1

I've read Analytics: CloudFlare vs Google - What discrepancy is too big? but still can't put together an answer that the CEO of any company would accept. The numbers in CloudFlare analytics are much larger than those reported by Google Analytics.

If a website needs to give numbers to advertisers, should GA be used considering the JS involved is more close to the Google DFP way of thinking rather than the CF way? Should I give them the CF stats and when they question the discrepancy in numbers between what Google DFP says and CF says, tell them what CF says about their analytics?

Improve this question

2 Answers 2

Reset to default
2

We cover the topic pretty well about CloudFlare analytics versus Google Analytics.

We are, for example, capturing traffic like bots that Google would not pick up.

Improve this answer
2
  • We're not particularly bothered about bots as they do not represent real users. We want to sell advertising on our site so if we use CF numbers then this will be something like 30% above what GA suggests but we certainly will not be getting 30% more hits to the adverts Commented Jul 8, 2014 at 12:36
  • And hotlinked images right? Commented Feb 4, 2017 at 12:53
1

Based on the the information in Analytics: CloudFlare vs Google - What discrepancy is too big?, the discrepancy between Google Analytics and CloudFlare is almost entirely automated bot traffic.

There could also be some visitors that trigger CloudFlare but not Google Analytics because they have JavaScript disabled in their browsers. This number should be very small. In 2010, Yahoo found that only 2% of their US visitors had JavaScript disabled:

I suspect that the number of users that disable JavaScript has fallen over time because more and more websites are rich media that require JavaScript and use AJAX. Many popular web sites just don't work with JavaScript disabled.

Advertisers (and your CEO) only care about real people that see your website and are influenced by the advertising. Bot traffic should be excluded from your visitor stats, but JavaScript disabled should be included. Google Analytics may be under reporting by up to 2%, but it sounds like CloudFlare could be over reporting by 2000% (with a 20x discrepancy). Unless there is a way to filter the bot traffic out of the CloudFlare numbers, I would use the Google Analytics numbers. They are much closer to the number of actual human visitors that use your site.

Improve this answer
2
  • I agree that disabling JavaScript has probably dropped significantly, however, there is still a segment that deals with security issues that by routine disables JavaScript. I use this example as a hint. There are user segments that disable JavaScript routinely- i.e. highly technically aware people. I read a factoid last year the ascribed as much as %40 of such user segments disabled JavaScript. This was a reputable source. I did not fully believe this of course! It is just something to be aware of and may apply more strongly to some sites and not at all to others. Just a thought. Commented Jul 2, 2014 at 17:04
  • hotlinked images probably triggers a hit on Cloudflare whereas they should not. So I agree that Google Analytics should be more accurate. Commented Feb 4, 2017 at 12:53

Start asking to get answers

Find the answer to your question by asking.

Ask question

Explore related questions

See similar questions with these tags.