Thank you for the response and the suggestion on the points that near clarification.
1) The statistics are for landing pages, so when I said that ecommerce pages were up 40% starting on May 20th, 2014, I mean specifically LANDINGS on those pages were up Using google analytics I am looking at the:
Behavior > Site Content > Landing Pages
To source the data for both the INCREASE and the DECREASE in traffic - for both the ecommerce pages and the blog pages.
Were the queries general or precise?
Hard to say. They were along the lines of "blue widgets" . They could be interpreted as either an ecommerce keyword (as if someone were looking to buy blue widgets), or a general info query.
But none of the top 200 listed keywords contain shopping-specific keywords such as "buy" or "shop" or even "cheap" or "inexpensive" or anything like that. Just a general "blue widgets"
There were only a couple of keywords that contained an identifiable product brand, such as "acme blue widgets" or "brand x orange widgets"
Is your blog targeting people who might respond to what your ecommerce pages are offering?
Yes and no:
The blog offers information and history about the various widgets, so it was generally informational. But it did / does have LINKS to the ecommerce pages. The click-through rate from the blog to the ecommerce pages is / was very low, and the conversion rate (turning a blog visitor into a product purchaser) is / was abysmal.
The blog is geared toward the long tail: "where to place your blue widgets?" and "What is the meaning of different widgets? are examples of what the blog targets.
But just to reiterate, the traffic increase and traffic decrease I cited above are (as best as I can tell) direct LANDINGS on the ecommerce pages.
Were the blog links sending visitors to page types other than ecommerce?
I am not sure I understand this. Maybe you can clarify for me?
Bounce rate was high (around 80% for blog pages, compared to only 30% for those who landed on ecommerce pages).
Visitors would see a small box on the right with text links to "Related Articles" and then below that links to the ecommerce pages.
What kind of data do you have to segment your traffic, and to examine it over time? Etc etc?
Really, I just have google analytics.
It may be that Google observed user behavior over time and decided that they weren't sending the right kind of traffic to your blog...
That's the thing though.
Landings on the BLOG continued upwards after September 9th, 2014. Landings on the blog were up 125% over the previous period from May 20th to September 8th, and then from September 9th through December 29th they were up an additional 20% over the previous period.
So google seems to not have a problem with the blog at all despite the near 80% bounce rate. (When I was using event tracking on the blog, I found that 75% or more were reading to the end of the content on the blog pages. They just weren't going on to look at other blog info pages or going to the linked ecommerce pages).
IMO, Google adjustments are designed to help Google searchers find what they're looking for as fast as they can...
As my ecommerce landing pages had only a 29% bounce rate, and they showed an ecommerce conversion rate of .55% (which I know is pretty low), I am guessing that google had a fairly good idea when someone had an ecommerce "intention" in mind, even though their query might have only been "blue widgets".