Was it random or organized?

[kexej28769@nongnue]

Aside from minor variations, the pattern of dominant intent URLs appears to be very similar to the previous analysis. It appears that the impact of de-indexing was widespread.

It's hard to determine whether this bug was random, affecting all sites somewhat equally, or somehow systematic. It's possible that limiting our analysis to "stable" URLs is skewing the results. On the benin number data hand, trying to measure the instability of inherently unstable URLs is a bit naive. I should also note that the MozCast dataset is biased toward so-called "head" terms. It doesn't include many queries in the very long tail, including natural language queries.

One question we can answer is whether large sites were affected by the bug. The graph below separates our “Big 3” in MozCast: Wikipedia, Amazon, and Facebook. This narrowed us down to 2,454 stable URLs. Unfortunately, the deeper we dive, the smaller the data set becomes:




At the same 90-100% zoom-in scale, you can see that the impact was less than for all stable URLs, but there is still a clear pair of dips on April 5th and April 7th. It doesn't seem like these mega-sites were immune.