Your CPI is fine. Your VALUE signal is 💩​ – Fix it!

Meta drives performance with VO through improved AI optimization

Wohooo! The big VO (Value Optimization – yes, I still call it VO) is finally here. Meta’s app campaigns now drive materially better outcomes when you optimize for value rather than volume.

Across 2025 improvements, advertisers optimizing for conversion value see +29% ROAS vs optimizing for conversion count.

Teams report stronger monetization, stability, and scale, and many are shifting the majority of spend to Value Optimization (VO). In parallel, Meta aligned its delivery with MMP reattribution windows (AppsFlyer flexible 0–180 days, 180-day “new user” exclusion for Adjust and Singular), cutting misclassified “old users” in new-user campaigns by ~20% in tests.

The net effect is a clearer signal to AI, higher-value payers, and cleaner reporting that aligns with your measurement.

The only “issue” I have with the picture above is why not to use both optimisations and still win? Instead of spending $200 and getting $500, I spend $400 and get $750 + more users.

What’s New and Why It Matters

• Value Optimization outperforms volume: Advertisers optimizing “maximize value of conversions” achieve +29% ROAS vs those optimizing for conversion volume. This makes VO the most performant Meta solution for ROAS-focused app advertisers.

Brutally Honest by Matej Lancaric is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.

• Better alignment with your MMP logic:
  • AppsFlyer: Meta honors a flexible reattribution window, 0–180 days, matching your AppsFlyer setting.
  • Adjust & Singular: Meta applies a 180-day exclusion for “new user” targeting to reflect their lifetime new-user logic.

Outcome

Extending the exclusion from 90 → 180 days produced ~20% fewer acquired users who fail your “new user” definition in 15 split tests (May–Sept 2025), 97% confidence.

So what: AI performs better when “success” is unambiguous. VO plus MMP-aligned windows reduces label noise, lets the model chase higher-value events, and makes third-party reporting line up with delivery.

Proof points & quotes

• FunPlus (Yang Jing, UA Manager):
  “We’ve seen value optimization performance improve by 5 percent over the past year and it’s now over 20 percent better than when we focus on the volume of conversion. This uplift means we’re acquiring higher-value players more efficiently and driving stronger ROAS and in-game monetisation across our titles.”
• CrazyLabs (Alexandr Snopkov, Lead UA Manager):
  “At CrazyLabs, value optimization campaigns have significantly boosted our portfolio performance. On iOS, VO has become the leading campaign type for our IAP-focused titles, such as Coffee Mania, driving standout results. Beyond the numbers, VO has brought the stability and scalability we were looking for in our UA strategy with Meta Ads.”
• Banditos Studio (Guy Galin, Head of UA):
  “In the last 90 days, we spent about 85% of our user acquisition investment with Meta on Value Optimization for our RPG, Heroes & Dragons. We are pleased with the results and plan to continue scaling the VO campaigns.”

The Data at a Glance

• +29% ROAS when optimizing for value vs volume.
• +5% VO performance improvement year-over-year at FunPlus, now 20%+ better than volume optimization.
• 85% of Meta UA spend on VO at Banditos Studio over the past 90 days.
• 0–180d reattribution flexibility (AppsFlyer). 180 d “new user” exclusion (Adjust, Singular).
• ~20% reduction in “not-actually-new” acquired users after extending exclusion 90 → 180 d in 15 split tests, 97% confidence.

Recommended next steps

1. Switch ROAS-critical titles to VO: Use “maximize value of conversions.” Confirm purchase value schema and currency.
2. Mirror your MMP settings in Meta:
   • If on AppsFlyer, set the reattribution window you actually want Meta to respect.
   • If on Adjust/Singular, plan on a 180-day new-user exclusion.
3. Audit event quality: Ensure purchase events fire reliably, include value, and map to monetization moments that predict LTV.
4. Test windows and budgets: Run short split tests to validate new-user quality and ROAS lift after window changes. Target similar audience sizes and budgets.
5. Scale with stability: As VO stabilizes CPI and value density, lean in with controlled budget ramps and creative iteration around payer intents.