be passed through the upload workflow. If you use a Customer Match upload partner for EEA data, the consent signal must flow through the partner.^[3][4] Outside the US, confirm local consent requirements (e.g., GDPR in the EU, PIPEDA in Canada) before including non-US records in a Customer Match upload.

US state privacy law (as of September 2026)

Multiple US states have enacted comprehensive consumer privacy laws with relevance to Customer Match workflows. California (CCPA/CPRA), Colorado, Connecticut, Virginia, Texas, and others require businesses to honor opt-out requests for the sale or sharing of personal information for advertising purposes, provide notice of data uses, and in some states (California, Connecticut) honor universal opt-out signals. A Customer Match upload that includes records of California residents who have submitted a verified opt-out request under CPRA is a potential regulatory violation independent of Google’s platform policy.

The practical implication: your CRM must maintain a suppression file of opted-out contacts, and your upload workflow must exclude those records before any data reaches Google. This is not a Google Ads configuration setting—it is a data governance process that must be implemented upstream in your CRM or CDP.^[4][6]

Consent handling in API workflows

Google’s API documentation requires populating the consent field in customer_match_user_list_metadata when creating OfflineUserDataJob requests. Omitting this field is a compliance gap even if your CRM records show consent. The consent status must be passed through the API payload, not inferred from the presence of a record in the export.^[6]

Practical consent architecture

• Record the consent basis, timestamp, and permitted ad uses in the CRM at the point of data collection—not at the point of upload.
• Map CRM consent fields to the API consent parameter so only users with valid marketing consent are included in exports.
• Maintain a real-time or near-real-time suppression process: when a user opts out, removes consent, or submits a deletion request, their record must be removed from active Customer Match lists within the timeframe required by applicable law.
• Audit your upload logs quarterly: verify which identifiers were sent, under what consent status, and whether suppression records were correctly excluded.^[3][4][6]

11. Common Mistakes to Avoid

The following errors are the most frequently observed in Customer Match implementations and each has a direct, measurable cost in match rate, compliance exposure, or wasted budget.

Mistake 1: Uploading a single identifier type when multiple are available

Email-only uploads are the most common configuration error. Google explicitly states that uploading all available identifiers for each contact in the same row improves match accuracy. An email-only file with 20,000 records and a 31% match rate produces 6,200 matched users. The same 20,000 records with email + phone + address could plausibly reach 45%+ match rate—9,000+ matched users—from the same underlying list without acquiring a single new contact.^[1][2]

Mistake 2: Treating match rate as a vanity metric rather than a diagnostic

Match rate is not a KPI to optimize for its own sake. It is a diagnostic signal. A drop from 48% to 39% between two consecutive uploads is an actionable alert that something changed in the data pipeline—a normalization script broke, a CRM export setting changed, or a new data source was merged with inconsistent formatting. Monitor match rate after every upload and investigate changes above 5 percentage points.^[1][6]

Mistake 3: Letting lists age out of the 540-day window

A Customer Match list that is never refreshed begins losing eligible members after 540 days. A list that drops below 100 members added or updated within the last 540 days loses eligibility entirely.^[3][4] This most commonly happens to one-time campaign lists that were created for a promotion and then abandoned. Assign every list a refresh owner and a refresh schedule at creation time.

Mistake 4: Using Customer Match Targeting mode without meeting the eligibility thresholds

Attempting to configure Targeting mode on an account with less than $50,000 lifetime spend or less than 90 days of history will result in the audience either reverting to Observation mode silently or generating a policy error. Verify eligibility before building campaign structure around Targeting mode.^[1][5]

Mistake 5: Using the Google Ads API for Customer Match without checking the April 1, 2026 access change

Developer tokens that had no Customer Match API requests between October 1, 2025 and March 31, 2026 are restricted from using the Google Ads API for Customer Match as of April 1, 2026. Requests fail with CUSTOMER_NOT_ALLOWLISTED_FOR_THIS_FEATURE. New integrations must use the Data Manager API. Failing to migrate causes upload failures that silently stop list refreshes.^[10][11]

Mistake 6: Not excluding recent purchasers from acquisition campaigns

Running prospecting campaigns without a Customer Match exclusion for recent converters is the most prevalent budget waste pattern in accounts with mature CRM programs. Every click from a user who purchased 10 days ago is an acquisition budget dollar that produced zero new customers.^[3][10]

Mistake 7: Uploading opted-out users

Failing to run a suppression pass before each upload means opted-out users remain in active Customer Match lists. This is a compliance risk under CPRA for California residents and under GDPR for EEA users, and a Google policy violation. Build suppression removal into every upload workflow, not as a periodic cleanup task.^[4]

12. What Changed Recently

The most consequential Customer Match changes in the September 2026 window are not cosmetic policy edits—they are structural shifts in how data reaches Google and how long it remains eligible. Three distinct changes are now fully in effect and require immediate operational attention: the Google Ads API Customer Match access restriction, the 540-day membership duration cap, and the lowered list-size minimum for Search campaigns. A fourth area—the tightening of consent signal requirements—continues to evolve and is covered separately in section 10. Each change below is described with its effective date, the specific rule, the operational impact, and a worked example showing the arithmetic.

The Google Ads API Customer Match Access Restriction (Effective April 1, 2026)

Starting April 1, 2026, Google restricted access to Customer Match functionality in the Google Ads API for developer tokens that recorded no Customer Match requests between October 1, 2025 and March 31, 2026. Accounts and integrations that fall outside that six-month activity window receive a CUSTOMER_NOT_ALLOWLISTED_FOR_THIS_FEATURE error when calling OfflineUserDataJobService or UserDataService.^[10] Google simultaneously announced that the Google Ads API will no longer accept new Customer Match adopters as of April 1, 2026, and that the Data Manager API is now the primary import path for any organization that is migrating or setting up Customer Match for the first time.^[11]

The practical consequence is a bifurcated landscape in September 2026: accounts with active pre-April integrations can continue using the Google Ads API OfflineUserDataJob workflow, while any account that lost its active-token status or is building a new integration must route uploads through the Data Manager API instead.^[11] If your organization has not audited which path it is on, do so before your next scheduled upload.

Outside the US, the same API restriction applies globally; the Data Manager API migration requirement is not limited to any single country or region.

The 540-Day Membership Duration Cap

Customer Match list memberships now carry a hard maximum duration of 540 days. Any user record added or last refreshed more than 540 days ago is automatically removed from eligibility, regardless of how large the overall list is.^[1][4] A list must maintain at least 100 members added or updated within the last 540 days to remain usable in campaigns.^[3][4] This rule makes static, one-time uploads operationally hazardous: a list uploaded once on January 1, 2025 will be fully ineligible by June 25, 2026 unless records have been refreshed.

Note a discrepancy in the research on list-size minimums: Google’s official help documentation states 100 members as the eligibility threshold,^[3][4] while a third-party source claims 1,000 active members are required before serving on any surface,^[9] and a Search Engine Land article reports that Search campaign minimums were lowered from 1,000 to 100 users.^[8] Because sources conflict, the most conservative interpretation is to treat 100 members within the 540-day window as the minimum for eligibility and list survival, and to treat 1,000 members as a practical floor for reliable delivery and bidding signal on most surfaces. Plan for at least 1,000 active, in-window records before expecting consistent campaign performance.

Search Campaign List-Size Minimum Lowered to 100 Users

A Search Engine Land report published May 23, 2025 documented that Google lowered the Customer Match list minimum for Search campaigns from 1,000 to 100 users.^[8] If accurate, this meaningfully expands the window during which a newly built or small-scale Customer Match list can begin influencing Search bids and reporting before it reaches the historical 1,000-member threshold. As noted above, a third-party source still cites 1,000 active members as a requirement for serving,^[9] so the conservative approach is to build toward 1,000 active records for cross-surface reliability while recognizing that Search may begin returning data at 100.

Data Manager API as the New Primary Upload Path for New Integrations

As of April 1, 2026, Google directs all new Customer Match integrations—and any existing integration that lost API eligibility—to the Data Manager API rather than the Google Ads API’s OfflineUserDataJob workflow.^[11] The Data Manager API provides a dedicated Customer Match flow that supports adding user data, checking upload results, and verifying list size through a separate endpoint from the Google Ads API.^[11] For teams building a new integration after April 2026, the Google Ads API path is no longer an option; the Data Manager API is the only supported route.^[11]

For teams maintaining a pre-April 2026 Google Ads API integration that remained active through the October 2025–March 2026 qualifying window, the existing OfflineUserDataJob workflow continues to function. Google recommends keeping individual jobs under 1,000,000 operations for optimal processing and batching user data additions in groups of up to 10,000 identifiers per call.^[6]

Consent Signal Requirements and Policy Tightening

Google’s Customer Match policy continued to tighten consent requirements throughout 2025 and into 2026, culminating in a documented major privacy update that requires advertisers to ensure consent signals are explicitly passed through the upload workflow rather than assumed from the presence of a privacy policy alone.^[4] For EEA users specifically, Google requires that consent signals be passed to the platform through the upload partner or API metadata fields.^[3] In the US, the practical implication is that your privacy policy must explicitly state that customer data is shared with third parties to perform advertising services on your behalf, and that consent must be verifiable at the record level if audited.^[4]

While there was no new Customer Match policy page published within the 30 days immediately preceding September 2026, the April 2026 API access restriction and the consent field requirement in customer_match_user_list_metadata together represent the most operationally impactful documentation changes visible in Google’s official sources as of this writing.^[6][10][11]

Summary of Changes in Effect as of September 2026

Change	Effective Date	Rule or Threshold	Who Is Affected	Source
Google Ads API Customer Match access restriction	April 1, 2026	Tokens with no CM activity October 1, 2025–March 31, 2026 blocked from OfflineUserDataJobService and UserDataService	Any integration built against the Google Ads API that was inactive in that 6-month window	[10]
No new Google Ads API Customer Match adopters	April 1, 2026	New integrations must use the Data Manager API exclusively	Any account or developer building a new Customer Match integration after April 1, 2026	[11]
540-day maximum membership duration	Ongoing; fully in effect 2026	Records not added or refreshed within 540 days are ineligible; list must have ≥100 in-window members	All Customer Match users with static or infrequently refreshed lists	[1][4]
Search campaign list-size minimum lowered	Reported May 2025	Minimum lowered from 1,000 to 100 users for Search campaigns	Advertisers using Customer Match in Search with small lists; conservative floor of 1,000 still recommended for cross-surface serving	[8]
Consent signal required in API metadata	Ongoing; documented in API reference	Consent field must be populated in customer_match_user_list_metadata for OfflineUserDataJob requests	All advertisers using the API upload path, especially for EEA users	[3][6]
Eligibility: 90-day history and $50,000 lifetime spend for Targeting	Ongoing policy	Accounts below either threshold limited to Observation and Exclusions only	Newer accounts and lower-spend advertisers	[1][7]

The single most urgent action for any Google Ads account using Customer Match in September 2026 is to audit whether the existing upload integration is routing through the correct API path (Google Ads API for pre-April eligible tokens, Data Manager API for all others), verify that no records older than 540 days are being relied upon for campaign delivery, and confirm that every record in the upload file was collected under a privacy policy that explicitly discloses third-party advertising data use.^[4][10][11] These three checks address the changes with the highest probability of causing silent list failure or policy non-compliance without triggering an obvious ad-serving error.

---

References

1. [1] https://support.google.com/google-ads/answer/10010286?hl=en support.google.com
2. [2] https://support.google.com/google-ads/answer/10589050?hl=en support.google.com
3. [3] https://support.google.com/google-ads/answer/6379332?hl=en support.google.com
4. [4] https://support.google.com/adspolicy/answer/6299717?hl=en support.google.com
5. [5] https://support.google.com/google-ads/answer/10550383?hl=en support.google.com
6. [6] https://developers.google.com/google-ads/api/docs/remarketing/audience-segments/customer-m… developers.google.com
7. [7] https://www.mbadv.agency/google-ads/audience-targeting www.mbadv.agency
8. [8] https://searchengineland.com/google-slashes-customer-match-list-minimums-in-search-campaig… searchengineland.com
9. [9] https://adstralis.agency/en/blog/remarketing-customer-match-google-ads-australia/ adstralis.agency
10. [10] https://twominutereports.com/blog/google-ads-best-practices twominutereports.com
11. [11] https://developers.google.com/data-manager/api/devguides/audiences/google-ads/customer-mat… developers.google.com