Marketers love control. For years, we’ve built campaigns by slicing and dicing audiences: age brackets, interests, income levels, geographies. But the reality is shifting fast. With Google Performance Max, AI-driven Search Ads, and Meta Advantage+ campaigns, the lever has moved from targeting to signals and creative variety.
At Socialee, we see this every day across industries—pharma, universities, gaming apps, e-commerce, and healthcare. The old playbook of “refine audience, repeat” just doesn’t deliver the same lift anymore.
The Platforms Have Changed the Rules
1. Google AI Search Ads
Ads are now being placed inside Google’s AI overviews. You can’t “target” these directly. Google decides which ads show based on existing campaigns, product feeds, and conversion signals. If your feed is weak or your signals aren’t clean, you simply won’t show up.
For example, when we worked with Zydus Hospitals on search campaigns, we realized refining cancer-related keywords alone wasn’t enough. The real lever was improving the landing page signals and structuring conversion tracking—only then did the ads gain visibility where it mattered.
2. Performance Max
PMax campaigns don’t honor audience boundaries in the traditional sense. You provide audience signals—like customer lists, in-market behavior, or custom intent—and Google uses them as hints, not rules. What matters more is how strong your assets and feed are, and how well your conversion tracking is set up.
With one of our e-commerce clients in organic foods, we shifted from micro-audience campaigns to a PMax approach. By cleaning up product feed attributes and expanding creative sets, we scaled from 25–30 monthly orders to consistent growth—without obsessing over audience layers.
3. Meta Advantage+
On Meta, Advantage+ automates placements and targeting. The system optimizes around conversion signals and creative combinations. Your job is to supply enough high-quality variations for the algorithm to learn quickly.
Take Yes Rummy—a gaming app. Instead of trying to refine interests like “online games” or “card games,” we built diverse creative libraries: UGC-style ads, gameplay snippets, trust messaging. Meta’s system picked the best combinations, and performance improved as we kept feeding it new variants.
Why Signals > Targeting
Signals tell the algorithm what good looks like. Conversion events, customer data, and clean product feeds are now your targeting. Instead of spending hours on micro-audience definitions, spend that time improving the quality of your signals:
- Define primary conversion events (purchases, leads, high-value actions).
- Upload clean customer lists and value-based audiences.
- Optimize your product feed titles, descriptions, and labels.
When the platform knows what success means, it finds the people.
Why Creative Variety > Creative Perfection
Most teams over-invest in “the perfect ad” and under-invest in variety. But AI-driven ad platforms don’t want one perfect asset. They want many good ones—so they can test, mix, and scale what works.
For example, with Amity University Online, we tested two creative themes: “career growth” vs. “flexibility with studies.” Instead of deciding for the platform, we built multiple variants under each theme. The system quickly picked winners—and the winning theme wasn’t what internal teams had initially bet on.
The New A/B Testing Playbook
Old way: test audiences against each other.
New way: test themes against each other.
- Phase 1: Run 3–4 themes with a handful of creative variants each.
- Phase 2: Identify the winning theme by looking at conversions, CPA, and ROAS.
- Phase 3: Scale the winner by adding 5–10 new variants.
Step by step Socialee’s framework:
Step 1: Define Themes
Pick 3–5 high-level themes that reflect your brand’s value propositions.
Theme Types to Consider (examples from Socialee’s client work):
- Problem-Solution (Rama Water Filter → “Hard water ruining appliances? Here’s the fix.”)
- Lifestyle Aspiration (Amity University Online → “Study without pausing your career.”)
- Trust & Proof (Zydus Hospitals → “20,000+ successful cancer treatments.”)
- Emotional Storytelling (Naysan Home Healthcare → patient family narratives)
- Offer/Value-led (Yes Rummy → “Play Free Rummy. No cash. Just fun.”)
- Innovation/Feature Highlight (Neo Mobility → “EV charging at your doorstep.”)
Step 2: Testing Logic
- Launch 3–5 themes simultaneously across the same audience/signal setup.
- Each theme should have 3–4 creative variants (videos, carousels, statics).
- Don’t mix messages inside one theme → keep it clean for signal clarity.
Step 3: Time to Wait
- Meta / Google PMax → Allow 7–10 days minimum before judgment (avoid learning phase bias).
- Google Search → 10–14 days, depending on daily search volume.
- Don’t touch campaigns too early; let the algo stabilize.
Step 4: KPIs to Track
Instead of vanity CTRs, track signal-led KPIs per theme:
- Awareness / Engagement Phase → CTR, Thumbstop Rate (Meta), Video Completion %
- Consideration Phase → Add-to-cart %, Lead Form Completion %, Micro-conversions
- Conversion Phase → Cost per Conversion, ROAS/CPA vs benchmark
- Scaling Signal → Consistency of results across 7-day rolling averages
Step 5: Scaling the Winner
- Pick top 1–2 winning themes (based on KPIs).
- Scale by:
- Adding creative variants under the same theme (different formats, lengths, hooks).
- Expanding platforms (e.g., theme works on Meta → adapt for Google PMax/Youtube).
- Broadening audience signals (but keep the theme constant).
This builds momentum. Instead of chasing audiences, you’re fueling the system with inputs it can actually use.
What This Means for Marketers
- Stop over-optimizing audiences. The platforms are better at finding incremental buyers than we are.
- Invest in data hygiene. Signals are your targeting.
- Build a creative engine, not one-off ads. Variety beats perfection.
- Measure at the right level. Look for winning themes, then double down.
Final Thought
The future of advertising isn’t about finding the “perfect audience.” It’s about giving algorithms the right fuel: strong signals and enough creative variety to learn fast and scale.
So here’s the question worth asking inside your team: are we still trying to outsmart the algorithm with targeting, or are we building the creative and signal engine it actually needs to perform?
Ready to move beyond outdated targeting tactics? Contact Socialee today. Our team will help you set up stronger signals, build creative variety, and create a scalable framework for real growth.