Objective

Instamart needed to lower Cost Per First Purchase and scale Instamart more efficiently. The team was running highly localized and category-based campaigns across multiple cities, audience segments, and creative types.

At that scale, manual optimization became unsustainable. The team was pulling performance data from Appsflyer and Meta dashboards daily to adjust budgets, creating heavy operational overhead, slower optimization cycles, and budget dilution across too many small campaigns. This fragmentation increased CAC and made it challenging to scale top-performing campaigns consistently.

Solution

Smartly introduced AI-powered Predictive Budget Allocation (PBA) to simplify Instamart’s fragmented campaign structure and automatically direct spend toward the campaigns most likely to drive first purchases.

To further enhance performance, PBA incorporated Appsflyer’s last-touch attribution data—the source of truth for actual first-purchase performance—to inform its predictive models and real-time budget decisions. This ensured spend was optimized against actual acquisition outcomes rather than proxy platform metrics.

By consolidating budgets, removing overlapping campaigns, and automating reallocations, PBA delivered a scalable, always-optimized setup that reduced CAC, unlocked incremental conversions, and eliminated the need for daily manual optimization.

Key benefits delivered:

• Budget allocation optimized against first-purchase performance
• Automated shifts toward the highest-impact campaigns
• Faster optimization with ~1 hour saved per day in manual work