Actually, more and more companies, including mine, are removing GA to simplify compliance and avoid tool overlap 💸
Keeping both adds cost and complexity, especially since GA isn’t required for GAds if conversions and audiences are tracked directly.
Sounds awesome! definitely curious to see how it shapes up.
Awesome! Thanks
Hi David,
Is there a way to get the Analytics Debugger extension on Microsoft Edge? I looked in the store and on your website but couldn’t find it. I think it used to be available but doesn’t seem to be now.
Thanks!
Before that, I was using GTM: smooth and intuitive. ✨ Switching to Commanders Act felt like going from automatic to manual. It required way more Javascript skills than GTM. Tough at first, but diving into the nitty-gritty of tracking taught me the mechanics inside out.
2 years ago, I was thrown into Commanders Act and Javascript tracking. I remember staring at the screen, wondering if I’d ever get the tags to fire correctly. 🫣 Now, I set up tags like muscle memory. It’s wild how much you can learn when you’re dropped in the deep end.
Thanks for sharing! We’re planning to run them together for a bit before switching off GA.
We’re planning to run them together temporarily to make sure everything works well before shutting down GA. Just want to be sure the migration goes smoothly!
Good to know! I’m not worried about the metrics since we’re planning to fully switch to the second tool at the end. I’m more thinking about the technical side, like if running GA4 and the other one together might slow things down or cause issues.
To those in mobile analytics :
Have you ever run 2 mobile analytics SDKs in parallel (e.g., GA4 + another analytics tool)? Curious if it helped or caused more trouble: performance hits, data overlap, or maintenance headaches? Worth it, or a lesson learned? Would love to hear your experiences!
Yeah!
I’d add _STREAM_ID for identifying streams in inserts and _CHANGE_TYPE for tracking changes in CDC queries.
I remember when you could just buy a computer program like Microsoft Word or Photoshop and just have it on your computer forever without having to pay monthly or yearly. What a time that was
Oh merci, Mehdi! It worked perfectly. Appreciate the help!
Trying to connect lines to shapes in Looker Studio feels impossible lately. The lines keep moving around, and it’s so frustrating to get them to stay where they’re supposed to. 😖
For example, I’m trying out Plotly’s dcc.Graph for interactive charts and using vaex to simplify big datasets. So far, it’s looking pretty good for keeping things fast and smooth!
To tackle the performance challenge, I’m experimenting with data aggregation techniques and async rendering. The goal is to keep visuals interactive, even with larger datasets.
First time working on dynamic data visualizations in Python with Visual Studio. It’s fun, but managing interactivity without slowing down performance is a tricky challenge! 🫥
This is a huge move and could clear the way for more US-based solutions to get the green light. Let’s wait n see!
Truth!
Team asked for a quick data pull. I said, “Sure, just 5 minutes.”
3 hours later, I’m knee-deep in SQL joins, questioning the meaning of life, only to realize… they gave me the wrong dataset to start with.
Data analysis: where “quick” is always a lie.
Yes, that sounds like an amazing idea! We’ve done some workarounds for continuous reporting, but nothing perfect. Plus, dealing with another analytics tool alongside Google Analytics has been a whole different challenge. 🙃
Open sourcing your SQL would be super helpful for sure.
It’s not always smooth—deadlines clash, priorities shift, but when IT and data teams find their rhythm, we don’t just build dashboards or pipelines. We build trust.
IT loves specifics: “What exactly do you need?” Data teams thrive in nuance: “It depends.” It can get messy, but the magic happens when we stop debating details and start co-creating solutions that work for both sides.
Working with IT teams as a Data Analyst in an agile setup feels like running a relay race where we forgot to agree on the handoff. They work in sprints, we deal with constant changes. The trick? Meeting halfway; planning ahead while staying flexible for when the data world surprises us.
Merci Mehdi !
Thanks, James!
& sometimes what’s missing is the stakeholder who swore this data was “urgent” … until they disappeared without a trace. 😏
Data taught me one thing: silence speaks volumes. It’s not always about what’s in the numbers>it’s about what’s missing. The question no one asked, the metric no one tracked. The best insights come when you lean into the gaps and ask, ‘What aren’t we seeing?’
That’s a good point! Batch uploads can mess with eventtimestamp, but it’s usually fine for most analyses. Adding a custom timestamp is a smart move if you need exact times, especially for debugging or precise workflows
2. Use the remainder (%) to calculate remaining minutes and seconds.
Combine them with CONCAT and LPAD for a clean HH:MM:SS format!
