3 ms·
This looks like a solution in need of a problem. Previously I had tried some similar things at scale, mainly because I wanted upsampled reports and GA charges $
by redm 7y ago
This looks like a solution in need of a problem. Previously I had tried some similar things at scale, mainly because I wanted upsampled reports and GA charges $100,000 for premium services. What I found is that raw logs are not reliably accurate. The volume of traffic in certain countries was accurate but not reliably “real” users even after accounting for known bots, search engines etc. GA has a way of accounting for these and giving you a better overall picture. The second thing is that GA has improved its service so you get upsampled reports now, even at scale for tier 1 reports. At low volume, there upsampled already.
I’m not sure why anyone would want to waste time with this.
- RA_Fisher 7y agoI prefer the raw logs. One reason is that Google performs "statistical improvement" like "sessonizing." This act destroys the value of point process data. It wouldn't be found in a stats textbook because the operation destroys valuable information. Also, the concept of Visitors and Users that GA uses isn't transparent to me. Another fun fact. Since the tracking happens on the client side, there's potentially a ton of truncated data that GA simply misses. Backend server instruments don't suffer the same way.
- redm 7y agoI think those are fair points; mileage will depend on your end-goals. We want to know how our traffic relates to real-world ad deliverability, real users in our funnel, etc. I'd agree with "Also, the concept of Visitors and Users that GA uses isn't transparent to me." but I'd add that whatever they do, is more accurate than what we could get from raw logs, as it relates to relatable business metrics.