Google Ads New Customer Acquisition Goal: When Bidding for New Buyers Pays Off, and How It Can Inflate Your ROAS

Look at the conversions column in a Google Ads shopping or Performance Max campaign and you are looking at a mix of two very different events. Some of those purchases came from people who have never bought from you. Others came from people who already had your email in their inbox, your app on their phone, and a basket they were going to fill anyway. The column adds them together and calls the total “conversions”. Smart Bidding, by default, treats them as equally valuable.

For a lot of accounts that is fine. For others it is the reason growth has stalled while the dashboard looks healthy: the campaign is efficiently harvesting people who were already yours, and the cost of finding new ones is buried underneath. Google has a setting aimed squarely at this problem, the new customer acquisition goal. It is also a setting that is easy to switch on, hard to evaluate, and quietly capable of making your ROAS look better than reality.

This post walks through what the goal does, the two modes it offers, what Google needs from you before it works, where it fits and where it does not, and how to tell afterwards whether it did anything useful. I have leaned on Google’s own help documentation and a handful of practitioner write-ups, and where those sources disagree or where I could not confirm something, I say so. Google changes this area often, so treat the specifics as things to verify in your own account before you act.

The problem the goal is trying to solve

Smart Bidding optimises toward the conversion data you give it. If every purchase carries the same value, the system has no reason to prefer a first-time buyer over a loyal one. In practice it will often favour the loyal one, because returning customers are easier to convert: they know the brand, they trust the checkout, and they often search for you by name. A campaign can post a lovely ROAS largely by showing up for people who would have found you anyway.

That is not a Google conspiracy. It is just what optimisation toward an undifferentiated target looks like. The system does what you asked, and what you asked was “more conversions at this cost” or “more conversion value at this return”. It was never told that a first purchase from a new person is worth more to the business than a fifth purchase from an existing one.

If you have ever read a brand-versus-non-brand analysis and felt uneasy about how much of your performance came from people already looking for you, this is the same instinct applied to the whole account. We covered the brand side of it in our post on bidding on your own brand name. The customer acquisition goal extends the question beyond keywords: it is about who the person is, not what they typed.

What the new customer acquisition goal actually is

Google groups this feature under what it calls customer lifecycle goals. According to Google’s help page on customer lifecycle goals, the new customer acquisition goal lets you tell Performance Max or Search campaigns to put extra weight on people who are new to your business. Some Google pages also list Demand Gen and Shopping in the broader lifecycle family, and different help articles give slightly different lists of eligible campaign types. That inconsistency is worth noting. Before you plan around it, open the Goals section of your own account and see which campaign types actually offer the setting.

The mechanism is straightforward. You tell Google who counts as an existing customer, usually by uploading customer lists. Anyone not on those lists is treated as new. Google also says it can detect new customers on its own from your tracked purchase activity, so autodetection runs alongside your lists when purchase conversions are being tracked. Then you pick how aggressively the bidding should favour the new group.

Two things are worth stating plainly at the start. First, the goal changes how the bidding model values different people. It does not change your targeting in the way an audience exclusion would. Second, the quality of the whole thing depends on how accurately Google can separate new from existing, which depends on the data you hand over. If that sounds like a theme, it is: nearly every problem with this goal traces back to the customer list.

The two modes, and why the difference matters

The goal has two modes, and choosing between them is the most consequential decision in the setup.

Bid higher for new customers. Google’s documentation calls this the new customer value mode. Smart Bidding still tries to win every kind of conversion, but it assigns extra value to a new customer’s conversion and bids more aggressively for those people. It requires a purchase conversion goal and value-based bidding, because the whole thing works by adding a number on top of the conversion value. Google’s own guidance recommends this mode for most advertisers, and the logic is easy to see: you keep serving existing customers where it is cheap to do so, and you lean toward new ones.

Bid only for new customers. In this mode, existing customers are effectively worth nothing to the bidding model. One practitioner summary describes it as assigning zero value to repeat purchases by existing customers. It is not a hard block. Third-party guides are clear that this mode does not guarantee you will never show an ad to an existing customer, because the system works from predictions and list matching, and neither is perfect. What it does is stop treating returning customers as something to pay for.

Google recommends the exclusive mode narrowly: when you have a dedicated acquisition budget, or when you are running a campaign built around a non-purchase conversion goal such as lead generation. A practitioner write-up from Practical Ecommerce warns that it limits traffic, which makes sense. You have narrowed the pool of people the bidding cares about, so there are fewer auctions worth winning.

There is a related wrinkle for stores. In Google’s documentation on using the goal with store goals, Performance Max campaigns running store goals are compatible only with the new-customers-only setting. And in that mode, offline store sales are not reported as new customers, per the same page. If you run a physical retail operation and expect to see in-store new customer counts, check this before you build a plan around it.

The number that does the work: customer acquisition value

In the bid-higher mode, you set a customer acquisition value. This is the extra amount Google adds to the conversion value of a new customer’s first purchase. It is not revenue. Nothing is paid out and nothing shows up in your payments. It is an instruction to the bidding model: a first purchase from a new person is worth this much more to us than the purchase itself.

Google suggests a way to think about it. Set the acquisition value so that it, plus the value of a typical first purchase, approximates the lifetime revenue of a typical new customer. Google’s help material walks through an example built on average spend per purchase, the share of customers who buy again, and the number of years they keep buying. If you are unsure what to use, Google offers a recommended default value in the interface.

Here is a rough way to reason about it with made-up numbers, purely for illustration. Suppose a new customer’s first order averages 60, and across the customers you acquired in a past period, the average total spend over the following years came to 180. The gap, 120, is what the acquisition value is trying to represent. You might choose something conservative like 60 and watch what happens, rather than loading it with the full gap and betting everything on a lifetime value estimate you have not stress-tested. These numbers are invented to show the arithmetic and say nothing about what you should expect.

Where do you get lifetime figures you trust? Not from a forecast, ideally. Pull them from your own order history: group customers by the month of their first order and look at what each group spent over the following months. If your business is young, you will have too little history, and the honest answer is a conservative acquisition value that you adjust once data accumulates. An inflated acquisition value does real damage, and the next section explains how.

Why it can inflate your ROAS (and what to do about it)

This is the part that gets missed. When you add an acquisition value, Google reports that added amount as part of conversion value in the platform. Search Engine Land makes the point in its coverage of the goal: the extra value sits on top of the real sale value, which can artificially inflate ROAS if you bid on a return-on-ad-spend target. If you bid on a cost-per-conversion basis, the discrepancy matters less.

Think through what that means. Say your campaign has a target ROAS of 400 percent and your real sales are producing 300 percent. After enabling the goal with a generous acquisition value, the platform’s reported ROAS climbs, because new customer conversions now carry bonus value that never reached your bank account. The campaign looks like it is hitting target. Meanwhile, if you are using that target to decide how much to spend, the system is also reading its own inflated numbers, which can push spending beyond what the real revenue supports.

There are a few sensible ways to protect yourself.

• Judge the campaign on your own data, not on the in-platform ROAS. Compare spend to actual revenue from your store or CRM. If the two diverge sharply after you turn the goal on, the acquisition value is probably doing the talking.
• Re-think the ROAS target. If the reported value includes a bonus, a target set against the old, plain value will mean something different. Expect to revisit it, and expect that to restart some of the calibration. Our explainer on the Smart Bidding learning phase covers why changes like this are not free.
• Start with a modest acquisition value. It is easier to raise it after a few weeks of clean results than to unwind a campaign that learned to chase an inflated number.
• Consider whether a cost-based strategy fits better. If you are unsure about the interaction, a cost-per-conversion approach is less exposed to the inflation effect, according to the coverage cited above. The trade-offs between those strategies are in our comparison of Target CPA, Target ROAS and Maximize Conversions.

One more practical note from Practical Ecommerce: for reporting purposes it is worth giving new customers some nominal value even if you do not want to affect bidding, something as small as a cent, so that the reporting columns populate. Whether that works smoothly in your account is worth checking rather than assuming.

What Google needs from you before it works

Switching the goal on is a few clicks. Getting it to behave is about the inputs.

A purchase conversion action. Google’s setup page says a purchase conversion action is required to optimise for new customers. A summary of that page also notes that only conversion actions set up through the manual implementation method are eligible, though I would confirm that against the current page since I am working from a secondary summary of it. If your purchase tracking is patchy, fix that first. Our guide on auditing conversion tracking is a reasonable starting point, and enhanced conversions matter here too, since better matching of purchases to people helps the new-versus-existing call.

Customer lists that are big enough, recent and shared properly. Google’s lifecycle goals page says a customer list needs at least 1,000 active members in at least one network to be used. Some third-party guides describe other thresholds, and one cites 100 as a floor, so the sources disagree; Google’s own page is the one to trust for eligibility. Lists must be shared with the account that runs the campaigns, and you can include up to five lists. Anyone not found on your lists counts as a new customer, which means a stale or thin list will mislabel plenty of loyal buyers as new.

Good match rates. Google’s troubleshooting guide for the goal stresses that accuracy depends on Customer Match data and recommends checking match rates. If your upload has many malformed emails, no phone numbers, or hashed values that do not line up, Google cannot recognise the people you thought you told it about. Our write-up on Customer Match covers list building in more depth, including what changed in 2026.

A sensible lookback. Google says it can build a new-customer audience from up to 540 days of campaign activity and tracked purchases when purchase tracking is in place. That is a long window, which is good for a business with infrequent purchases and less obviously good for one where people buy monthly. It is also a reminder that “new” in Google’s model means new relative to what Google has seen, not new to the human universe.

Patience. Google’s troubleshooting page says Performance Max and Search campaigns need one to two weeks to adjust after significant changes. That fits what most people see with Smart Bidding generally. The temptation to flip a setting on Monday and judge it on Thursday is strong. Resist it.

Where it fits, and where it does not

The goal makes the most sense in a particular set of situations. It makes much less sense in others.

It tends to fit when:

• You are an e-commerce or repeat-purchase business with a meaningful customer file, so the list requirements are easy to meet and the new/existing distinction is real.
• Your repeat customers buy mostly on their own, so paying a premium to reach them is largely wasted. A cheaper way to retain them (email, loyalty, remarketing) may exist, and you want ad spend focused elsewhere.
• Your growth plan is explicitly about acquisition, and you are willing to accept a higher blended cost per conversion today in exchange for more first-time buyers.
• You can estimate lifetime value from history, so the acquisition value is grounded in something.

It tends to fit poorly when:

• Your customer list is tiny or your business is new. Without a list of real size, Google treats nearly everyone as new, and the goal has nothing to separate.
• Your conversion volume is low. Smart Bidding needs enough data to do anything sensible even before you add a layer of valuation. Our post on how much conversion data Smart Bidding needs is relevant, because piling an acquisition value on top of too few conversions mostly adds noise.
• Most of your revenue is a single lead or sale with no repeat component, so there is no meaningful “existing customer” to distinguish.
• Your offline sales happen in stores and you need them counted as new customers, given the limitation above for store goals.

Lead generation sits in an awkward middle. Google points to the new-customers-only mode as a fit for non-purchase campaigns such as lead gen, but the setup page also says a purchase conversion action is required for the goal in general. Those two statements sit uneasily together, and I could not reconcile them from the sources I found. If you are a lead-gen account, test it on a small campaign and check what the interface lets you do before you commit. You will also need your CRM outcomes flowing back into Google to make any of it meaningful, which is the territory of offline conversion imports. A lead-gen account that cannot yet tell good leads from junk has bigger problems first; our guide to spam and junk leads covers those.

Performance Max, Search, and who gets the credit

Since the goal works in both Performance Max and Search, you need to think about how those two campaign types interact. If both can match the same query, the new customer settings on each can pull the auction in different directions. Our piece on Performance Max versus Search explains which campaign generally wins an overlapping query, and that matters because a Performance Max campaign set to bid only for new customers will not be the one serving a loyal customer if your Search campaign takes the auction first.

A tidy way to use the feature is to give each campaign a clear job. A Performance Max campaign with the bid-higher mode is meant to find and win new people at scale. A branded Search campaign is left alone to catch people looking for you by name, because you are not trying to stop those clicks, merely to avoid paying a premium to chase them elsewhere. Troubleshooting guidance from Google notes that if a Performance Max campaign serves less than expected after you enable the new-customer-only setting, you should confirm both the “new customers only” setting and the “optimize campaign for acquiring new customers” box are enabled. Two switches that need to agree is the sort of thing that goes wrong quietly.

If your account structure is already tangled, adding this goal makes things harder to read. The simpler structure we describe in our post on how many campaigns you actually need makes it much easier to know what a given setting is doing.

Measuring whether it worked

The trickiest part is knowing, a month later, whether the goal helped or whether you just changed what the dashboard shows. A few approaches help.

Use the new customer reporting. Google adds reporting for this: you can segment conversions by new versus returning customers and add a new customer lifetime value column. Google’s own example is a campaign with five new-customer conversions at an acquisition value of 25 each, which shows 125 in the new customer lifetime value column. It is a good check that the settings you think you applied are the ones in effect.

Calculate acquisition cost the plain way. Google suggests customer acquisition cost as cost divided by the number of new customers. That is a number you can compare against what a new customer is worth to you, from your own order history, with no in-platform bonus value in the way. If CAC goes down and total new customers goes up, the goal is probably earning its place. If new customer counts rise while CAC rises faster than lifetime value justifies, you may have set the acquisition value too high.

Watch total revenue and total spend, not just the campaign. The sneaky failure is that new-customer bidding shifts money away from returning customers who would have converted cheaply, and that revenue does not magically reappear from new buyers. Compare the whole account’s blended revenue and blended spend over a window before and after, not merely the campaign the setting lives in. If you use email or direct channels heavily, check whether they picked up the returning customers you stopped paying to reach.

Expect noise, and respect conversion lag. The last couple of weeks of any report undercount conversions that have not landed yet. We covered that in our post on conversion lag. It matters doubly here, since you are already waiting one to two weeks for the campaign to adjust.

Run it as a proper test if you can. The cleanest way to know whether the goal helped is a split test rather than a before-and-after comparison, because seasonality, promotions and stock levels all move numbers independently of your settings. Google Ads has a campaign experiments tool built for this, and our guide to running experiments without fooling yourself explains how to set one up so the result means something.

One caveat on sources. A single agency blog post dated August 2026 describes a reporting-only “report on new customers acquired” option that adds new customer columns without changing bids. I could not find that in Google’s official documentation, so I would treat it as unconfirmed. If you see an option like it in your account, it may be a way to get the measurement without touching bidding, which is worth trying first if you are nervous about the ROAS effects discussed above.

A sensible rollout, step by step

If after all this you want to try it, here is a measured way to go about it.

1. Clean up the foundations. Confirm purchase tracking is accurate, conversion values are right, and your tracking audit comes back clean. Do not stack a new valuation layer on numbers you do not trust.
2. Build and check your customer list. Export customers with consistent, complete contact details. Aim well above the 1,000-member minimum if you can. Upload it, wait for it to process, and look at the match rate. If matching is poor, fix the data before going further.
3. Pick one campaign. Choose a campaign with enough volume that results will be readable in a few weeks, and one where you could live with a rough patch. Avoid starting with your best performer.
4. Start in the bid-higher mode with a modest acquisition value. Hold the exclusive mode in reserve for a campaign that is explicitly an acquisition campaign with its own budget.
5. Record your starting point. Note spend, real revenue, CAC, new-customer counts and ROAS before you change anything, ideally from your own store data as well as the ad platform.
6. Leave it alone for a couple of weeks. No tweaking the bid target, budget or creative in the meantime if you can avoid it.
7. Judge on real revenue and blended results. Compare against your starting point and, if you ran an experiment, against the control arm.
8. Adjust slowly. Raise or lower the acquisition value in small steps, and let each step settle before the next.

The monitoring problem nobody mentions

Setups like this fail in boring ways, not dramatic ones. A customer list stops refreshing and slowly goes stale. A purchase tag breaks after a site update and the model loses its signal. Someone changes a ROAS target without realising the reported value now includes a bonus. A campaign’s spend creeps up for a week before anyone looks. None of these announce themselves, and a weekly glance at the dashboard might catch one of them two weeks late.

This is the kind of thing continuous daily account monitoring is for. Growera’s Google Ads manager monitors your account continuously and reviews it daily, so the kind of drift described above, like a conversion signal dropping or spend moving away from where it should be, is more likely to get noticed while it is still small. It does not decide whether the customer acquisition goal is right for your business. That judgment depends on your margins and your customer file, and you are better placed to make it. What it can do is make sure you notice when something underneath the goal has stopped working. If you want to see how that fits your account size, the pricing page lays out the plans, and the e-commerce page describes how it applies to stores, where this goal is most relevant.

Common mistakes to avoid

Treating the acquisition value as a guess you can inflate. The bigger the number, the harder the system chases new customers and the more it distorts reported ROAS. Tie it to evidence from your own history.

Forgetting the list needs maintenance. A list uploaded once and never refreshed is a list that slowly tells Google your newest loyal customers are strangers.

Reading the platform’s ROAS as the truth. After enabling the goal, in-platform ROAS includes added value. Reconcile against revenue you can actually see.

Choosing the exclusive mode out of enthusiasm. It sounds like the pure version of what you want. For most advertisers, Google’s own advice points toward the bid-higher mode, and the exclusive mode limits traffic.

Changing five things at once. If you switch the goal on and also alter the budget, the target and the creative, you will never know which change caused what.

Applying it where it does not fit. A small account with few conversions and a tiny customer file will not benefit from a layer of valuation that depends on volume and data it does not have.

Frequently asked questions

Does the goal stop my ads showing to existing customers? Not entirely. Even the new-customers-only mode works through prediction and list matching, and third-party guides are explicit that it does not guarantee zero impressions for existing customers. It changes what the bidding model is willing to pay for them.

Do I need a customer list? Google can detect some new customers from your tracked purchases, but it recommends supplying lists, and its troubleshooting guidance ties accuracy to Customer Match data and match rates. In practice, supply one.

How big does the list need to be? Google’s lifecycle goals page states a minimum of 1,000 active members in at least one network. Some third-party sources quote other numbers, so go with Google’s.

Which bidding strategies work with it? The bid-higher mode needs value-based bidding and a purchase conversion goal. The new-customers-only mode works with the main Smart Bidding strategies according to Google’s documentation, with a purchase goal recommended for accurate reporting.

How long before I see results? Google says Performance Max and Search campaigns need one to two weeks to adjust to significant changes. Add time for conversion lag before you judge.

Is it worth it for lead generation? Possibly, but the sources conflict about whether a non-purchase setup can use it, so test on a small campaign and check what your account allows.

The short version

The new customer acquisition goal exists because a plain conversion count cannot tell a new buyer from a returning one, and Smart Bidding left alone will happily favour whoever is cheapest to convert. The goal gives you a way to say that new customers matter more, either by bidding higher for them or by bidding only for them.

It earns its place when you have a real customer file, repeat purchasing, a clear acquisition target and lifetime value figures you can defend. It works poorly on thin data. Its main hazard is that the acquisition value is added to reported conversion value, which can make ROAS look better than your bank balance does, so judge results by real revenue and your own cost per new customer, not by the in-platform return.

If you try it, fix tracking first, check the match rate on your list, start in the bid-higher mode with a conservative value, change one thing at a time, and give it a couple of weeks before deciding. Then look at the whole account rather than the single campaign. And since the setup depends on lists, tags and targets that can drift without anyone noticing, keep something watching them. If you want the broader picture of where this sits in a daily workflow, the Growera homepage has an overview.