About 30 percent of e-commerce ad spend in many Google Ads accounts quietly goes to re-acquiring customers who would have bought again anyway, and in 2026 the algorithm has no reason to stop unless you tell it to. Smart Bidding optimizes for conversions, and a returning customer is the cheapest, most predictable conversion there is — so left alone, it keeps buying back people you already own. The New Customer Acquisition goal exists to break that habit and push spend toward genuinely net-new buyers.
This guide explains how the goal works, the two modes you can run it in, the first-party data it depends on, how to set new-customer value from lifetime value, and how to prove it grew real revenue rather than re-tagging it. To see how much of your current spend is chasing existing customers, run our free 5-axis Google Ads audit.
Updated 2026-05-24 with current New Customer Acquisition modes, Customer Match behavior, and Performance Max acquisition reporting observed across US, UK and European accounts.
- It values first-time buyers — the goal stops Smart Bidding from spending your whole budget re-acquiring people who would have returned anyway. 2. Two modes — 'value new customers' adds bonus value to acquisitions, 'new customers only' bids for first-timers and excludes returning buyers. 3. First-party data is the prerequisite — without a fresh Customer Match list the 'new' label is guesswork. 4. Set value from LTV — anchor the bonus to lifetime value, not a round number. 5. Measure incrementality — judge it on net-new revenue from a geo holdout, not on the conversion label.
What is the New Customer Acquisition (NCA) goal?
The New Customer Acquisition goal is a setting layered on top of your existing Smart Bidding strategy that changes how the algorithm treats first-time buyers versus returning ones. By default, automated bidding sees every conversion as equal, so it gravitates toward the cheapest and most predictable conversions — and nothing is cheaper or more predictable than a customer who already loves your brand. The goal corrects that bias.
The core problem — Smart Bidding is an efficiency engine, not a growth engine. Pointed at raw conversion value, it will happily spend a large share of budget buying back loyal customers through branded search, remarketing, and Performance Max, because those convert at high rates and low cost. That looks great on a ROAS dashboard and terrible for actual growth.
What the goal changes — New Customer Acquisition introduces a distinction the algorithm did not have before: who is new and who is not. Once it can tell them apart, it can either reward acquisitions with extra value or bid only for them, depending on the mode you select. The mechanism is simple, but it only works if Google can reliably identify your existing customers.
Why 2026 matters — As third-party signals continue to erode, Google leans harder on your first-party data to make this distinction, which makes the goal both more powerful and more dependent on clean inputs. For the broader stack this sits in, see our e-commerce playbook for 2026.
How do 'value new customers' vs 'new customers only' modes differ?
The goal ships in two modes, and choosing the wrong one is the most common setup mistake. They sit on a spectrum from gentle nudge to hard exclusion, and the right pick depends on what a given campaign is actually for.
Value new customers — This is the default for most accounts. The campaign keeps bidding on everyone, returning customers included, but every time a genuinely new customer converts, Google adds an extra value on top of the order. The algorithm then leans toward acquisition without abandoning profitable repeat revenue. You keep full coverage and blended efficiency while gently tilting spend toward growth.
New customers only — This mode is aggressive. The campaign bids only when the system predicts a first-time buyer and effectively suppresses returning customers. It is the right call when a specific campaign or budget exists purely to grow the base — a prospecting Performance Max, a top-of-funnel push, a new-market launch. Expect lower total volume in exchange for cleaner, unambiguous acquisition.
How to choose — Default to value mode when the same campaign must serve both acquisition and retention and you cannot afford a volume drop. Move to new-only mode when you can ring-fence an acquisition budget and want every dollar spent on someone who is not already a customer. Many mature accounts run both: value mode on the core engine and new-only on a dedicated prospecting line. The bonus you assign in value mode is where most of the calibration work lives, which is the next section.
What first-party data does NCA require to work?
Neither mode works without a reliable answer to one question: is this person already a customer? Google builds that answer from two sources, and the quality of your data decides whether the new-customer label is trustworthy or close to random.
Observed conversion history — Out of the box, Google uses the conversions it has already recorded on your account to guess who is returning. This is better than nothing but shallow. It misses anyone who bought offline, purchased on another channel, or converted before your tracking existed — all of whom will be wrongly labeled as new and earn an undeserved bonus.
Customer Match — The fix is to upload your full existing-customer base as a hashed first-party list through Customer Match. This gives the model a far truer picture of who is already yours, including offline and pre-tracking buyers that observed history never sees. Our Customer Match and first-party data guide covers the upload, hashing, and consent details end to end.
Keep it fresh — A stale list is almost as damaging as no list. If you upload once and never refresh, last month's new buyers stay tagged as prospects, so the goal pays a bonus to acquire people it already acquired. Schedule a recurring refresh — weekly or monthly — so recent customers move into the existing-customer set promptly. Treat the list as living infrastructure, not a one-time export.
How do you set new-customer value correctly?
In value mode, the single number that matters most is the extra value you assign to a new customer. Set it too low and the goal barely changes bidding; set it too high and you overpay for acquisitions before the repeat behavior that justifies them is ever proven.
Anchor to lifetime value — The right starting point is the gap between a first order and the customer's full lifetime value. If a first order is worth 60 dollars in profit and lifetime value is 180 dollars, the incremental worth of acquiring that customer is roughly 120 dollars beyond the first sale. That gap, not the order value, is what the bonus represents. To model this properly, use our LTV calculator and our lifetime value modeling guide.
Stay conservative — Do not assign the full gap as a bonus on day one. Most disciplined advertisers set the new-customer value to a slice of it — often 30 to 50 percent — so the bid lifts toward acquisition without betting the whole future margin on a repeat rate you have not yet earned. Raise it later if cohorts prove out.
Use real cohort data — Compute the number from actual repeat behavior, not from a marketing aspiration. If your repeat rate is weak, the honest lifetime value is lower and the bonus should shrink to match. A bonus built on a hoped-for repeat rate quietly distorts every bid in the campaign and can make a losing acquisition look profitable. This is where value rules connect, covered in our conversion value rules guide.
How does NCA fit Performance Max and Shopping?
For e-commerce, New Customer Acquisition matters most inside Performance Max and Shopping, because those campaigns carry the bulk of acquisition budget and blend prospecting with remarketing in ways that hide existing-customer spend.
Performance Max — Performance Max is the classic place where the goal pays off, because its remarketing and customer-list signals naturally pull it toward existing customers unless you intervene. Turning on New Customer Acquisition — usually in value mode first — keeps the engine efficient while tilting it back toward net-new buyers. New-only mode suits a dedicated prospecting Performance Max where retention is handled elsewhere.
Shopping — Standard Shopping campaigns benefit too, especially when branded queries pull in loyal customers who would have bought regardless. The goal lets you keep harvesting that demand while paying a premium specifically for the first-time buyers mixed into the same auctions.
Keep retention separate — The cleanest structure isolates pure retention and win-back into their own campaigns, so the new-customer bonus never distorts spend you intend for existing customers. When acquisition and retention share one budget under one bonus, you lose the ability to read either clearly. A deliberate split keeps measurement honest and is the foundation for the incrementality work in the next section.
How do you measure incremental new revenue, not just labels?
The most expensive mistake with this goal is trusting the new-customer label as proof of growth. A label only says a conversion was tagged new; it does not prove that customer would have been lost without the bonus. Real measurement asks whether the goal created customers you would not otherwise have won.
The label trap — If you simply count new-customer conversions and watch the number rise, you can convince yourself the goal works when it may only be re-tagging buyers you would have acquired anyway. Labels describe; they do not prove causation. Treat the new-customer count as a diagnostic, never as the verdict.
Run a holdout — The honest test is a geo holdout or a Google-supported experiment: one matched group runs New Customer Acquisition, an equivalent group does not, and you compare net-new customer counts and revenue between them. The difference is the incremental effect — the customers and dollars the goal actually added rather than relabeled.
Watch the cohort — Track first-time-buyer rate, new-customer cost per acquisition, and cohort repeat rate over 60 to 90 days. If new customers rise in the test region without cannibalizing returning revenue, the goal is creating growth. To set a defensible acquisition target before you scale, use our LTV calculator, and to surface where spend is leaking into re-acquisition, run our free 5-axis audit.
The New Customer Acquisition decision table
Work this table top to bottom — it maps the situation you are in to the right mode, the data you need, and the first move, ordered from setup prerequisites to measurement.
Watching new-customer conversions climb feels like proof the goal is working, but the label only says a conversion was tagged new — not that you would have lost that buyer without the bonus. With a stale Customer Match list, the system can even relabel existing customers as new and pay a premium to re-acquire them. Always confirm growth with a geo holdout that compares net-new revenue against a control, and refresh your first-party list before you trust a single label.
How to prioritize your New Customer Acquisition rollout
You will rarely flip every switch at once. The mistake is skipping the data foundation, or chasing aggressive new-only mode before you can even trust the new-customer label. Sequence the rollout so each step rests on the one below it.
Data first — Upload and schedule a refresh of your Customer Match list before anything else, because every downstream decision depends on Google knowing who is already a customer. With no fresh first-party list, both modes degrade toward ordinary bidding and the bonus is wasted.
Value mode before new-only — Start in value mode so you keep coverage while the algorithm learns the new-customer signal. Calibrate the bonus from lifetime value, stay conservative, and only graduate a ring-fenced budget to new-only mode once the data is trustworthy and the acquisition economics hold.
Prove, then scale — Run an incrementality test before you pour budget in, and read it on net-new revenue and blended profit over 60 to 90 days, not on the conversion count. Size a defensible acquisition target with our LTV calculator, and to see exactly how much of your spend is currently re-acquiring existing customers, run the SteerAds free 5-axis audit.
Sources
Official sources consulted for this guide:
-
support.google.com — about new customer acquisition
-
support.google.com — about Customer Match
-
blog.google — ads and commerce updates
-
ads.google.com — Google Ads
FAQ
What is the New Customer Acquisition goal in Google Ads?
It is a campaign setting that tells Smart Bidding to treat first-time buyers differently from returning ones, so the algorithm stops spending your full budget re-acquiring people who would have bought anyway. It works in two modes. The 'value new customers' mode keeps bidding on everyone but adds an extra conversion value — often 50 to 100 percent of an order — every time a brand-new customer converts. The 'new customers only' mode bids exclusively for first-time buyers and suppresses returning ones. Both depend on Google knowing who your existing customers are, which is why first-party data is the real prerequisite.
What is the difference between 'new customers only' and 'value new customers'?
The 'value new customers' mode is the safer default: every conversion still counts, but new customers carry a bonus value, so the algorithm leans toward acquisition without abandoning profitable repeat revenue. The 'new customers only' mode is aggressive: it bids only when the system predicts a first-time buyer and effectively excludes returning customers from that campaign. Use value mode when you still want blended efficiency and full coverage; use new-only mode when a specific campaign or budget exists purely to grow the customer base and you can tolerate lower total volume for cleaner acquisition.
Does New Customer Acquisition need Customer Match?
Effectively, yes. Google identifies returning customers from two sources: its own observed conversion history on your account, and the first-party lists you upload through Customer Match. Observed history alone is shallow — it misses buyers who purchased offline, on another channel, or before tracking existed. Uploading a complete, hashed customer list through Customer Match gives the model a far truer picture of who is already yours. Without that list, the 'new' label is mostly guesswork, and both NCA modes degrade toward ordinary bidding. Refresh the list regularly so recent buyers are not mislabeled as new.
How much extra value should I give a new customer?
Anchor it to lifetime value, not to a round guess. If a typical customer's first order is worth 60 dollars in profit but their lifetime value is 180 dollars, the incremental worth of acquiring them is roughly 120 dollars beyond the first sale. Many advertisers set the new-customer bonus to a conservative slice of that gap — often 30 to 50 percent — to avoid overpaying before the repeat behavior is proven. Recompute it from real cohort data, not from a marketing target, and lower it if your repeat rate is weak. A wrong bonus quietly distorts every bid.
Is New Customer Acquisition worth it for a small catalog?
It depends almost entirely on your mix of new versus returning buyers. If 80 percent of your conversions are already first-time customers, NCA adds little — you are barely paying to re-acquire anyone, so the bonus changes few decisions. The goal earns its keep when a meaningful share of spend currently goes to existing customers who would have returned regardless, which is common for subscription, consumable, and high-repeat categories. Small catalogs with thin conversion volume also struggle, because Smart Bidding needs enough first-party and conversion data before the new-customer signal becomes reliable rather than noisy.
Can New Customer Acquisition reduce my total conversions?
Yes, and that can be the correct outcome. The 'new customers only' mode deliberately stops bidding on returning customers, so raw conversion count and even blended ROAS can fall while net-new revenue rises. That is the trade you are choosing: fewer, more valuable acquisitions instead of cheap repeat purchases the algorithm was claiming credit for. Judge it on incremental new-customer revenue and blended profit, not on the headline conversion number. If total conversions drop but new customers and margin climb, the goal is doing exactly what it should be doing.
How do I know if New Customer Acquisition actually grew new revenue?
Measure incrementality, not labels. The label tells you a conversion was tagged new; it does not prove that customer would have been lost without the bonus. Run a geo holdout or a Google-supported experiment where one group uses NCA and a matched group does not, then compare net-new customer counts and revenue between them. Watch your first-time-buyer rate, new-customer cost per acquisition, and cohort repeat rate over 60 to 90 days. If new customers rise in the test region without cannibalizing returning revenue, the goal is creating real growth rather than re-tagging it.