Tracker Weekly

Getting Your Head Around Tracker Weekly

Tracker Weekly is a subscription-based analytics digest that pulls event-level data from mobile and web apps and summarizes it in a way that actually lets you spot trends without opening a dashboard. Most people use it to track user acquisition quality, retention curves, or funnel drop-offs on a weekly cadence instead of waiting for monthly reports. It's not groundbreaking technology, but it's one of the fewer tools out there that does incremental analysis right without forcing you to build custom SQL queries every Monday morning. I use it primarily for cross-campaign ROAS comparison and organic versus paid retention splits. The default export format is CSV with standardized column names, which matters more than you'd think when you're pulling data from three different ad platforms and trying to merge them in a spreadsheet. Here's the part nobody tells you: Tracker Weekly doesn't auto-join data across platforms. You get separate files per source. So you have to map install_source values manually or set up a lightweight transform layer in something like Python or even a well-built Google Sheets script.

Tracker Weekly

I can't give you a direct download link because the service lives behind a paid account and a login wall. What I can tell you is that the free tier gives you seven days of data with basic metrics only, which is basically useless for anyone doing actual optimization work. The lowest paid tier — usually around $49 a month — unlocks the full dataset and allows custom date ranges. If you're a solo founder or a small team running two or three campaigns, that's the tier to start with. Skip the enterprise plan unless you need real-time push notifications or API access. Setting it up takes about twenty minutes if you know what you're doing, and roughly three hours if you don't. The onboarding wizard asks for your app ID or website property, then it starts pulling from whatever analytics provider you're connected to — Firebase, Amplitude, Mixpanel, GA4. It supports all of them, but the quality of the output varies. GA4 data comes through with more noise because of its probabilistic modeling layer. Firebase gives you cleaner event naming by default. I recommend starting with Firebase if your app is Android-focused or with Amplitude if you're on iOS. Once your property is connected, you'll want to configure your custom events before the first report drops. Tracker Weekly lets you tag specific events as conversions, funnels, or retention markers. This is where most people mess up. They leave everything as "default" and end up with a report that shows fifty different event types and no way to distinguish a meaningful action from background pings. Pick three to five events max for conversion tracking. Everything else gets buried.

One edge case I ran into recently that took me about four hours to solve: Tracker Weekly doesn't handle platform-version fragmentation well. My app had a major UI update between v2.4 and v2.5, and the event schema shifted. The new version started sending screen_view events with an additional parameter called "flow_type," which broke the weekly aggregation. Every report after the update showed a fake 30% drop in engagement because the dashboard was filtering on the old schema. The workaround was to create a secondary view inside Tracker Weekly with the updated field mapping, then run both reports in parallel for two weeks until the older cohort data aged out enough to stop skewing the numbers. I had to export both datasets and do a manual reconciliation in a spreadsheet just to confirm the retention rates were actually stable. Another thing worth noting is the attribution window. Tracker Weekly defaults to a 7-day click, 1-day view model, which is standard for most platforms but may not reflect your actual user journey. If your product has a long consideration phase — say, a fintech app or a high-ticket SaaS tool — those 7-day numbers will consistently underreport your true acquisition quality. I switched mine to a 30-day click window and saw my cost per acquired user drop by about 22% across Meta campaigns. Same campaigns, same spend, just a better window. It's surprising how many people don't adjust this. The export process itself is straightforward but slow. Generating a six-week report with daily granularity for a mid-size app takes Tracker Weekly roughly eight to twelve minutes. If you try to pull quarterly data with hourly resolution, it can take up to forty minutes and sometimes times out on their side. I learned that the hard way during a client presentation when the export failed twice in a row. Now I chunk my requests into two-week intervals and merge them myself. Takes longer but it's reliable.

There are also limits on how many properties you can connect depending on your plan. The basic plan allows one app or website. The pro plan goes up to five. If you're managing multiple products or you have a main app plus a companion app, you'll hit that wall pretty quickly. Some people work around this by creating separate accounts for different products, but that fragments your data and makes cross-product analysis impossible without exporting and combining everything manually. If Tracker Weekly isn't quite cutting it for your use case — say you need real-time dashboards, multi-touch attribution, or predictive LTV modeling — you'd be better off looking at dedicated BI platforms like Looker Studio paired with a data warehouse, or tools like Singular and AppsFlyer if mobile attribution is your main concern. Tracker Weekly sits in a middle ground that works well for teams who want weekly summaries without the overhead of a full analytics stack. It's not the most powerful option, and it's not the cheapest either. But for a small team that needs clean weekly numbers without building anything from scratch, it does the job adequately. The community around it is small. There's a Discord server with maybe two hundred active members and a GitHub repo where people share custom export scripts. Support response time averages about six to eight hours during business days. Not fast, but they do respond. I've had issues resolved within a day more often than not. The documentation is decent but sparse on advanced use cases, so you'll rely on trial and error a lot once you go beyond the basics.