Ephricon Web Marketing

When Google first announced that they would no longer always provide keywords resulting in a natural search visit the SEO community panicked. Eventually Matt Cutts, head of Google search’s webspam efforts, said that this would only be a small amount, people started to calm down, but no one knew how much a “small amount” was.

Since then I have been closely monitoring the data for a few websites to see what Mr. Cutts meant by “small amount.” The first complete week after this was rolled out (the week starting October 23rd, 2011) seemed to confirm this claim with my clients- with the smallest percentage of “(not provided)” keywords being 0.6% of the total visits from Google natural search and the most being 3.7% of Google natural search visitors.

When it came to the second week after the roll-out (week of October 30th, 2011) it was clear that Google was gradually trickling this change throughout its users- with percentages of not provided keywords ranging from 4.0% to 21.6% of visits from Google natural search. Since the second week, these hi/lo percentages have continued to rise but have remained consistent over the last couple of weeks. This is leading me to conclude that the percentage of provided keywords from Google natural search is leveling-off, as you can see in the chart below:

Below is the largest (MAX), average (AVG) and smallest (MIN) percentages of not provided keywords from Google natural search visits over the last few weeks:

Important facts about this data:

1. The percentages mentioned above reflect a percentage of Google natural search visits- not a percentage of all natural search visits or a percentage of all visits to the website.
2. This is an average percentage of about 8 different websites, ranging in size from approximately 500 visits per week to 70,000 visits per week.
3. Most of these sites generate a majority of the visits from the United States (the first place Google is rolling-out this change) but two of these sites have a majority of visits from outside of the US). This means these percentages might grow as this change becomes more international.
4. These sites reflect a wide diversity of industries; half of these sites are business-to-business oriented and the other half are business-to-consumer oriented.
5. While the latest week of this data (the week starting December 18th, 2011) shows that on average 23% of the keywords were “not provided”, these websites ranged from 14.2% to 33.0% of the keywords that were not provided.

Are you seeing similar data in your website analytics? Do you think this is leveling-off or going to get larger?

Yesterday we talked about how Google is now merging the local listings and organic listings such that in many cases they are inseparable.  Similarly, we’re noticing several other interesting developments with regard to actual rankings of sites… not just what the listing looks like when it does rank.

Like most really cool discoveries I have in the world of SEO, I stumbled upon this (haha, “Stumbled Upon”) by looking at my analytics.   Our site produced 3 web-form leads in a 45 minute span yesterday which is well higher than the norm.  So this morning I was curious if they came from a particular referring site, a particular tweet, or a particular keyword.  I suspected they were not from organic search, since traffic from organic search tends to be more consistent over time.  Traffic from referring sites tends to be highly volatile – when a prominent blogger creates a new post and links to you you’ll see a ton of traffic for a day, maybe two, and then as that post drops down and is no longer the “new thing” you’ll see that referring traffic die down quickly.

(more…)

So we may have jumped the gun here as Google appears to be testing these changes to the SERPs, but if we waited until Google stopped testing things before commenting on them then we’d still be without Gmail accounts and most other Google products.

Korey and I share outline the recent changes and share our initial thoughts, below:

Anyhow, it appears that sometime yesterday Google started testing changes to how they show the “local” (map-driven) listings in the search results.  Specifically, they seem to be doing away with the “7-pack” or “one-box” format where a map is showing (typically with up to 7 listings) immediately above the organic results and below the top 3 paid results.  Instead, they are showing the organic results higher on the page, immediately below the top 3 paid results, but this time with the local listings and related info (reviews, link to the places page, map marker, physical address) all tied to the organic listing itself.

(more…)