4 ms·

Google hates Survey.io - 32% drop in SE traffic

15 points by vaksel 17y ago

jbyers 17y ago
http://en.wikipedia.org/wiki/Correlation_does_not_imply_causation http://en.wikipedia.org/wiki/Correlation_does_not_imply_caus... The numbers here are too small and the data too limited to make that claim. Also search engine changes aren't instantaneous -- I've seen it takes days or weeks for changes found in a crawl of a small site to percolate through Google. Did Google even crawl the pages in question on the day Survey IO went live?
- vaksel 17y ago
  survey.io widget is site wide, so it was showing up on all pages on the site. Google crawls the site every day, so changes are more or less instantaneous. A lot of times, it takes like 7 minutes from the moment a question is posted, to the moment it shows in Google. But yeah I'm just as baffled as you, as to why it had such a big impact
mogston 17y ago
I don't believe that survey.io would have had that type of effect on your website traffic. We've been using it on http://technews.am http://technews.am for quite some time with no impact on visitor numbers.
marklubi 17y ago
Wow, if that inclusion is the reason then it's uncanny. On Monday I pushed a minor update to one of my websites that included the survey.io code and watched traffic coming from Google tank. Just rolled back that inclusion... we'll see if that was the reason for the traffic drop. Edit: quick clarification... the survey.io code was the only change that would be showing up to spiders. The other updates included in the change were all within areas that require authentication
hshah 17y ago
Hi, This is Hiten from KISSmetrics / Survey.io - We do not know what is going on with this and have not heard of this issue before. We're taking a look at this issue right now. At the moment, we think it might have to do with the Google Analytics code on the survey.io embeds, so we will be removing the GA code from them shortly. If anyone has any feedback or ideas please feel free to shoot me an email: hshah at kissmetrics.com
- hshah 17y ago
  The Google Analytics code has now been removed from the embed. I took a quick look at the stats for our own ProductPlanner.com and didn't notice any dips as a result of the survey.io code. If anyone has any more details or ideas, please let me know.
ahi 17y ago
This may have nothing to do with with Survey.io, but I've noticed some weird survey behavior in the past week. When I click through some newly opened tabs, occasionally there is a large blank survey page up, like the javascript choked on the survey and never managed to get to the requested page. There's no obvious way to click through to the actual site. Reloading the page usually gets it to go away. I have not been paying close enough attention to say whether the survey was from survey.io.
patio11 17y ago
My first advice for any perceived SEO issue: Take a deep breath. You're probably seeing random variation. (Incidentally, I don't know what time zone you are in relative to your users, but my first thought would be "A sharp drop that lasts two days and then abruptly reverses, hmm, was either of those days Saturday?" And indeed, it looks like it was.)

15 points by vaksel 17y ago

http://en.wikipedia.org/wiki/Correlation_does_not_imply_causation http://en.wikipedia.org/wiki/Correlation_does_not_imply_caus... The numbers here are too small and the data too limited to make that claim. Also search engine changes aren't instantaneous -- I've seen it takes days or weeks for changes found in a crawl of a small site to percolate through Google. Did Google even crawl the pages in question on the day Survey IO went live?

vaksel 17y ago
survey.io widget is site wide, so it was showing up on all pages on the site. Google crawls the site every day, so changes are more or less instantaneous. A lot of times, it takes like 7 minutes from the moment a question is posted, to the moment it shows in Google. But yeah I'm just as baffled as you, as to why it had such a big impact

mogston 17y ago

I don't believe that survey.io would have had that type of effect on your website traffic. We've been using it on http://technews.am http://technews.am for quite some time with no impact on visitor numbers.

marklubi 17y ago

Wow, if that inclusion is the reason then it's uncanny. On Monday I pushed a minor update to one of my websites that included the survey.io code and watched traffic coming from Google tank. Just rolled back that inclusion... we'll see if that was the reason for the traffic drop. Edit: quick clarification... the survey.io code was the only change that would be showing up to spiders. The other updates included in the change were all within areas that require authentication

hshah 17y ago

Hi, This is Hiten from KISSmetrics / Survey.io - We do not know what is going on with this and have not heard of this issue before. We're taking a look at this issue right now. At the moment, we think it might have to do with the Google Analytics code on the survey.io embeds, so we will be removing the GA code from them shortly. If anyone has any feedback or ideas please feel free to shoot me an email: hshah at kissmetrics.com

hshah 17y ago
The Google Analytics code has now been removed from the embed. I took a quick look at the stats for our own ProductPlanner.com and didn't notice any dips as a result of the survey.io code. If anyone has any more details or ideas, please let me know.

ahi 17y ago

This may have nothing to do with with Survey.io, but I've noticed some weird survey behavior in the past week. When I click through some newly opened tabs, occasionally there is a large blank survey page up, like the javascript choked on the survey and never managed to get to the requested page. There's no obvious way to click through to the actual site. Reloading the page usually gets it to go away. I have not been paying close enough attention to say whether the survey was from survey.io.

patio11 17y ago

My first advice for any perceived SEO issue: Take a deep breath. You're probably seeing random variation. (Incidentally, I don't know what time zone you are in relative to your users, but my first thought would be "A sharp drop that lasts two days and then abruptly reverses, hmm, was either of those days Saturday?" And indeed, it looks like it was.)