Meta App Install Campaigns: How to Run Facebook & Instagram Ads That Convert

Every year, someone declares that Meta ads are dying. Every year, the apps that invest in running them well gain a significant competitive advantage over those that abandon ship.

Meta's advantage for app install campaigns comes down to three things. First, scale: with over 500 million daily active users in India across Facebook and Instagram, Meta offers an addressable audience that no other social platform matches. Second, intent signals: Meta's advertising system has access to billions of behavioural data points—content consumed, pages liked, purchases made—that allow it to find users similar to your best customers with remarkable precision. Third, creative surface area: Facebook Feed, Instagram Feed, Reels, Stories, and the Audience Network give you six distinct placements with different user mindsets and optimal creative formats.

Yes, iOS privacy changes (App Tracking Transparency) reduced signal fidelity from Apple devices, and SKAdNetwork attribution is less granular than the old IDFA-based measurement. But for Android-dominant markets like India, and for apps that have implemented Meta's Aggregated Event Measurement and SDK correctly, Meta remains the highest-ROI paid channel available.

Meta offers two primary campaign types for apps: App Promotion campaigns (the standard app install objective) and Advantage+ App Campaigns (Meta's automated, AI-driven format). In 2026, Advantage+ is now the default recommendation for most apps, but understanding when to use each matters.

Advantage+ App Campaigns. This format hands targeting, placement, and bid optimisation almost entirely to Meta's algorithm. You provide the app, the optimisation event (install, registration, purchase), the budget, and the creatives. Meta does the rest. In independent tests across dozens of apps we manage, Advantage+ campaigns deliver 20–35% lower CPIs than manually structured campaigns at equivalent budgets. Use Advantage+ as your primary campaign when you have at least 50 installs/day for the algorithm to learn from.

Manual App Install Campaigns. Use manual campaigns when you need to test specific audiences (competitor audiences, custom audiences from your CRM), control placement-level spend (e.g., Reels-only), or run experiments that Advantage+ would interfere with. Structure them as: one campaign → multiple ad sets (one per audience) → 3–5 ads per ad set (creative variants).

Recommended campaign structure for scale:

• Campaign 1: Advantage+ App Campaign — ₹500–₹2,000/day — broad install optimisation
• Campaign 2: Manual — Lookalike audiences (1%, 2%, 3% of top users) — install or registration optimisation
• Campaign 3: Manual — Retargeting lapsed users — re-engagement objective

Custom Audiences. Upload your existing user list (emails, phone numbers) to Meta to create a Custom Audience. Use this to exclude existing users from install campaigns (avoid wasting budget) and as the seed for Lookalike Audiences.

Lookalike Audiences. A Lookalike Audience finds Meta users who share behavioural and demographic characteristics with your seed audience. Create lookalikes from three different seeds and test them in separate ad sets:

• Lookalike of your top 5% highest-LTV users (most valuable—if you can pass this signal via your MMP)
• Lookalike of all registered users (broader, higher volume)
• Lookalike of your email waitlist or website visitors (useful if your app is new)

Start with 1–2% lookalike (most similar, smaller audience) and expand to 3–5% as you scale. In India, even a 1% lookalike of 100K users gives you an audience of several million—more than enough.

Interest targeting. For new apps without an existing user base, interest targeting is your starting point. Choose 3–5 tight interest clusters relevant to your app's core use case. Avoid broad interests like "Technology" or "Shopping"—they deliver high volume but poor quality. For a trading app: "Zerodha", "NSE India", "Sensex", "stock market investing".

Broad targeting (for Advantage+). When running Advantage+ campaigns, set no audience restrictions—let Meta's algorithm find users globally within your target geography. This feels counterintuitive but consistently outperforms targeted approaches at scale because Meta's behavioural signals are simply richer than any manually constructed audience.

Creative is the primary performance variable in Meta campaigns. Two advertisers with identical targeting and identical budgets can see 3–5x differences in CPI based purely on creative quality and format. Here is what works in 2026:

Reels-style vertical video (9:16). Reels is Meta's fastest-growing placement and has the highest organic-to-paid blending. Creatives that look like organic Reels—shot on a phone, natural lighting, voiceover or text overlay, no polished production aesthetic—achieve 40–60% lower CPCs than polished brand videos on this placement. Hook the viewer in the first 1–2 seconds with a problem statement or an unexpected visual.

Demo walkthroughs. A 15–30 second screen recording showing the app solving a specific problem. This format works because it sets accurate expectations and self-qualifies viewers—only people genuinely interested in the problem will watch through and install. Lower CTR than emotional/hook-led creative, but higher post-install engagement.

Testimonial/UGC videos. Authentic, user-filmed testimonials ("I used to do X manually, now [App] does it in 2 minutes") outperform produced testimonials almost universally. Recruit 3–5 real users to create short testimonial clips. Pay them ₹2,000–₹5,000 for usage rights. Test raw vs. lightly edited versions.

Static images. Do not neglect static. A clean static image with a bold benefit headline, a relevant screenshot, and a prominent CTA button still works exceptionally well on Facebook Feed and the Audience Network. Static is also the lowest-cost creative format to produce and iterate—ideal for rapid creative testing.

Carousel ads. Carousels work well for apps with multiple distinct features—each card highlights a different use case. Effective for productivity, finance, and utility apps where the breadth of features is a selling point.

Meta offers several optimisation events for app campaigns. Choosing the right one dramatically affects the quality of users you acquire.

Optimise for installs (CPI). Use this when you are early stage and need volume to generate learning data. Meta will find users most likely to install—but this optimises for quantity, not quality. Expect high install volume but potentially lower retention.

Optimise for a post-install event (CPA). Once you have 50+ post-install events per week (registrations, purchases, tutorial completions), switch to CPA optimisation. Now Meta finds users most likely to complete that valuable action. CPIs will be higher, but the users you acquire are significantly higher quality—typically 1.5–2x better D30 retention than install-optimised cohorts.

Value bidding (ROAS optimisation). If your app has in-app purchases or subscriptions and you pass purchase value events to Meta, you can optimise for revenue. Value bidding tells Meta to find users who are not just likely to purchase, but likely to spend more. This is the gold standard for monetised apps—set a minimum ROAS target and Meta will prioritise high-LTV users. Requires at least 100 purchase events per week to work effectively.

iOS and SKAdNetwork considerations. Apple's SKAdNetwork (SKAN 4.0) means attribution for iOS campaigns flows through Apple's privacy-preserving system with a 24–48 hour reporting delay and limited signal fidelity. Configure your Meta Aggregated Event Measurement (AEM) setup through your MMP (AppsFlyer, Adjust) before launching iOS campaigns. Prioritise post-install events with high enough volume for SKAN reporting, and supplement with modelled attribution data from your MMP to make optimisation decisions.

Without proper attribution, you are flying blind. Here is the tracking stack you need before scaling Meta campaigns:

Meta SDK integration. The Facebook SDK (or Meta SDK) must be integrated into your app and configured to fire the standard app events: fb_mobile_activate_app, fb_mobile_complete_registration, fb_mobile_purchase. This enables Meta's algorithm to receive install signals and optimise. Test via Meta Events Manager before launching.

MMP integration (AppsFlyer or Adjust). A Mobile Measurement Partner sits independently between your app and all your ad networks. It provides deduplicated, fraud-filtered attribution data across Meta, Google, and any other paid channels. This is essential once you are running more than one paid channel simultaneously—without an MMP, you will double-count installs and misallocate budget.

Deep linking. Configure Meta's Deep Link destinations in your ad creatives so that users who tap your ad open directly to the relevant screen within the app (rather than the app home screen). For example, a "Start trading" ad should deep-link to the trading screen post-onboarding. Deep links improve Day 1 retention because users land exactly where the ad promised to take them.

UTM parameters for web traffic. If your Meta ads also drive web traffic (to a landing page before the store), tag all URLs with UTM parameters and connect them to your web analytics. This gives you a complete picture of the conversion funnel from ad impression to app store to install.

Weekly creative rotation. Creative fatigue is the #1 cause of rising CPIs on Meta. Monitor frequency (average number of times a user in your audience has seen your ad) and CPM trends. When frequency exceeds 3–4 for your core audience or CPMs rise more than 20% week-over-week, it is time to introduce new creative. Build a production pipeline that delivers 3–5 new creative variants every week.

Audience exclusions. Exclude existing installs from all campaigns. Upload your current user list as a Custom Audience and add it as an exclusion in every ad set. This prevents budget wastage on users who already have your app—a surprisingly common and expensive mistake.

Budget scaling rules. Scale winning ad sets gradually—increase budget by 20–30% every 48–72 hours. Increasing budget too quickly (more than 50% in 24 hours) resets the learning phase and typically causes CPI to spike temporarily. If you need to scale faster, duplicate the ad set and run both in parallel rather than increasing the original's budget sharply.

Dayparting and placement analysis. Review performance by hour and day. Many consumer apps see CPIs 30–40% lower during evening hours (7 PM–11 PM IST) when users are browsing casually. If your campaign data shows this pattern, use ad scheduling to concentrate spend during peak performance windows.

A/B test landing pages. If your Meta ads drive to a landing page before the store, test different page variants with Meta's built-in A/B testing. A headline change or different hero image can lift click-to-install conversion by 15–25%.

Running Meta campaigns at scale requires constant attention—creative, audiences, bids, and tracking all need regular review. If you want an expert team managing your Meta UA, get in touch with Vmobify. Our user acquisition service includes full Meta campaign management, creative production, and MMP setup. We have run Meta app install campaigns for over 100 apps and driven more than 30 million installs across categories.