How should we launch a new product or offer on Meta: new campaign, existing campaign, or a pre-launch waitlist?
Our answer
Launch a new product or offer in its own time-bound campaign when its price or buyer differs and the budget can carry it. Add same-price variants to existing ad sets. Email customers first, and judge the launch on blended results.
- Confidence
- Low
- Strongest proof
- 5 Meta rules
- 17 practitioners, 1 opinion or claim
- Sources
- 12 sources
- 23 statements
MMO
CMCP
New here? Take the 1-minute tour
Why this is the answer
3 reasons and our judgmentA new offer has a 50/50 chance of working, so Common Thread Collective, an ecommerce ad agency, runs it as a time-bound marketing moment campaign, not in evergreen (Joy Sharma). Winners then graduate into evergreen.
Separate launch campaigns found buyers the funnel missed. Gut Culture's 90-day bundle did badly inside the funnel but found customers at an effective cost per purchase on its own (Connor MacDonald). Meta says a new ad set leaves learning after about 50 results a week.
Launch revenue lands on other products. Ridge targets total sales divided by total ad spend (MER) of about 2.5 times per product, yet its best launches run about 1.8 (Connor MacDonald). HexClad's launch campaign also sold 52 12-piece sets (Connor Rolain).
Our judgment
- Give a new product its own campaign only when its price, order value or buyer differs. We drew the line at a price, order value or buyer that differs from what evergreen sells, and send variants with the same economics to existing ad sets. This joins Common Thread Collective's marketing moment rule, the Gut Culture bundle, Cody Plofker's Jones Road split, Andrew Faris's economic units and Jon Loomer's budget condition.
- Fund it to about 50 results a week, or don't split it. We put that number on Jon Loomer's budget condition, using the point at which Meta says ad sets usually leave learning. Read early results loosely, since Meta says ad sets in learning are less stable and usually cost more.
- Sell to existing customers by email and SMS before launch ads go live. We applied Common Thread Collective's marketing moment sequence, which it describes for sales events, to product launches. Its forecasting model says launches can lift returning customer revenue in the days around them.
- Judge a launch on blended results and on what launch buyers bought across the range. The Ridge and HexClad launches sold other products that product-level numbers miss, and Common Thread Collective finds the new-customer effect takes time.
Our judgment, built from Common Thread Collective, Connor MacDonald, Cody Plofker, Andrew Faris, Jon Loomer, Meta and Connor Rolain (Marketing Operators podcast): our reading of these sources, not a finding from any one of them.
What the sources say
23 statements from 12 sourcesMeta says ad sets exit the learning phase after about 50 results in a week, warns that many ads and ad sets slow learning, and recommends combining similar ad sets.
A DTC Podcast speaker says launching a product on pre-order is the best way to read market demand before committing to production.
“put your product that you're going to release on pre-order. It's the best way to launch a company, get a temperature read on the market.”
Meta's Ads Super Burst lets advertisers book a 3 to 24 hour window at TV-like reach, with category exclusivity and guaranteed delivery, rolling out in phases to general availability in October 2026.
“Ads Super Burst lets you book a window of 3 to 24 hours and reach your audience at TV-like scale inside that window”
Meta says ad sets in the learning phase are less stable and usually cost more per result, so wait to edit until learning ends and avoid frequent budget changes, but do not try to avoid learning entirely.
“During the learning phase, ad sets are less stable and usually have a higher CPA.”
Meta says targeting, creative or optimization event changes and adding a new ad are always significant edits, while a budget change may or may not restart learning, depending on its size. No safe percentage is given.
“if you increase your budget from $100 to $101, that isn't likely to cause one or more ad sets to reenter the learning phase.”
An outside observer counted AG1's public Meta Ad Library: about 3,053 ads live and 2,275 launched in 30 days, with every product launch getting new pages, creators and offers. No spend or results.
In a HexClad creator launch, a marketer reports pure whitelisting worked best out of the gate, with frequency of 4.5 on some ads and a 98% new website visitor rate on one. Creators were new to the US ad account.
“This has a 98% new website visitor percentage.”
HexClad offered $30 off a complementary product as an upsell on its $130 item; about 27% of buyers took both, lifting Northbeam-reported AOV to about $217.
“it's like 27% of people are buying both of these products together”
Ridge's CMO says its target is a 2.5x MER on an individual product, with its best launches around 1.8. HexClad's launch also drove attributed sales of other products, such as 48 griddles and 52 12-piece sets.
“Our best launches are like a 1.8.”
Connor MacDonald says launching a 90-day subscription offer as a separate campaign, telling Meta you want $90 average order value customers, works better than A/B testing it against a 30-day offer inside Meta.
“If you launch it separately and you tell Meta explicitly, I am trying to acquire someone at this $90 average order value that Meta's actually far more effective at that.”
Cody Plofker says Jones Road ran kits and higher-AOV offers on value optimization and others on conversion optimization; at Winks a 90-day offer was cut after two days for lack of signal. No data.
“because we just need more signal”
Andrew Faris describes a jewelry client running one consolidated campaign budget with 12 to 15 ad sets, each a group of products with similar order value, margin and lifetime value, with its own cost cap or bid.
“a consolidated CBO”
Johnny Cox spent $1,500 to $2,000 on Meta ads to a sign-up page for about 2,000 leads in two weeks, then a three-day pre-launch sold $12,000, validating the product before stocking it.
Common Thread Collective runs two campaign types: always-on evergreen campaigns split between acquisition and retention, and time-bound campaigns for sales and launches with set start and end dates.
“Marketing moment campaigns are time-bound, attached to sales, product launches, or any other event”
CTC splits accounts into evergreen campaigns (acquisition and retention) and marketing moment campaigns. Anything time-bound, such as a sale or product launch, goes in a marketing moment campaign.
“Anything that is time-bound starts here, ends here, goes in marketing moment.”
Jon Loomer says start simple and split campaigns only for separate goals, distinct products or locations, or seasonal items, and only if each piece can still reach about 50 optimized events a week.
“Simplicity is the starting point, not an absolute rule.”
Common Thread Collective models how launches, emails and promotions move new and returning customer revenue for each brand. In its example, a launch lifts returning customer revenue fast while new customers take longer.
“Every brand has a different relationship of marketing event types”
CTC's Joy Sharma tests a new offer in a marketing moment campaign against set order value and CAC targets, not in evergreen, and moves it to evergreen only if it wins. CTC sells this service.
“I'm fine running it in the marketing moment and if it wins, I'll graduate into evergreen.”
CTC's Joy Sharma says once a brand has offer-market fit, static images win: launch a high-value offer with statics only, then move winners into evergreen campaigns. CTC sells this service.
Skullcandy's Evan Catlett describes an embargoed tease, launch and sustain plan for its April 2025 earbuds launch, giving creators product early under NDA so content was ready at launch. No data.
“so that we would be really poised with a strong content pipeline”
Flighted, a Meta agency, says new campaigns without cost history should start on Highest Volume, then add cost or ROAS goals set from the trailing 30 days, not aspirational targets. No data shown.
“For new campaigns without baseline data, Highest Volume is typically the right starting point”
CTC says before a sale, email existing and VIP customers early access 48 hours before ads, email and text again at 24 hours, then run ads to prospecting only for 2 days. No data.
“launching ads before you've extracted value from your existing customer base”
Meta's docs say reminder ads need an Instagram upcoming event and stop delivering once the event is past. The page lists supported objective and optimization pairs and reports no results.
“A reminder ad will stop delivering once its event is in the past”
In the full answer
For readers who sign in- What to do this week3 steps
- The numbers behind it3 figures from cited sources
- Where each source standsA chart placing each source on the question
- The strongest disagreementThe best case against the answer, from a cited source
- How sure we areWhat its confidence rests on, and what would change the answer