From a student project to 4.6 million downloads: My app @MoviebaseApp journey started in 2017 as a simple homework project for learning Android. I was a beginner and the first version was a mess, but I realized there is more. The vision? 🎯 Building the world's best movie app

I just came across this book while preparing for my move. It helped me a lot to understand Universal Analytics! Time to say goodbye and ready for recycling in 3, 2...1 #measure

The first #MeasureCamp Zurich Ticket release starts in 30 minutes. Be prepared!

This is how it looks like:

My college spotted a new feature in #BigQuery which hasn’t been announced yet: Data Canvas, a simple miro style canvas to execute queries and visualize results via AI. Similar to https://t.co/RRz4ewXf65 but definitely not on the same level. Seams like its not fully rolled out…

There is no doubt that Tableau a super powerful tool, but the amount of workarounds you need for simple stuff like a WoW Scorecard drives me nuts.

Tableau folks: “Other data viz tools like Looker Studio are a joke…” Also Tableau folks: “Oh you need a Scorecard? Here is a 4000 character guide, it’s a bit hacky! #DataVisualization

It’s an undocumented way of testing Snowbridge and I was so exited that is works. I might write my very first blog post about it soon - stay tune.

The best way to test @snowplow 's event forwarding tool “Snowbridge” locally: HTTP rewrite rule to a local SP Micro instance with a proxy e.g. @RequestlyIO Micro & Snowbridge images piped via docker so that Micros stdout output is Snowbridge source. Boom: "real time" forwarding!

I think that makes sense in a way, since Google need to physically restore the data blocks / the partition, but it’s not so obvious in the first place.

Storage is cheap is a common argument for sloppy measurement design. A not so comment fact is, that update and delete statements are not so cheap for partition tables in #BigQuery since Google is charging the whole partition if one row is updated/deleted.

Is 100 Slot days already „Big Data“ 🤓? #BigQuery

Such a nice debugging experience: proxying analytics requests with @proxyman_app on a local @Snowplow Micro instance spun up via @warpdotdev

I came across this Safari feature flags and was a bit shocked. Happy end: not new and category „internal“ flags – not „experimental“ flags. #measure

Still a preview, but not impressive IMO: ▫️quality of the generated queries although Duet has a lot of context e.g. table schema and column descriptions ▫️ Autocompletion: sometimes a time saver but too often weird suggestions or suggestions when you don’t need them.

What is your experience with Duet AI in #BigQuery so far? What I like: ▫️ no tool switch to generate queries via AI is nice ▫️ ability to use the actual tables without fear to expose data

The debugging browser extension @snowplow Inspector got a huge update. ✅ new UI with JSON, schema & error view ✅ Synch with BDP console (Pipeline Config, enrichments etc.) ✅ Create Test Suites based on Tracking Scenarios Definitely useful additions to simplify debugging.

Alternative: @SimoAhava once mentioned the POST request header „Content-Length“ on measure slack. In SSGTM it could be used to: A) measure request size below 16 kb in GA4 B) truncate or strip lengthy event parameter from the request C) log requests beyond 16kb in another tool.

Error handling is difference between client-side & server-side #GA4 setups. The former is getting a 413 error above 16 kb, but the latter not. The SST GA4 Tag is throwing an EVENT_TO_LARGE error in the console.

Simple approach to find #GA4 requests which are exceeding the 16kb payload size limit in sever-side tagging setup with App Engine: Just execute this query in Log Explorer (caveat: request logging must be enabled).