Skip to content
whitead

Fix: Customer Match lists are stale

finding google ads updated 2026.05.25 8 min read
// contents · 7 sections
  1. Why this matters
  2. How to verify the issue
  3. How to fix it
  4. How to confirm the fix worked
  5. Diagnostic checklist — run all five within 7 days of the refresh
  6. Related rules + concepts
  7. Sources

Your Customer Match lists have not been refreshed in months. Google is silently decaying the matched-user count, your match rate is drifting, and the Smart Bidding signal those lists were supposed to feed is going dark.

Why this matters

Customer Match is the only first-party audience surface Google Ads exposes to Smart Bidding, Performance Max audience signals, and new-customer-acquisition goals. Once the list goes stale, three things degrade in parallel. First, the matched-user count contracts as email addresses churn, phone numbers reassign, and Google's own user graph turns over; the segment that activated at 60% match rate six months ago typically loses several percentage points per quarter without a refresh [3]. Second, since April 7, 2025 Google automatically expires any membership added or refreshed more than 540 days ago — list size now decays on a fixed clock regardless of how the data was uploaded [1][2]. Third, Smart Bidding and PMax stop trusting the segment as a signal: when a high-LTV customers list shrinks below the per-network activation threshold (~1,000 active matched users), Google demotes it from audience signal to noise, and the optimization loop reverts to broad behavioural targeting [3].

For e-commerce, this shows up as PMax shifting spend toward acquisition-only auctions even when you have asked it to bias toward retention. For B2B, the closed-won and lapsed-account lists that feed Enhanced Conversions for Leads and OCI quality checks lose coverage, so Smart Bidding makes value-based decisions on a thinner ground truth. The fix is procedural, not technical — but operationally it is the audience-hygiene rule most agencies ship without and then quietly accept the drift.

How to verify the issue

  1. Open the audience library. Tools → Audience manager → Your data segments. Filter to Type: Customer list.
  2. Inspect each active list's last-upload date. The column Last modified (or Date updated depending on UI version) shows when the membership was last refreshed. Anything older than 90 days is yellow; older than 180 days is the audit trigger; older than 540 days has already lost rows automatically [1][2].
  3. Check size and match rate. Open each list, look at List size per network (Search / YouTube / Display / Gmail) and Match rate on the segment detail page. A list previously at 60% match that has drifted to <50% is decaying; a list reporting <100 active members on any network has tipped under Google's serving threshold [5][6].
  4. Cross-reference recommendations. Recommendations tab — Google surfaces REFRESH_CUSTOMER_MATCH_LIST recommendations for any list it has flagged internally as stale. If the recommendation exists, the rule fires regardless of your own last_modified reading.
  5. Verify upload cadence in the source system. If the list is supposed to sync from a CRM (HubSpot, Salesforce, Klaviyo) or via Google Ads Data Manager, confirm the job is actually running. Connector breakage is the most common silent failure mode — the UI shows the segment exists, but no rows have flowed in 60+ days.

How to fix it

  1. Decide refresh cadence per list (15 min). Customers who churn fast (subscription, retention, lapsed) need a 30-day refresh; high-LTV and closed-won lists tolerate 60-90 days; one-time seed segments still need a touch every 180 days to stay safely inside the 540-day ceiling. Document the cadence per segment.
  2. Wire up Google Ads Data Manager (preferred, 1-2 hours). Tools → Data manager → Connect a product. Native connectors exist for Salesforce, HubSpot, Snowflake, BigQuery, and several CDPs; configure scheduled sync at the cadence above. Data Manager is now the supported path for first-party data; the legacy Google Ads API OfflineUserDataJobService Customer Match upload path is being deprecated on April 1, 2026 [7].
  3. For accounts without a connector — use the manual UI path (30 min). Open the segment → Edit → Upload new file. Use CSV with the canonical headers: Email, Phone, First Name, Last Name, Country, Zip. Include multiple identifier columns in the same row — uploads with two match keys lift match rate by ~28% on average, three keys lift it by ~35% [4][6]. Lowercase emails, strip whitespace, and use E.164 format for phone (+15555551234) before hash if you hash client-side [6].
  4. Accept Customer Data Terms once per account. Admin → Account settings → Customer data terms. Without acceptance, uploads silently fail policy validation regardless of refresh frequency [8].
  5. Set a calendar reminder per list at the cadence interval. The 540-day ceiling means even your "stable" lists need a touch — schedule it.
  6. Document who owns the refresh. In agency operations this is the single highest-frequency forgotten task. Put the segment name + owner + cadence in the account runbook.

How to confirm the fix worked

Diagnostic checklist — run all five within 7 days of the refresh

  • Every active Customer Match list shows Last modified within the cadence window you set (30-90 days for most, 180 max).
  • Match rate is reported on every list (rate only displays when ≥100 unique users matched via API upload [5]) and is stable or up vs the prior reading.
  • List size per active network is ≥1,000 active matched users where the list is used for targeting; ≥100 minimum to serve at all [6][3].
  • No REFRESH_CUSTOMER_MATCH_LIST recommendation outstanding on the account.
  • If Data Manager is wired, the last sync timestamp on the connector is within the last 24-48 hours and shows no error state.

Match rate >70% on a list with three identifier columns is healthy; 50-70% is typical and acceptable; below 50% means either decay or a formatting bug in the source export [4][6]. The fix here is rarely a one-shot — it is establishing the cadence and the calendar, and assigning the owner.

Methodology note. Whitead audits this rule together with three others because they co-occur: (a) Enhanced Conversions for Leads being off on a B2B account that already has the CRM segments to feed it, (b) Performance Max running without any audience signals attached to asset groups, and (c) new-customer-acquisition goal disabled despite a customers (all) list existing. The four findings share a single root cause — the team built Customer Match as a one-time launch task and never operationalized the refresh. Fixing the freshness rule on its own restores the targetable size, but the real audit value is the bundled fix: pair the refresh cadence with audience-signal attachment to PMax asset groups, EC-for-Leads activation on the closed-won list, and the new-customer-acquisition toggle. The combined unblock typically moves the account from "Customer Match exists but does nothing" to "Customer Match feeds Smart Bidding daily" — which is the state PMax was designed to assume. Treat the 540-day ceiling as the deadline floor, not the target; the target is the cadence appropriate to how fast that segment churns.

  • Customer Match — glossary on upload, hashing, match rate, and per-network activation thresholds.
  • Enhanced Conversions — the other first-party data surface; same Customer Data Terms gate.
  • Enhanced Conversions for Leads — closed-won uploads feed both EC-for-Leads and a closed-won Customer Match list.
  • Modeled Conversions — when first-party signals weaken, Google fills with modeled estimates; refresh keeps the real signal dominant.
  • Smart Bidding — the downstream consumer of Customer Match audience signals.

Sources

  1. Google Ads Developer Blog — Update to Customer Match membership expiration starting April 7, 2025. https://ads-developers.googleblog.com/2025/02/update-to-customer-match-membership.html (2025-02-12)
  2. PPC Land — Google sets new 540-day limit for Customer Match data retention. https://ppc.land/google-sets-new-540-day-limit-for-customer-match-data-retention/ (2025-02-12)
  3. Google Ads Help — About Customer Match. https://support.google.com/google-ads/answer/6379332 (accessed 2026-05-25)
  4. Google Ads Help — Customer Match Best Practices. https://support.google.com/google-ads/answer/10010286 (accessed 2026-05-25)
  5. Google Ads Help — About Customer Match match rates. https://support.google.com/google-ads/answer/10534785 (accessed 2026-05-25)
  6. Google Ads Help — Fix Customer Match issues with list upload, small list size, or low volume. https://support.google.com/google-ads/answer/7474166 (accessed 2026-05-25)
  7. Google Ads Help — Your guide to Customer Match. https://support.google.com/google-ads/answer/10550383 (accessed 2026-05-25)
  8. Google Ads Help — Customer Match policy. https://support.google.com/adspolicy/answer/6299717 (accessed 2026-05-25)
  9. Store Growers — 7 Strategies to Leverage Customer Match in Google Ads. https://www.storegrowers.com/google-customer-match/ (2026-01)
// was this useful?
// anonymous · no personal data stored
← Back to Wiki