|Evidence Google normalizes traffic to network of sites?|
My Australian client site sells home products. They have several domains:
Group A – one domain with an online store with huge number of longtail keyword optimised pages as well as location specific keywords competing with other local stores.
Group B - 10 separate domains targeting specific towns in the area (their content very similar to each other but not similar to the main sit above)
Group C – 1 domain that is a vertical directory listing all stores in the area including competitors in the area.
All the domains are on the same IP address.
Group B & C Domains have links pointing to Group A domain but nowhere else.
There is no google ads on any of the sites
There is no google analytics on any of the sites
When looking at traffic from google to each site everything looks normal, but when combined I notices an interesting observation over the last 8 weeks.
Wednesdays 134 visitors +/- 1
Thursdays 143 visitors +/- 1
Fridays 108 visitors +/- 1
Saturdays 120 visitors +/- 2
The consistent and predictable numbers on set days is suspicious. Or have I got SEO fever?
It seems google is using traffic normization and lowering the organic traffic on one domain when sending additional traffic to another domain in the network presumably by excluding or lowering the SERP ranking.
Has anyone experienced this before?
If theory is correct then for local businesses, SEO is more about proving to google that you deserve more traffic than your competitors than on/off optimisation.
If you’re an expert or see anything I’ve missed, Id be very interested in hearing your theory.
This might be aligned with the ideas in the Traffic Shaping thread [webmasterworld.com]. You're certainly not the first to spot "suspicious" signs of this nature.
Can you break down the data by the hours of each data? In other words, is there a traffic cut-off after the target number is reached? Or is the traffic distributed in an expected curve throughout the whole 24 hours?
wow this is interesting...traffic shaping for a set of sites sharing the same ip?
Yes, the next step would be to see how the traffic shapes for the different sites during different times of the day.
|The consistent and predictable numbers on set days is suspicious. |
I see it too. Post Panda I can predict my traffic on any day to +/- 5%. The patterns are predictable. On days when I'm making huge money and the day looks strong I fizzle and come in as expected. I'm sure Google is shaping. Any ideas on breaking the pattern?
Question to those who are seeing shaping... were your sites penalized by Panda? If so that might be part of the Panda penalty.
For one of my pandalized sites, Google Analytics stats show 165, 166, and 167 visitors daily for May 4, 5, and 6th from US Google organic search. That's weird, as well as stats from WMT Top Pages section showing exact number of 1,600 clicks/day globally for 15 days out of the last 30 days. I wasn't observing these stats before Panda update, so I don't know if it's "normal".
P.S. Google organic search data in Analytics, when filtered with Country/Teritory, for April is not available (shows 0 visits).
Here's an interesting observation I made while looking through the analytics of my pandalized site:
Total number of traffic driving keywords for mon-tues this week: 5,863
Total number of traffic driving keywords for mon-tues two weeks ago: 5,861
Total number of traffic driving keywords for mon-tues four weeks ago: 5,863
Coincidence? Possibly, but it's at least a LITTLE strange...
(I also note a big black hole for April when using any advanced segments, so I can't filter this for just Google keywords)
|were your sites penalized by Panda? |
Yes, the site that I speak of was penalized by Panda - 60% loss in traffic. Super quality site with tons of authority. Probably over optimized on heavy KWs. Even before Panda the traffic patters were very regular. 2MM unique/mo.