Google Analytics Understated Impact of Social on SEO

[goodroi]

Google Analytics is a great resource for many webmasters to make smarter SEO decisions. Just be careful that you aren't blindly following the data without understanding its limitation. A big limitation is the constant erosion of inbound data to GA due to privacy, security & technology issues.

We just finished running a test and found Reddit traffic was under-reported by 384% (of course ymmv greatly depending on site/industry). This is because many redditors use the mobile app which didn't send good data to GA. Once we adjusted the reports we were able to better correlate social success with SEO boosts.

To be clear I'm not saying saying social is the simple secret to SEO rankings or that GA is a bad product. I'm saying that GA can be more valuable if we remember its limitations. Making adjustments can empower us to make smarter SEO decisions. Some of those decisions might be to utilize social as part of a more holistic approach to online marketing. From my experience going viral on social can generate many strong ripples that indirectly boost multiple SEO factors.

To recap: GA not perfect, social is useful, & mobile apps can hide analytics.

[RedBar]

found Reddit traffic was under-reported by 384%

That's a strange way to present a figure. Do I assume you mean that your Reddit traffic was actually 3.84x more, or 384% higher, than reported.

I'm not questioning your figures, I'm simply making sure that I have understood you correctly.

[justpassing]

This is why it's better to rely on your own stats / logs than on third part service.

[whoa182]

I shared one of my articles to reddit just recently and I had a spike in traffic for the article (approx 3000 uniques in 2 days). It seemed like a large proportion of the traffic was labeled as Direct traffic rather than social or referrer Reddit.

[goodroi]

I wouldn't focus too much on the 384% since it is greatly YMMV depending on website, subreddit, etc. I wouldn't be surprised if others find the inaccuracy ranges between a factor of 1x to 6x. Reddit is trying to push its mobile users from browser to mobile app which greatly impacts analytics. IMHO this is not an isolated strategy. Many companies want to push mobile users into their app where they can better control the user experience & be granted access to more user information (contacts, location tracking, etc).

[Shaddows]

No, RedBar had the same reaction as me. You can't under-report by 384% as that would be a negative figure.

If you had 100 real visitors, a 75% under-reportage would be 25 visitors reported.

A 90% under-reportage would be 10 visitors shown. 101% under would be minus one (-1).

The true figure could be 384% bigger than reported = ~78.45% under-reported.

(As with RedBar, not disputing your numbers, just trying to make sense of them to "hook in" to your otherwise valuable insight)