|
|
|
|
|
|
| Who benefited from the Google bug? |
|
| In our previous study dated 2nd February, we had analyzed the impact of Google traffic loss on Saturday Jan 31st. A system error left millions of visitors puzzled when links to all search results were flagged with the warning 'This site may harm your computer'. The breakdown resulted in a 90% drop of Google visits during peak time around 4:04 PM. Traffic started to come back to normal around 4:20PM. We estimate the total duration of the breakdown to be around 50 min. Apart from inconvenience experienced by Google users, what is the real impact of this incident on global internet traffic? |
| Cross section: |
- Study performed on Jan 31st 2009
- Cross-section of 54689 websites
|
|
| Google breakdown results in a loss of 21% of the traffic during 50 min with a peak of 29.3%… |
The graph displayed below represents internet traffic evolution during the Google breakdown on Jan 31st between 3:29 and 4:20 PM. In order to manage this study, we have brought the number of visits coming from Google searches to 100 visits at 3:00 pm (just before the break down) on Saturday Jan 31st, amongst sites audited by our AT Internet solutions. We can clearly observe a significant drop in traffic during this period. We estimate the average global internet traffic loss to be around 21.3% with a peak value of 29.3% (around 3:43PM). Google visits dropped by 71.4% during the whole period, with a peak value of 91.2% (around 4:04PM). At the end of the breakdown we also noticed a slight traffic increase compared to the average before the bug. The traffic rise only lasted between 4:21 and 4:27PM. The return to normal was around 4:30PM.
|
| … resulting in an increase of direct accesses and other search engines, Yahoo! first |
As a result of this breakdown, did Google visitors use other search engines / methods to search the web? The graph below tells us that the answer is yes: an increase in visits from other search engines or direct accesses (direct entry of the URL, bookmarks/favorites or e-mail via messaging software like Outlook) corresponds to the period in which Google broke down.
- The graph below uses 100 as a baseline for Google visits, at 3pm on Jan 31st.
- Visits from other search engines are at a baseline of 10, at 3pm. During the peak of the Google breakdown, such visits enjoy an increase to approximately 23.
- Visits from direct accesses are at a baseline of 124 at 3pm, peaking at 139 during the Google breakdown.
|
The transfer of visits lost by Google as a result of the breakdown, between 3:29 and 5:00 PM, is estimated as follows:
- 16.2% for direct access visits. In this case, Google is only used as a navigator bar because internet users have accessed websites through bookmarks or typing URL.
- 13.9% for visits via other search engines.
As for the remaining 70% of visits, we surmise that the visitors simply abandoned their search.
|
|
Among search engines benefiting from the transfer of Google visits, Yahoo! is the leader at 41.7%, followed by Live Search (24.7%) and Orange (13%). Yahoo! remains an alternative in internet users mind.
It is important to note the irony of this breakdown: Google could be considered a benefactor from a financial point of view. During the breakdown Google sponsored links were still functioning, thereby proposing an alternative on which Google is remunerated.
We must observe that the breakdown Google affected traffic worldwide.
So, who profited from the Google bug? Perhaps nobody except Google as a result of expected higher sponsor link activity; as the increase in alternatives (direct accesses or other search engines) is relatively small. Is Google deeply rooted in users’ minds that they can’t search differently? In contrast, the quasi-monopoly which Google enjoys in the global internet economy is certainly brought to light. This evokes a discussion regarding the political game regarding the internet, the ways we access it and use it: does the power of the internet not reside in the absence of a centrally controlling force? Do the means to access post office, our local City Hall, Shopping malls be the sole responsibility of public services?
|
Methodology
The study was done on the global traffic generated by the AT Internet audited sites at recording times (GMT + 01:00) between 3:00 pm and 5:00 pm.
|
|
|
|
|
|
|
|