Meta Ads Optimization for Advantage+ Shopping (Now Advantage+ Sales): The 2026 Guide

Legacy Advantage+ Shopping is retired, and the existing-customer budget cap did not survive. This is how to rebuild that limit and read the campaign on profit.

• What changed between ASC and Advantage+ sales
• Three ways to limit spend on past buyers
• How to size budget from your own CPA
• A weekly creative batch and a segment check

How do you optimise Advantage+ Shopping in 2026? You build Advantage+ sales instead. Meta renamed the product. Keep Advantage+ budget, audience, and placements on, replace the removed customer cap with an exclusion or a capped second ad set, give each ad set room for about 50 purchases between edits, and judge new-customer profit.

What the rename actually changed

Advantage+ Shopping, often called ASC, is no longer a separate product. Meta rebuilt it as Advantage+ sales. In Ads Manager you create a Sales campaign that runs in Advantage+ mode.

Industry coverage reported a single sales setup on test in February 2025, with a developer rollout from May 2025. API v24.0 (8 October 2025) stopped new ASC campaigns. v25.0 (18 February 2026) blocked create, duplicate, and update through the Marketing API, and from 19 May 2026 that covered every version. Reports citing Meta say leftovers pause at v26.0. Cap campaigns move only by duplicating in Ads Manager. Those v26.0 details could not be confirmed on a Meta page.

Feature	Legacy ASC	Advantage+ sales
Ad sets	One	No stated limit
Ads	Up to 150 combinations	50 per ad set
Existing-customer cap	0 to 100 percent	Removed
Audience inputs	Location only	Hard controls plus suggestions
Custom-audience exclusions	Not available	Available

Advantage+ mode stays on only when campaign budget, audience, and placements are all Advantage+. Narrowing the audience, excluding placements, or giving ad sets their own budgets switches it off. Excluding existing customers does not. Location, minimum age, and language are hard limits. Age, gender, detailed targeting, and custom audiences are suggestions. Use hard limits for rules such as the countries you ship to. Migrating a legacy ASC sends it back into learning, and the cap does not come along. Do it in a quiet fortnight, not before Diwali, Ramadan, or Black Friday–Cyber Monday, when CPMs are already rising.

Three ways to replace the cap

Define segments under Advertising Settings, then Audience Segments, before launch. Existing customers should include the CRM list plus purchasers. Engaged audience should include site visitors and social engagers. Meta’s sales campaigns page recommends this for reporting. Read Breakdown, By Demographics, Audience Segments for spend, purchases, and cost per result on New, Engaged, and Existing.

For pure acquisition, one ad set excludes the customer list. That fits when email and WhatsApp already handle retention. For a split, two ad sets share campaign budget. Cap the existing-customer ad set with a spending limit and exclude those buyers from the other. For monitoring, one ad set has no exclusion and you review segments weekly. That suits a short customer list. Stay on monitoring while existing customers take under 20 percent of spend. If they take more than 25 percent for two weeks and new-customer cost per purchase is rising, split.

A skincare brand at ₹30 lakh a month, about ₹1,00,000 a day, saw existing customers take ₹35,000 a day, or 35 percent. After two weeks above 25 percent, with new-customer CPA rising, it caps the existing-customer ad set at ₹15,000 a day and lets the prospecting ad set, which excludes the list, take ₹85,000. At ₹1,100 per new purchase, ₹85,000 buys about 77 a day. At ₹400 per repeat purchase, ₹15,000 buys about 37. Both clear about 50 events within a couple of days. Custom audiences are only suggestions, so recheck the segment split after a week.

Budget, one creative drop a week, and profit

Each ad set needs about 50 optimisation events since its last significant edit, ideally inside a week. Flat daily minimums in many guides are not in Meta’s documentation. Weekly budget is about 50 times target CPA. At ₹1,200, that is ₹60,000 a week, or about ₹8,600 a day. ₹10 lakh a month, about ₹33,000 a day, feeds two or three ad sets at that CPA, not five. Meta’s phrase is “around 50 optimization events since its last significant edit.” The week is a working approximation.

A new ad, a targeting or bid change, a large budget move, or a pause of seven days or more sends the ad set back to learning. See how to leave learning faster. Raise budget by at most 20 percent, wait three or four days, and only while profit holds. If CPA and frequency rise first, the limit is creative.

Ship new ads once a week: different hooks and formats, vertical video for Reels and Stories, static images for Feed. Pause only after spend of at least twice target CPA, and only with no purchases or a CPA above twice the ad set average. Dynamic Media is on by default for catalog ads created through the API. If the same ads fade, check creative fatigue before you cut budget.

Each week, read existing-customer share of spend and new-customer cost per purchase. In March 2026, click-through credit was limited to link clicks and other interactions moved to a one-day engage-through column, so older ROAS benchmarks do not compare. Run an A/B test or Conversion Lift before most of the budget moves. Before an edit, review 7–10 days by ad set, ad, placement, region, device, and hour.

Send value net of tax. A ₹2,360 India order with 18 percent GST is ₹2,000 of revenue. Adjust cash on delivery for return to origin, use OTP or WhatsApp checks and pincode limits, watch UPI drop-off, and aim for Event Match Quality of 6 or higher on hashed phones. Strip VAT in the UK, EU, and Gulf. Exclude US sales tax. Treat Canadian GST or HST added at checkout the same way, and strip Australian GST. Shopify nets refunds. Meta does not, unless you send them. Compare with break-even ROAS. The Conversions API guide covers duplicate events.

Ignore “set the cap to 20 percent,” “one ad set only,” and “25 conversions a week.” No Meta source turned up for that last figure. Use manual delivery for retargeting-only, strict age rules, or when no ad set can reach about 50 purchases between edits. Week zero: Pixel, net-of-tax values, segments, structure. Weeks one and two: read only. Week three: one creative batch and the existing-customer share. Week four: up to 20 percent more budget if profit holds, or a restructure above a 25 percent existing share.