5 ms·

Google searches for "VPN" are spiking this week in the US

25 points by zissou 14y ago

mey 14y ago
Yet drastically down compared over a larger timeline. https://www.google.com/trends/explore#q=VPN&geo=US&cmpt=q https://www.google.com/trends/explore#q=VPN&geo=US&c...
- hndude 14y ago
  My guess is that that gradual decline over time may be related to more people getting online. Put another way - when the internet's use was limited to more tech-savvy people, this term was searched a higher percentage of the time than today.
Pezmc 14y ago
SimCity is being released soon, VPN for getting access 3 days early if you're in Europe may be the case.
- nthitz 14y ago
  I think it's more to do with the USA's new six strikes program.
- eclipticplane 14y ago
  Crap, no. I have work to do...
nthitz 14y ago
Much more drastic are searches for seedboxes https://www.google.com/trends/explore#q=seedbox&geo=US&date=today%203-m&cmpt=q https://www.google.com/trends/explore#q=seedbox&geo=US&#... edit: for those that don't know, seedboxes are servers you rent that just run bittorrent. Makes it easy to skirt around these 6 strikes programs
ars 14y ago
No they aren't. Look at a longer graph. What you are seeing is just random noise. Flagged as incorrect.
- jasallen 14y ago
  It's not 'incorrect' -- clearly the flat trend has existed long enough for this 'spike' to represent something other than random noise. Would you say that the dips on weekends are 'random noise'? No, they too represent a trend, though if we zoom out far enough they would get lost in the lack of precision.
cjbprime 14y ago
Could be related to the Yahoo! policy change, since part of the story is that Marissa Mayer ran analytics on how their VPN was being used and concluded that many remote workers weren't working much: "Mayer explained the rationale at Yahoo’s “Friday FYI,” its equivalent of Google’s TGIF. ”We’ve checked and some people who work from home haven’t even logged into the VPN…” she apparently said."
lessnonymous 14y ago
So what's with the Dakotas and Wyoming? No interwebs? Edit: Missed Alaska and Vermont
antoko 14y ago
Great job on wild speculation HN! https://www.google.com/trends/explore#q=VPN&geo=US&date=1%2F2011%201m&cmpt=q https://www.google.com/trends/explore#q=VPN&geo=US&d...
- Diamons 14y ago
  That's 2011.
  - antoko 14y ago
    that's a 2 year timeline jan2011-jan2013 showing clearly that this latest "spike" isn't anything of the sort. EDIT my apologies, I was supposed to link this https://www.google.com/trends/explore#q=VPN&geo=US&date=1%2F2011%2025m&cmpt=q https://www.google.com/trends/explore#q=VPN&geo=US&d...
    - noecker 14y ago
      There actually is a decided uptick in vpn activity if you expand the date selection correctly. It is a decided uptick; the highest weekly search volume in the past 2 years. https://www.google.com/trends/explore#q=VPN&geo=US&date=1%2F2011%2027m&cmpt=q https://www.google.com/trends/explore#q=VPN&geo=US&d...
codezero 14y ago
This is probably related to the Copyright Alert System, or "Six Strikes" rule that many ISPs have adopted. VPN is the best way around this and it costs next to nothing.
zissou 14y ago
OP here. Should have posted sooner before it looks like my post got flagged? I think the shock[1] is being caused by the Copyright Alert System put into place on Monday by Comcast. Here's why: Comcast is the largest ISP in the US, so a change of their piracy policy (the exogenous shock in my story/model) is likely provide a large enough segment of Internet users (definitely bigger than the set of Yahoo employees or SimCity fans) with the appropriate incentives to search for information about VPNs. The logic here is simple. [1]There is clear weekly periodicity in the 90 day chart. If you were to detrend the data and remove the autocorrelation, the spike would look even sharper. Based on the data and the normalizations on the data that Google does, 90 day data is much better than 12 month data because the latter is not daily data anymore -- it is data aggregated weekly. Since it is obviously clear that the data has weekly periodicity, aggregating the data at the weekly-level ignores the fact that how many searches there are for "VPN" on a given day depends on what day of the week it is. Calling this wild speculation, incorrect, random noise, or comparing it to weekly or monthly data is just not very good statistics in my mind. EDIT: Also, for clarification Google Trends data is search volume data normalized on an index from 0 to 100 where 100 is the relative maximum over the series. Tuesday's index value was 100, with Monday's being the second highest over the 90 day period. So technically, yes, search volumes for VPN are currently spiking this week. I'm really curious to see what happens when Wednesday's data is released.

25 points by zissou 14y ago

mey 14y ago

Yet drastically down compared over a larger timeline. https://www.google.com/trends/explore#q=VPN&geo=US&cmpt=q https://www.google.com/trends/explore#q=VPN&geo=US&c...

hndude 14y ago
My guess is that that gradual decline over time may be related to more people getting online. Put another way - when the internet's use was limited to more tech-savvy people, this term was searched a higher percentage of the time than today.

Pezmc 14y ago

SimCity is being released soon, VPN for getting access 3 days early if you're in Europe may be the case.

nthitz 14y ago
I think it's more to do with the USA's new six strikes program.
eclipticplane 14y ago
Crap, no. I have work to do...

nthitz 14y ago

Much more drastic are searches for seedboxes https://www.google.com/trends/explore#q=seedbox&geo=US&date=today%203-m&cmpt=q https://www.google.com/trends/explore#q=seedbox&geo=US&#... edit: for those that don't know, seedboxes are servers you rent that just run bittorrent. Makes it easy to skirt around these 6 strikes programs

ars 14y ago

No they aren't. Look at a longer graph. What you are seeing is just random noise. Flagged as incorrect.

jasallen 14y ago
It's not 'incorrect' -- clearly the flat trend has existed long enough for this 'spike' to represent something other than random noise. Would you say that the dips on weekends are 'random noise'? No, they too represent a trend, though if we zoom out far enough they would get lost in the lack of precision.

cjbprime 14y ago

Could be related to the Yahoo! policy change, since part of the story is that Marissa Mayer ran analytics on how their VPN was being used and concluded that many remote workers weren't working much: "Mayer explained the rationale at Yahoo’s “Friday FYI,” its equivalent of Google’s TGIF. ”We’ve checked and some people who work from home haven’t even logged into the VPN…” she apparently said."

lessnonymous 14y ago

So what's with the Dakotas and Wyoming? No interwebs? Edit: Missed Alaska and Vermont

antoko 14y ago

Great job on wild speculation HN! https://www.google.com/trends/explore#q=VPN&geo=US&date=1%2F2011%201m&cmpt=q https://www.google.com/trends/explore#q=VPN&geo=US&d...

Diamons 14y ago
That's 2011.
- antoko 14y ago
  that's a 2 year timeline jan2011-jan2013 showing clearly that this latest "spike" isn't anything of the sort. EDIT my apologies, I was supposed to link this https://www.google.com/trends/explore#q=VPN&geo=US&date=1%2F2011%2025m&cmpt=q https://www.google.com/trends/explore#q=VPN&geo=US&d...
  - noecker 14y ago
    There actually is a decided uptick in vpn activity if you expand the date selection correctly. It is a decided uptick; the highest weekly search volume in the past 2 years. https://www.google.com/trends/explore#q=VPN&geo=US&date=1%2F2011%2027m&cmpt=q https://www.google.com/trends/explore#q=VPN&geo=US&d...

codezero 14y ago

This is probably related to the Copyright Alert System, or "Six Strikes" rule that many ISPs have adopted. VPN is the best way around this and it costs next to nothing.

zissou 14y ago

OP here. Should have posted sooner before it looks like my post got flagged? I think the shock[1] is being caused by the Copyright Alert System put into place on Monday by Comcast. Here's why: Comcast is the largest ISP in the US, so a change of their piracy policy (the exogenous shock in my story/model) is likely provide a large enough segment of Internet users (definitely bigger than the set of Yahoo employees or SimCity fans) with the appropriate incentives to search for information about VPNs. The logic here is simple. [1]There is clear weekly periodicity in the 90 day chart. If you were to detrend the data and remove the autocorrelation, the spike would look even sharper. Based on the data and the normalizations on the data that Google does, 90 day data is much better than 12 month data because the latter is not daily data anymore -- it is data aggregated weekly. Since it is obviously clear that the data has weekly periodicity, aggregating the data at the weekly-level ignores the fact that how many searches there are for "VPN" on a given day depends on what day of the week it is. Calling this wild speculation, incorrect, random noise, or comparing it to weekly or monthly data is just not very good statistics in my mind. EDIT: Also, for clarification Google Trends data is search volume data normalized on an index from 0 to 100 where 100 is the relative maximum over the series. Tuesday's index value was 100, with Monday's being the second highest over the 90 day period. So technically, yes, search volumes for VPN are currently spiking this week. I'm really curious to see what happens when Wednesday's data is released.