TL;DR: For six months, Meta advertisers have been running Andromeda with two structural gaps. There was no way to validate a new concept fast enough for a 2-3 week fatigue window, and no way to close the lead-to-appointment latency the funnel opened. Last week Meta shipped Push Delivery to This Ad and made Advantage+ Leads globally available with a Lead Capture AI Agent and Book Time inside Instant Forms. The gaps are closed. The operating model is finally coherent. If your account is still running the pre-Andromeda structure, you are working against the platform, not with it.
At Third Pillar Marketing Group, we work with growth-focused founders across health-tech, consumer hardware, spirits, and DTC. Meta is a live channel for most of them, and every account has been navigating the same tension since Andromeda's full rollout. The math on creative volume, ad-set structure, and lead capture kept getting more demanding, and two of the operational pieces you needed to actually run the recommended play were missing. Last week both pieces shipped.
The consensus Andromeda playbook, refreshed this week by Common Thread Co., The Optimizer, and Segwise independently, converges on the same structure. One ad set, 20 to 30 conceptually distinct creatives, tested at the concept level rather than the variant level. Brands doing this reported roughly 65 percent higher ROAS than brands testing under 10 new creatives a month.
That was the target. The problem was the mechanics.
Gap one: when you added a new creative to a live, high-performing ad set, Meta's delivery system starved the challenger and kept spending on the incumbent winner. The old workaround was to duplicate the ad set to force the spend, which reset the learning phase and blew up your reporting. So most operators ran fewer new concepts than the mandate required. The playbook said 25 creatives. The reality shipped 8 to 12.
Gap two: parents, patients, and other high-consideration buyers converted best when they could hold a discovery call, book a demo, or schedule a fitting inside the moment of intent. Instant Forms captured the lead, but the calendar hold happened 24 to 48 hours later, off-platform, after email nurture. That latency cost you conversions, especially in categories where the buyer researches at 11pm and forgets by 9am.
Push Delivery to This Ad now lives under Show More Settings when you add a new ad to an existing ad set. You configure a percentage of daily budget, up to 7 days, that Meta must direct to the new creative. It does not reset the ad set's learning phase. It does not require a duplicate. It is a manual delivery override, and it is the first genuine one Meta has shipped in years.
Advantage+ Leads went globally available the same week. Two features inside it are the real story. The Lead Capture AI Agent responds instantly on Messenger to any lead who submits a form, so the response time on a warm inquiry drops from hours to seconds. Book Time surfaces your live calendar availability inside the Instant Form itself, so a parent researching your service at 11pm can hold a discovery call slot before they close the tab. Meta reported 60 percent lower CPL and 125 percent higher lead volume for dual-form campaigns using Book Time versus website-form-only baselines.
You do not have to trust Meta's numbers. What you should trust is that the mechanic exists. The gap between intent and calendar hold just collapsed to zero on-platform.
Meta advertisers now run one ad set with 20 to 30 conceptually distinct creatives, use Push Delivery to force spend on every new concept for 3 to 5 days, and use Advantage+ Leads with Book Time to convert intent into a held calendar slot without leaving the platform. If your account structure does not look like that by end of Q3, you are running a version of Meta that no longer exists.
We are running the same four-move play with our clients through Q3.
First, consolidate ad-set structure. If you have five ad sets running variants of the same concept, collapse them. Meta's Creative Similarity Score above 60 percent already treats them as one entity for competition purposes. You are paying for the illusion of structure. Move to one ad set per intent segment with a rotating library of 20 to 30 distinct concepts.
Second, build the concept library. Not variant headlines. Not cropped image variants. Actual distinct angles: fear-relief, cost objection, waitlist frustration, sibling story, credential proof, comparison to the obvious alternative, use-case-in-context, price transparency. Aim for 20 to 25 concepts by end of quarter. Every concept starts as a hypothesis about a specific buyer moment.
Third, adopt Push Delivery as the default new-concept validation cadence. When a concept enters a live ad set, Push Delivery forces 25 percent of daily budget to it for 5 days. If Hook Rate hits 35 percent and Hold Rate hits 30 percent, keep it. If it does not, kill it. You now have a working validation cycle inside the platform for the first time.
Fourth, run a controlled Advantage+ Leads test with Book Time. Do not migrate your entire lead-goal budget. Set up a parallel Advantage+ Leads campaign with a dual-form structure (Instant Form with Book Time, and a website form for buyers who want the longer path). Match your best current lead-goal ad set for spend. Run it 14 days. If CPL drops and calendar holds convert to actual booked meetings at the same or higher rate, it is your new default.
Want to apply this to your Meta account? Kick off with this prompt in Claude. It turns the four moves above into a working audit you can run in about an hour.
You are my fractional CMO. I need help auditing my Meta Ads account
against a shift that happened last week: Push Delivery to This Ad
shipped, and Advantage+ Leads went globally available with Book Time
inside Instant Forms. Meta's Andromeda operating model finally has
both structural gaps closed.
Context I'll give you:
- Business: [describe your business, category, target buyer, primary offer]
- Monthly Meta Ads spend: [amount]
- Campaign types currently running: [Sales / Leads / Advantage+ / etc.]
- Main conversion event: [lead / booked call / purchase / demo]
- Current ad-set structure: [# ad sets and rough creatives per set]
- Rough size of active creative library: [# distinct concepts, not variants]
- Current CPL or CAC: [number]
- Do you use Instant Forms today, and if yes, how many fields:
[yes/no + count]
Run this four-move audit and give me a concrete action list, not a
summary. Ask me one clarifying question at a time when you need more info.
1. AD SET CONSOLIDATION AUDIT.
Identify ad sets I should collapse. Flag any pair or trio of ad sets
running variants of the same concept. Recommend the new consolidated
structure by intent segment, and explain what learning-phase risk I
accept during the switch.
2. CONCEPT LIBRARY GAP ANALYSIS.
Score my current library against a 20-25 distinct concepts floor.
List the specific angles I am missing (fear-relief, cost objection,
waitlist, sibling story, credential proof, use-case-in-context,
transparent pricing, comparison to alternative). Prioritize the top 5
angles I should produce first based on my buyer.
3. PUSH DELIVERY VALIDATION PROTOCOL.
Design a weekly validation cadence: which day new concepts enter,
what percentage of budget Push Delivery holds, for how many days,
what Hook Rate and Hold Rate thresholds trigger keep-or-kill.
Include a log format so I can track hit rate over 8 weeks.
4. ADVANTAGE+ LEADS TEST DESIGN.
Design a 14-day controlled test comparing Advantage+ Leads with
Book Time versus my current best lead-goal ad set. Specify budget,
success metric (CPL, or booked calendar holds, or booked-to-showed
rate), Instant Form field limit, and the platform-specific
compliance rules I need to respect for my category.
Output format: a numbered action list I can hand to whoever runs the
account this week. No preamble. Direct recommendations, not options.
If you want a version of this prompt already loaded with your business context, competitive set, and account state, that is the kind of starting artifact we build for our fractional-CMO clients.
Six months ago, this operating model was aspirational. Last week it became executable. Founders who move first spend Q3 rebuilding structure and library while everyone else waits to see what happens. Watching platform shifts, translating them into a working plan, and holding the account owner accountable to that plan is a full-time job. Most founders do not need a full-time marketing hire. They need someone who can watch the platforms, run the plays, and manage the vendors already spending against those platforms.
That is what we do at Third Pillar. If your Meta spend is meaningful and your account still looks like it did in Q1, we should talk.
Ready to talk about whether your Meta operating model matches the platform you are actually running on? Book a call.
Product and operations are handled. If marketing is the missing piece, let's talk.