Two ad platforms spent October building the same feature. Both need a fact about your business that most businesses cannot produce.
01 — What happened
In the first week of October, at Advertising Week New York, Meta made Customer Lifecycle Strategy available to all advertisers. Common Thread Collective dates the announcement to 7 October; other trade coverage puts it a day earlier. It had been in closed beta since the spring.
It is a setting at the top of a Sales ad set: reach new and existing customers — the default — or “Acquire new customers”. Meta’s own interface notes that the acquisition option may increase your cost per result.
Then, around 13 October, it changed in a way that matters more than the rollout. Choosing “Acquire new customers” now applies an exclusion labelled “Automatically created: Customers from your conversion data (730 days)”. According to Jon Loomer’s testing, that exclusion cannot be switched off, although the window can be shortened to 180, 90, 60 or 30 days. Until then, exclusions had been entirely manual.
Meta decides who is already a customer from your pixel data, your Conversions API feed and any customer lists you upload. Common Thread Collective reports that it treats someone as new if there is no purchase event from your brand in the previous 180 days.
On 22 October, Google announced three changes to customer lifecycle goals:
- High Value Mode — already available in Search and Performance Max — extended to Shopping campaigns.
- Customer retention goals coming out of beta for Performance Max, Search and Shopping, with Re-engagement, Loyalty and Retention-only modes. Search Engine Land reports settings including “Bid higher for lapsed customers (Recommended)”.
- Percentage multipliers rolling out of beta, from 1.1x to 5x, across Search, Shopping, Performance Max and Demand Gen in both Google Ads and SA360.
Google’s documentation is blunt about the input it needs: “Define your existing customers through first-party data when enabling customer acquisition in Google Ads.” Lapsed customers are whatever you say they are — you define the window and upload the list through Customer Match. PPC Land reports the multiplier interface requires a current-customer segment with 1,000 or more active members.
What is genuinely new is not a targeting option. It is that customer status has become a bidding input on both platforms, in one month.
02 — Why it matters
Both platforms now ask a question most businesses have never answered formally: who is already a customer, and when did they last buy? That used to be internal bookkeeping. It now sets what you pay at auction.
And if you do not answer, the platform answers for you. Meta’s automatic exclusion is derived from the purchase events that reached that ad account, and it cannot be turned off. For an online retailer whose every sale passes through one checkout, that inference is roughly right.
For everyone else it is structurally incomplete. A sale closed on the phone, agreed on a quote, invoiced monthly or taken in person never produced a purchase event on your pixel. To the platform that customer is a stranger, and will be bought again at acquisition prices by a campaign configured to find people who have never bought from you.
The error runs one way. Customers your records can prove are excluded; customers they cannot prove are targeted as prospects. That gap is the distance between your real customer base and the slice of it visible inside an ad account.
This is the mechanism behind the retirement of keyword control: the platform takes the decision, and the data you feed it becomes the only lever left. The difference is that conversion signals shape optimisation over weeks. Customer status decides, at each auction, whether you bid.
03 — The economics
The number that changes is cost per acquisition — which most businesses have never really measured.
Take an illustrative business with 200 purchases a month attributed to paid ads, 70 of them from people who have bought before. The reported cost per purchase divides total spend by 200. The true cost of a new customer divides the same spend by 130 — roughly 54% higher. Nothing has changed about the business; the denominator was wrong.
Meta’s warning that the acquisition option “may increase your cost per result” is therefore accurate but misleading. The cost does not increase. It stops being averaged down by purchases you were going to get anyway. The comfortable blended figure was partly a retention number wearing an acquisition label — the same error as quoting retention statistics from someone else’s business instead of measuring your own.
Google’s multipliers push harder. A range of 1.1x to 5x is not a setting; it is a demand that you state how much more a new customer is worth than a repeat one. That ratio depends on repurchase rate, margin and customer lifespan — figures in your records, not in an ad account. Pick 5x because bigger sounds more ambitious and you have repriced every auction on a guess.
Then there is the cost of the incomplete list. Suppose the same business can prove, from pixel data alone, only 60% of its existing customers. Acquisition-only mode excludes that 60% and leaves the other 40% in the prospecting pool. You have narrowed the audience — Meta’s cost-per-result warning applies in full — while still paying acquisition prices for people already on your books. Worse economics than the blended set-up, behind a figure that now looks clean.
The 1,000-member threshold has a quieter implication: below it, the retention-side features are unavailable. For a business that has taken orders for years without consolidating who placed them, the first cost is not media. It is assembling a list that should already exist.
04 — What businesses should do
If every sale runs through one online checkout. Treat this as a settings change, and do it. Check purchase event match rates in Events Manager first, because the exclusion is only as good as the match. Then re-baseline reporting: compare new-customer cost to new-customer cost, never to last quarter’s blended figure. The number will look worse. It is not worse.
If any sales close off-platform. This is most service and trade businesses, wholesalers and anyone who quotes. If your business takes 500 enquiries a month and closes 60 of them by phone or against a quote, the ad platforms know about none of those 60. Do not enable acquisition-only mode yet: it narrows your audience against a definition of “existing customer” that is missing most of your existing customers.
Decide the ratio before you touch the multiplier. Work out what a new customer is worth relative to a repeat one, from your own repurchase and margin figures. If you cannot calculate it, the correct multiplier is the lowest, not the highest.
Set the retention window to your repurchase cycle. Meta defaults to 730 days. A business whose customers reorder every three months, running a 730-day exclusion, has excluded its warmest buyers from the campaign most likely to reach them.
If you cannot produce a customer list at all, start there. Not with the ad settings. The list is now a bidding asset.
05 — Where technology fits
No AI system is required here, and no one should sell you one for this.
For a single-channel online business with a working pixel and fewer than a thousand customers, the honest answer is nothing at all. Check your match rates, choose a sensible window, move on.
Where sales close away from the website, the answer is narrow and unglamorous: one integration. The record of who bought and when needs to flow from wherever the sale is recorded — invoicing, booking system, point of sale, the CRM — into the ad platforms as a customer list or an offline conversion import. That is plumbing, not intelligence, and should not be scoped as though it were.
The prerequisite is the awkward part: an integration needs a single place that knows who your customers are. In most businesses it does not exist — buyers are split across an invoicing tool, a spreadsheet, a booking calendar and somebody’s inbox — exactly the cost of disconnected systems appearing as a new line item.
A CRM earns its place here only if it becomes the authoritative answer to “is this person a customer, and when did they last buy?” If it is where notes go, it will not help. Consolidation first, connection second.
06 — RevenueStack’s take
Our view: the development is not the setting. It is that two platforms have started pricing on a fact they can only get from you — and rather than wait, they chose a default and billed you on it.
Meta’s automatic exclusion cannot be switched off. That is the whole pattern in one control. When a platform needs a fact about your business and you have not supplied it, it does not pause — it infers one and leaves you the cost of it being wrong. Conversion signals, then keyword control, now customer status. The defensible position each time is not better campaign management. It is being the authoritative source on your customers.
The more useful point: this is a gift disguised as a configuration change. Most businesses have never split acquisition from retention in their advertising numbers, and have been quietly pleased with a blended figure that flattered them. Both platforms now force the split. The number will be higher and less comfortable, and for many businesses it will be the first true cost of a new customer they have ever seen.
Take the number seriously before you take the feature seriously. A business that knows what a new customer costs can decide whether to buy more of them. A business working from a blended average is only spending confidently.
Sources
- Google Ads expands customer acquisition and retention controls — Search Engine Land, 2026-10-22
- Google Ads Updates To Customer Lifecycle Goals For Holidays — Search Engine Roundtable, 2026-10-22
- Google Ads lets retailers multiply new customer values up to 5x — PPC Land, 2026-10-22
- About customer lifecycle goals — Google Ads Help
- Meta Adds Automatic Customer Exclusions — Jon Loomer Digital, 2026-10-13
- Meta's Customer Lifecycle Strategy Is Now Live for All Advertisers — Common Thread Collective, 2026-10-07
- Customer Lifecycle Strategy, Attribution Updates, and More — Jon Loomer Digital, 2026-04-02
- Customer Lifecycle Strategy Is Not What Advertisers Thought — Jon Loomer Digital, 2026-06-04
