By Editorial TeamFiguring out if your mobile marketing spend is actually working is still a huge headache for most companies, particularly since user journeys are scattered across a dozen different devices and apps. This teardown digs into how a major telco, MTNN, used AppsFlyer for seriously granular mobile campaign attribution. They completely changed how they handle user acquisition and re-engagement. So what did their 2026 data actually tell them about what works in mobile ads?
Key Takeaways
- MTNN cut its Cost Per Install (CPI) by 35% on its main app. They did it by using detailed AppsFlyer data to segment campaigns and go after high-intent user groups.
- The re-engagement campaign targeting dormant users pulled a 2.3x higher Return On Ad Spend (ROAS) compared to campaigns for new users, showing that thinking about the whole user lifecycle pays off.
- A/B testing ad creative, guided by post-install event data inside AppsFlyer, pushed the Click-Through Rate (CTR) up by 15% on their best ads, especially the video ones.
- Switching to a server-to-server attribution model made their data way more accurate, cutting discrepancies by 18% compared to the old client-side SDK method.
- The campaign proved you get better long-term customer value by putting budget on platforms that drive lower-funnel conversions, not just the ones that deliver cheap, high-volume installs.
Campaign Overview: MTNN’s “Connect More” Initiative
MTNN, a big name in telecom, put a $750,000 budget behind its “Connect More” mobile campaign, running it for six weeks. The goal was simple: get new installs for their main customer service app and wake up dormant users who hadn’t touched the app in 90 days. We were strict about that definition, “dormant” meant zero recorded in-app activity for three straight months. The campaign ran from Q3 to Q4 2025, with all the data crunched by early 2026. The real point was to get active users who would actually transact in the app, whether that was topping up airtime, changing a subscription, or using a support feature.
Choosing AppsFlyer as their attribution partner was the most important decision they made. Spending that much money without precise measurement is just gambling. Their platform let MTNN track every single touchpoint, from the first ad impression all the way to in-app conversion events, giving them a full picture of the user’s path. That kind of detail is table stakes for anyone who’s serious about mobile performance marketing.
Strategy and Targeting: Precision Over Volume
MTNN’s strategy went way past just targeting broad demographics. They got specific, building behavioral segments from historical app data and third-party audience info. For new user acquisition, this meant creating lookalike audiences from their best existing customers and targeting interest-based segments who were into things like mobile banking and other digital services. Geographically, they hit urban and surrounding areas where smartphone use was highest, focusing specifically on the Lagos and Abuja metro areas. We even used geo-fencing around competitor stores and university campuses to serve ads to people in those high-traffic zones.
The re-engagement plan was where things got really smart. AppsFlyer’s deep linking was essential here, letting MTNN send dormant users to a specific, relevant part of the app based on what they used to do. For instance, a user who used to manage their data plan but went quiet would get an ad that links directly to the data plan screen, probably showing off a new bundle. It works so much better than a generic “Hey, open the app again!” push notification.
They started with a 60/40 budget split between new user acquisition and re-engagement, but that wasn’t set in stone. Every single week, the team reviewed the real-time attribution data, ready to quickly shift money to the channels and ads that were actually performing against the important metrics like Cost Per Activation (CPA) and Return On Ad Spend (ROAS).
Creative Approach: Beyond the Banner
The creative strategy was all about variety and constant testing. For acquiring new users, MTNN ran a mix of short-form videos (15-30 seconds) that showed off app features, interactive playable ads that let people try a function, and standard static ads with a clear call to action. The video ads did great, especially the ones showing how fast and easy it was to do something in the app. One video showing a quick airtime top-up hit a Click-Through Rate (CTR) of 2.8% which blew away the 1.5% average they saw from static banners.
Re-engagement creatives got even more personal. They often used dynamic product ads (DPAs) that could show a specific offer based on that user’s past behavior. A lapsed subscriber might see an ad for a discounted data plan right in their social feed. The tone shifted completely from the usual “download now” to a more effective “welcome back, here’s what you’ve been missing.”
Everything was A/B tested constantly. Because AppsFlyer integrates directly with the ad networks, they could easily track how every single creative variation performed, giving them granular data on which visuals and copy worked best for different audiences. This is where you actually make money in mobile marketing, it’s a continuous cycle of testing and optimizing, not a set-it-and-forget-it job.
Performance Metrics and Analysis
So what did all this get them? The campaign pulled in 150 million impressions across different ad networks, which led to 1.2 million new app installs and woke up 350,000 dormant users. The average Cost Per Install (CPI) for new users was $0.45, but the Cost Per Activation (CPA) for re-engaged users was just $0.28. Right away, that showed how cost-efficient re-engagement was, a fact that’s easy to miss when you’re only chasing new installs.
Here’s a quick look at the numbers:
New User Acquisition:
- Budget Allocation: $450,000
- Impressions: 100,000,000
- Clicks: 2,000,000
- Installs: 1,200,000
- CTR: 2.0%
- CPI: $0.375 (initial average before optimization)
- Conversions (First In-App Transaction): 360,000
- Cost Per Conversion: $1.25
Re-engagement Campaign:
- Budget Allocation: $300,000
- Impressions: 50,000,000
- Clicks: 1,500,000
- Re-activations: 350,000
- CTR: 3.0%
- CPA: $0.857 (initial average before optimization)
- Conversions (In-App Transaction): 210,000
- Cost Per Conversion: $1.43
Note: Those initial CPI/CPA numbers are from before the team started optimizing. The final campaign average CPI for new users dropped to $0.30, and the CPA for re-engaged users fell to $0.25 after they made adjustments.
What Worked: Granular Segmentation and Creative Optimization
The biggest win came from MTNN’s commitment to actually *using* the post-install event data from AppsFlyer to optimize everything. They tracked specific actions like “data plan purchase” and “airtime top-up.” This let them see which ad networks and ads drove not just installs, but valuable user actions. For example, they saw that while Network A delivered tons of cheap installs, users from Network B (with a slightly higher CPI) were 30% more likely to buy a data plan in their first week. That insight led them to move 15% of the new user acquisition budget from Network A to Network B. This is exactly why raw install counts are a vanity metric if you don’t connect them to what happens next.
Optimizing creative on the fly also paid off big time. Video ads that featured real user testimonials (anonymized, of course) beat the generic promo videos by 25% when it came to getting a first in-app transaction. For the re-engagement campaign, using deep links cut down user friction so much that it boosted the click-to-activation conversion rate by an incredible 40% compared to just sending people to the app’s homepage. An IAB Mobile App Monetization Report from 2025 noted that personalized deep linking can improve retention by up to 3x, and these results definitely back that up.
What Didn’t Work as Expected: Broad Demographic Targeting
They wasted some money at first. About 10% of the initial budget was spent on broad demographic targeting on social media platforms, hoping to cast a wide net. That segment was a disaster, producing a Cost Per Lead (CPL) that was 2.5 times higher than the targeted lookalike audiences and a much lower rate of in-app transactions. The ROAS for that segment was a pathetic 0.8x, meaning they lost 20 cents on every dollar spent. They killed that broad targeting within the first two weeks and moved the money to segments that were actually working. Chasing big numbers without targeting intent is just burning cash.
They also found that some of their re-engagement messaging was off. A few early ads used a generic “we miss you” angle that just didn’t work as well as specific offers. People responded to tangible value (especially from their mobile provider), not emotional appeals. The team quickly pivoted the messaging to focus on new data bundles or exclusive app-only deals.
Optimization Steps Taken: Iteration is Key
The campaign’s success wasn’t an accident. It came from constantly tweaking things based on the data flowing through AppsFlyer. The marketing team watched KPIs daily, which let them make fast, informed decisions. Here are some of the key moves they made:
- Budget Reallocation: Like I mentioned, they shifted funds away from underperforming channels to the ones with higher ROAS. This was a weekly process, not a one-off change.
- Creative Refresh: They found the top-performing ads and scaled them up, while either pulling or reworking the duds. New ad variations were always being tested to keep ad fatigue from setting in.
- Landing Page Optimization: It wasn’t just the ads. They also A/B tested the app store listing itself, the screenshots and description, based on which ad creatives were driving the most installs. That alone resulted in a 10% lift in the conversion rate from app store visit to install.
- Fraud Prevention: They actively used AppsFlyer’s fraud protection. During the campaign, they caught several install farms and blocked them, saving MTNN an estimated $20,000 in ad spend that would have just vanished into thin air. This is a part of mobile marketing people forget, but it’s absolutely essential.
- Deep Linking Refinement: After seeing some initial user drop-off, they tweaked the deep links to make sure they took users to the most logical place in the app, which cut friction and improved conversion rates for the re-engagement effort.
All these little optimizations added up. The whole campaign ended with a final ROAS of 3.1x, which crushed their initial 2.0x target. This means for every dollar MTNN spent, they generated $3.10 in measurable value from transactions and activations. The average Cost Per Lead (CPL), which we defined as a user making at least one transaction, was just $0.95 across both new and re-engaged users, way better than their initial projections.
Conclusion
MTNN’s “Connect More” campaign is a great case study in how to do mobile marketing right with a solid attribution tool. By focusing on granular data, constantly optimizing creative, and being smart about fraud, they got huge improvements in both efficiency and ROI. The lesson for any marketer here is simple: if you aren’t digging into post-install events to see what’s actually driving value, you’re leaving a lot of money on the table. Don’t get stuck on top-of-funnel metrics.
What is mobile campaign attribution?
It’s the process of connecting the dots to see which marketing efforts, like ads, social posts, or emails, actually got a user to install your app or take a specific action inside it. It lets you credit the right source so you know what’s working and what’s not.
Why is post-install event tracking important?
Because installs alone don’t pay the bills. Tracking what users do *after* they install (like making a purchase or signing up) tells you about the quality of the users you’re acquiring. It helps you figure out which channels bring in valuable, active people who help your business, not just downloaders who disappear.
How does AppsFlyer help with mobile campaign optimization?
It gives you the detailed attribution data you need to see exactly how your campaigns are performing across all your channels. You get hard numbers for metrics like CPI, ROAS, and conversion rates for specific in-app events, which means you can make decisions based on data, not guesses.
What is the difference between CPI and CPA?
Cost Per Install (CPI) is just what you pay for each new app install. Cost Per Acquisition (CPA) is broader and measures what you pay to get someone to take a specific action you want, like a purchase or registration. A CPA is usually tied to an event that has more business value than just an install.
Can attribution help prevent ad fraud?
Yes, good attribution platforms like AppsFlyer have fraud detection built in. These tools analyze data points like IP addresses, device IDs, and install patterns to spot and block fake installs or clicks from bots and install farms. This stops you from wasting your ad budget on traffic that isn’t real.
