Mobile Advertising Performance Tracking

In-platform reporting is just a keyhole view of your business. It’s not the whole picture. I recently sat down with Justin Parker from ORIGIN® and Logan Brown from Triple Whale to discuss how DTC brands should be completely rethinking attribution moving into 2025. The key insight? Up to 40% of what shows as 'direct traffic' is actually coming from paid social. Here’s what happened when Origin tested this by pausing ads for 72 hours: - Direct traffic dropped 40% - Organic social fell 35% - Brand searches declined 50% This matters because most brands are killing winning campaigns based on incomplete data. A typical customer journey can look like this: - Sees ad on social - Opens Instagram to research - Checks reviews on Google - Types URL directly - Buys 3 days later Platform attribution: 0  Real impact: 100% The old way of tracking performance through platform-reported ROAS and last-click attribution is fading. Top brands are shifting to a new model focused on: - Total traffic impact - Cross-platform journey tracking - Brand search correlation Why? Well Origin was able to profitably DOUBLE their ad spend after seeing the real impact of their campaigns. They stopped killing winners based on bad attribution data. For DTC brands looking to scale, track these three metrics when scaling ads: - Direct traffic correlation - Brand search volume - Non-attributed sales If these rise with ad spend, your ads are probably working - regardless of what platform metrics show...

Imagine this: You’re running Meta ads. Sales are coming in. Things feel like they’re working. But you have no idea why. You don’t know which campaign is driving purchases. You don’t know if it was the reel, the carousel, or the story ad. You’re just… guessing. This is how most ecom brands run ads. Blind. And when performance drops? They panic. Shut things off. Start over. I’ve been there. I’ve run campaigns that looked like they were printing money… Until I checked the actual numbers. Turns out half the sales were coming from organic. One ad was burning cash. The rest were coasting. The fix? Dial in your tracking. No guesswork. Just data. Here’s how I do it now: • Use UTM parameters on every ad • Set up custom events with Google Tag Manager • Use Meta’s CAPI (Conversions API) alongside the pixel • Double-check data inside Triple Whale or Northbeam • Align your KPIs platform vs. backend vs. reality You can’t optimize what you can’t see. And Meta’s not going to do it for you. Track smarter. Spend smarter.

If you are running App campaigns on Meta, you should know this. Till 2021, Meta used to share the last-click level data for all your app installs. i.e., you get the details on any install's last-click time, location, and many more, as the data in your MMP. You can download this data from your MMP and use it to build models for ROAS, LTV, and fraud detection. Meta has referred to it as Advanced Mobile Measurement (AMM). However, they stopped it in 2021 in response to Apple's App Tracking Transparency. After stopping AMM, they only shared the aggregate with MMP. You will only get to see how much each source contributed to App installs. Now they are back. Last week, Meta announced that it will restart AMM this month. Interestingly, they will also start sharing the engaged view(leading to conversion) data as part of this from July 21st. One of the use cases for this data is in App fraud detection, specifically identifying parasitic ad networks. These ad networks detect when users click on Meta ads and generate a malicious click immediately after the Meta click, before the installation, thereby taking the installation attribution. We developed a parasitic ad network fraud detection rule for Google in 2022, utilizing UAC clickstream data while I was with Unacademy. It helped us save 20% of our Ad network budget without affecting the scale of new user acquisition. With AMM being available, now you can do the same for Meta as well. ♻️ Repost if you think this will be useful for your connection & followers. 🙏🏼

In the following situation, traffic from Google and Meta Ads are reported as direct/organic traffic in your #GA4 reports. Apple's IOS 17+ updates include a new feature called 'Link Tracking Protection', which automatically removes 'gclid' and 'fbclid' tracking parameters when a user uses the Safari private mode or Apple’s native Mail app or Messages. Without 'gclid' and 'fbclid' tracking parameters in place, traffic from Google and Meta ads could be reported as direct/organic in your GA4 reports. Other than that, IOS 17+ updates also directly affect Google's ability to track the performance of your Google Ads and Meta’s ability to track the performance of Meta Ads. The majority of Facebook users use the native Facebook app. So, Meta advertisers are unlikely to be impacted much by these updates. But these updates are not good news for Google Ads advertisers if the users are clicking on ads in Safari private mode as it will automatically disable the auto-tagging feature of Google ads, which provides a wealth of information to Google. Tagging your ad URLs with UTM parameters is more important than ever if you wish to continue to track the performance of Google and Meta ads in GA4 reports, as IOS 17+ updates do not remove the UTM parameters. Consider server-side tagging as a potential solution to mitigate tracking limitations imposed by browser and device restrictions. Stay updated with evolving privacy features and adapt tracking strategies accordingly.

Still letting ads run just because they used to perform? Here’s the reality: Most accounts waste 20–30% of their ad budget on creatives that should’ve been killed days ago. We built a dead-simple framework to fix that The 3/7 Rule for Ad Performance: Step 1: Track your top KPIs (CTR, CAC, ROAS) on both a 3-day and 7-day rolling average Step 2: Compare them daily Step 3: Act based on direction: - 3-day drops >20% vs. 7-day? You’re likely hitting fatigue — rotate creative. - 3-day improves >15% vs. 7-day? Scale it — you’ve got a winner. - 3-day flat vs. 7-day? Monitor — no major action yet. This works across Meta and Google. No fancy dashboards required. We use this logic inside our AI system to auto-flag underperformers and auto-scale winners — before the algorithm catches up. The result? 20% more efficient spend. No guessing. Just clarity.

One challenge that keeps coming up in our conversations with performance marketers: “How do we leverage AMC to get an idea about incrementality directionally?” Here’s a playbook that I’ve seen work well: 1. Pulled event-level AMC data (impression → click → purchase paths). 2. Used Xnurta’s no-code models to analyze pre/post-exposure cohorts 3. Layered on attribution modeling to track impact across touchpoints (including non-click engagements). The insight? A large portion of ROAS was being driven by mid-funnel ads, not last-click tactics. These weren’t the flashiest ads, but they were moving the needle in customer decision windows. Results: Surfaced 2x more value from custom audience retargeting than expected Enabled budget shift away from redundant low-funnel spend—without sacrificing efficiency This is the kind of clarity that lets growth teams scale smarter. And the best part? No black boxes. Every step traceable, every model transparent. We're moving into a world where performance isn’t about stating outcomes, it’s about explaining those outcomes. #AmazonMarketingCloud #Incrementality #Attribution #AI #Xnurta #PerformanceMarketing #AdTech #RetailMedia

#PPC Tip: By setting up a tracking template in Microsoft Ads, you capture each ad click accurately in #GA4, allowing for detailed performance analysis. Here’s a quick guide to setting up a tracking template: 1. Log into Microsoft Ads and decide if you're applying the template at the account level or for specific campaigns. 2. Navigate to the tracking section under "Account settings" for account-wide templates, or "Settings" for campaign-specific ones. 3. Enter your tracking template, incorporating UTM parameters, for example: {lpurl}?utm_source=bing&utm_medium=cpc&utm_campaign={campaignid}&utm_content={adgroupid}&utm_term={keyword} 4. Click "Test" to ensure your template is working as expected, then click "Save". This is what the tracking template above would track in GA4: Source of the traffic (utm_source=bing): Identifies the traffic as originating from Bing, allowing you to distinguish it from other sources in GA4. Medium of the campaign (utm_medium=cpc): Specifies that the traffic came through a cost-per-click (CPC) campaign, helping you analyze the performance of paid versus organic traffic. Campaign ID (utm_campaign={campaignid}): Dynamically inserts the ID of the specific campaign the click came from, enabling you to track and compare the effectiveness of different campaigns. Ad Group ID (utm_content={adgroupid}): Provides information on the ad group within the campaign that the user clicked on, offering insights into which sets of ads are performing best. Keyword (utm_term={keyword}): Captures the specific keyword that triggered the ad, allowing for granular analysis of keyword performance and its impact on user engagement and conversions. Keep an eye on your campaign performance in GA4 post-implementation, and be ready to adjust your template for any new tracking needs.

"Your ads are underperforming due to audiences." A brand told me this during our first call. Then we checked their Meta pixel. 0% event coverage. Their entire conversion tracking system had silently collapsed weeks ago. Not a single purchase was being recorded properly. Most marketers obsess over creative tweaks and bidding strategies while completely ignoring the technical foundation of their campaigns. When I audit underperforming accounts, pixel health is my first checkpoint. It's consistently the most overlooked aspect of performance marketing. What we typically find: - Events firing but not matching with CAPI - Duplicate conversion counts skewing ROAS data - Missing parameters causing audience fragmentation - Broken connections between platforms and Meta These invisible issues can drain your marketing budget faster than any creative problem. Monthly pixel health checks should be as routine as checking your campaign performance. It takes 30 minutes but can save thousands in wasted spend. Comment "Pixel" below and I'll share a video on conducting a proper health check for your tracking setup.

If you’re spending $20K+ on ads and still relying on pixel tracking… you’re behind. Here’s what top-performing accounts are doing instead: Most B2B marketers don’t even see it. For years, the go-to method was simple:  - Drop a pixel.  - Track pageviews.  - Feed that back into the ad platforms. But pixels are breaking. Privacy regulations, browser changes, and platform updates are killing pixel reliability. The signal is fuzzy. The attribution is off. And your performance is quietly suffering. That’s where server-to-server tracking (CAPI or conversion API) comes in. Instead of tracking user behavior via JavaScript on your site… CAPI sends conversion data from your backend systems to LinkedIn, Facebook, and Google. It’s cleaner. More accurate. Less reliant on browser behavior. Here’s what we tell clients: If you’re spending under $20K/month, pixels might be “good enough” If you’re spending $20K/month or more, you need to make the switch It’s not always plug-and-play. Sometimes, you’ll need dev resources or a specialist to set it up. But the lift is worth it. Because the ad platforms reward accurate feedback loops. More signal = better optimization = better performance. We’ve already moved several accounts over. In each case, performance improved within weeks. Not just in CPA, but in the consistency of results. If you wait 2 years to make this move, you’ll be late. If you start now, you’ll be ahead of your competitors. Server-to-server isn’t “advanced.” It’s just the new normal in 2025.

Today I fill your AppLovin obsession with a breakdown of how the Creative Optimizer really works. 🧠💥 After deep testing and analysis, here’s what drives actual spend inside the platform and why many teams optimize for the wrong things: ✅ Install Rate (IR) is the #1 signal. The system favors creatives that convert. ✅ Creative eCPM is calculated as: eCPM = CTR × IR × Payout That means the higher your IR, the better your eCPM and the more your creative gets delivered. Here’s a real campaign snapshot: Creative A → IR: 3.2%, eCPM: $28, CPI: $0.85 Creative B → IR: 2.9%, eCPM: $26, CPI: $0.90 Creative C → IR: 3.5%, eCPM: $30, CPI: $0.80 🟢 All three became top spenders, even though their ARPU was below target. Why? Because Applovin’s algorithm prioritizes early-funnel performance and it’s up to you to cut or scale based on ROAS and retention after. Top spenders are not your best D1 ROAS creatives they’re your highest IR performers. CPM and CPI help you measure efficiency, but they don’t determine delivery. IR does. ⚠️ If your ARPU or ROAS is off don’t blame the algorithm. It did its job: drive installs. You need to step in and layer performance insights after delivery kicks in. 💡 Takeaway: Creative success starts at the install. But real growth comes from knowing when to pivot from IR to ROAS. Watch your creative eCPM closely — that’s your signal to scale. Note : There is a project to work on the creative optimizer this year to improve it. So yes everyone can do UA, no everyone shouldn't do UA on applovin without experience or someone to help because it can be dangerous. #Applovin #UserAcquisition #MobileAds #CreativeTesting #AdTech #GameMarketing #PerformanceMarketing #CreativeOptimizer