Data freshness in GA4 basically means how recently Google has collected, processed, and reported your GA4 metrics. In simple terms, it tells you how up-to-date your reports are. If it takes 20 minutes to process, then your data is 20 minutes fresh. Pretty straightforward, right?
But here's the catch: freshness isn't always consistent because Google processes data at different intervals. There are three main types of freshness: realtime, intraday, and daily:
Realtime: It's updated basically in real-time (we're talking typically less than a minute here). It lets you keep tabs on what users are doing right now, like live events happening on your site or app. But - and this is a big but - it's got limited dimensions and metrics compared to other intervals. You won’t get the complete picture here, just a quick glance. I usually use the real-time dashboard for confirming my page views and other events (a lazy debug mode)
Intraday: Intraday updates are like the in-between sibling. They’re not instant, but they’re faster than daily. Intraday data gives you a glimpse of what happened throughout the day, and it’s refreshed multiple times. You'll get updates roughly every hour if you’re a Google Analytics 360 user. For standard free GA4 users, it’s more like every 4-8 hours. It’s not perfect though - sometimes you’ll see gaps in event-scoped data, especially for traffic sources. These gaps are normal and usually filled when the daily data becomes available.
Daily: This is where all the data pieces come together. Daily processing wraps up everything for the day and presents the whole story. It's the most complete version of your data, but it takes longer to get there - usually 12 to 24 hours. It’s got all the features you’d want and the complete set of dimensions and metrics, so if you’re running detailed analyses or using Explorations, this is what you’re waiting for.
What to Expect: Data Processing Intervals
Each day, Google processes data in three steps - realtime, intraday, and daily. It’s kind of like baking bread. You’ve got the first quick mix (realtime), then the rising dough (intraday), and finally the baked loaf (daily). Here’s how they break down:
- Realtime (Less than 1 min): Great for watching live activity or debugging.
- Intraday (4-8 hours): A midway point where Google’s starting to assemble most of your data, but it’s not quite all there.
- Daily (12-24 hours): The complete dataset for that day. You want this if you need everything accurate and all metrics to make sense.
A Few Gotchas with Intraday Data
Intraday is super handy, but it’s not perfect. Since it’s in the middle of the processing cycle, you can expect a few quirks:
- Missing Traffic Source Info: Sometimes traffic source dimensions (like source, medium, or campaign) might show gaps. That’s normal until the daily data fills in the blanks.
- Paid and Organic Attribution: Google might temporarily use the paid and organic attribution model before it processes the daily data. You’re basically in a holding pattern until the final verdict.
- Cardinality Issues: Ever see an "(other)" row in your data? Intraday data might apply stricter cardinality limits, so during that period, you’re more likely to see those annoying "(other)" rows if you’ve got high cardinality dimensions.
When Does Daily Data Actually Show Up?
If you’re wondering when you can see a complete data picture for yesterday, here’s a rough schedule for when daily data typically becomes available:
- 12:00 am: Google stops collecting data for the day.
- 5:00 am: Most BigQuery exports are ready for 360 customers.
- 11:30 am: Daily data is ready in Explore.
- 3:30 pm: Reports are ready with daily data.
These are just rough guidelines, not a guarantee. Processing times can vary -especially if you’ve got a large property or there’s some slowdown.
Data Categories and Limits
Google also categorizes your property based on the volume of events it processes:
- Normal: Less than 25 billion events.
- Large: Between 25 billion and 250 billion events.
- XLarge: More than 250 billion events.
If you’re a "Normal" or "Large" property, you can end up temporarily as an "XLarge" if you’re collecting 15 billion events per day for a week. The size of your property affects processing time, so big properties might take longer to get that daily data ready.
TL;DR
GA4 data freshness is all about how fast data gets collected and processed. You’ve got three types: real-time (super quick, limited), intraday (multiple times a day, not complete), and daily (full data, but it takes a while). Data might be delayed, and big properties can experience longer processing times.
