Google Ads new customer recommendation support arrived in Google Ads API v26.1 on 28 October 2026. The new recommendation can expose an incremental conversion value intended to help eligible campaigns prioritise new-customer acquisition.
This is not a reason to apply an automated value without review. The input influences value-based bidding, so weak lifetime-value assumptions or inaccurate customer detection can redirect real budget. This guide explains how marketing, finance and technical teams should evaluate the recommendation before accepting it.
What Google Ads API v26.1 added
Google’s official API release notes added the OPTIMIZE_FOR_NEW_CUSTOMERS recommendation type. Its recommendation resource exposes a recommended_incremental_conversion_value. Google specifies that the number uses standard units of the account’s local currency, not micros.
An API client can apply the recommendation with an account-default incremental value or a campaign-level override. The value is added to the conversion value of a customer identified as new. It therefore changes the signal used by Smart Bidding; it is not merely a reporting label.
API v26.1 is a minor release, so it does not require a new version suffix. However, a client library and internal data model must recognise the new fields before a system can read or apply them safely.
Separate the recommendation from the business decision
A recommendation is a platform-generated input, not an approved commercial forecast. Treat it as a review trigger. Confirm which account and campaigns it relates to, how new customers are defined, which purchase conversions are primary and whether the proposed value reflects current economics.
Google’s customer lifecycle goal guidance describes New Customer Value mode as a way to bid higher for new customers while continuing to reach existing customers. It is different from New Customer Only mode, which restricts bidding to customers classified as new.
Choose the mode first, then decide whether an incremental value is credible. Kayaar’s target-based bidding review helps identify conflicting goals, targets and measurement assumptions before a material bidding change.
Build an independent value range
Finance or commercial owners should calculate a defensible range before seeing the recommendation. Start with gross profit or contribution margin rather than headline revenue. Use a defined cohort and observation window, then subtract expected fulfilment, service, return and retention costs.
Input
Review question
Risk if wrong
First-order value
Is the purchase value net of discounts, cancellations and returns?
Smart Bidding overvalues low-quality sales.
Repeat contribution
How much additional margin is observed in a consistent cohort window?
Optimistic lifetime value inflates acquisition bids.
New-customer rate
Can the business reliably distinguish new from returning buyers?
Existing customers receive an unintended value uplift.
Payback limit
How much acquisition cost can cash flow and margin support?
Volume grows while profitability or liquidity declines.
Confidence range
What conservative, expected and high cases are supportable?
A single estimate hides material uncertainty.
Compare the independent range with Google’s suggested value. Investigate a large difference instead of averaging the numbers. The final setting should have a named business owner, evidence date and scheduled review.
Segment the calculation where economics differ materially. A subscription, one-off purchase, wholesale order and promotional first order may not justify the same uplift. If one account contains several business models, prefer controlled campaign overrides over a convenient account default. Record why the chosen scope is representative and which campaigns remain excluded.
Audit how new customers are detected
New-customer optimisation depends on classification accuracy. Review Customer Match lists, purchase history, tag parameters and Google’s automatic detection. Confirm list freshness, consent, geographic scope, identifiers and exclusions. Guest checkout, account merging and purchases across online and offline systems can create “unknown” customers.
Reconcile Google Ads reporting with CRM or commerce records using a stable period. Measure the proportions of new, returning and unknown customers, then sample mismatches. Kayaar’s lead-to-sale journey guide shows how to map platform events to qualified commercial outcomes.
Do not upload broader customer data merely to improve a match-rate metric. Apply documented consent, access, retention and deletion controls. Kayaar’s Data Manager readiness audit provides a governance checklist for first-party data connections.
Review conversion and bidding foundations
Confirm that purchase is a primary conversion, values use the correct currency and refunds or cancellations are handled consistently. Look for duplicate tags, imported transactions and conversion-action changes. An incremental uplift cannot repair a broken base value.
Document the current bid strategy, target ROAS, budget status, conversion delay and recent change history. Applying a new-customer value and changing the budget or target simultaneously makes the outcome difficult to explain.
Use Kayaar’s Google Ads measurement stack guide to validate tags, enhanced conversions, offline records and reporting before changing an optimisation signal.
Control the API implementation
Retrieve the recommendation and store its resource name, recommended value, currency, account, campaign scope and retrieval time. Present the information to an authorised reviewer with the independent value range and detection-quality findings.
Do not build a job that automatically applies every recommendation. Require an approval record and decide whether the value belongs at account level or as a campaign override. An account default can affect a wider portfolio than the reviewer intended.
Validate units carefully. Google specifies ordinary local-currency units rather than micros for this field. Test the request in a controlled account and confirm the saved setting. Log the response and change ID without recording credentials or raw customer data.
Measure the change without overclaiming
Capture a pre-change baseline for spend, total conversion value, new-customer conversions, new-customer acquisition cost, repeat-customer contribution and unknown-customer share. Allow for conversion delay and learning before judging results.
A higher new-customer ratio does not prove incremental growth. Check total contribution, margin, cohort retention and sales quality. Use an experiment where available and commercially appropriate; otherwise document confounding changes and avoid causal language.
Review early operational signals without making daily value changes. Frequent adjustments can destabilise learning and turn normal variance into a series of reactive decisions. Use a pre-agreed review cadence.
Set rollback conditions before launch, such as a material rise in acquisition cost, a fall in qualified revenue or an increase in unknown classifications. For an independent review of value assumptions and campaign safeguards, contact Kayaar.
FAQs
What is the Google Ads new customer recommendation?
It is an API v26.1 recommendation that can provide an incremental conversion value for optimising eligible campaigns toward new customers.
Is the recommended value expressed in micros?
No. Google states that it uses standard units of the Google Ads account’s local currency, not micros.
Should the recommendation be applied automatically?
No. Review customer detection, conversion quality, lifetime-value evidence and commercial limits before an authorised person approves the change.
What is the difference between account and campaign values?
An account default can cover the broader account, while a campaign override applies a specific incremental value to the selected campaign.
Does a higher new-customer ratio prove incremental growth?
No. Validate total contribution, acquisition cost, retention, customer-classification accuracy and other changes before claiming incremental growth.
