Check out our new blog post "3 Quick Steps to Set Up Email Reports in the New Google Analytics" at http://t.co/QFj2lH5L #seo #
Archive for the ‘Stats & Data’ Category
Quantitative Effects of the Penguin Police
Penguin sparked fear in many SEOers, with blog posts like Anatomy of a Disaster, Penguins, Pandas and Panic at the Zoo, and innumerable accounts of sites losing over half of their traffic. Penguin is Google’s latest search algorithm update, and is supposed to level the playing field – increasing ranking for websites that have great content but aren’t well-optimized for the search engines, and penalize websites with less great content that are over-optimized for the machines.
Facts dispel fear (or at worst justify it and indicate a path towards greener pastures), so I wanted to distill what we know to date with hard quantitative data. Surprisingly the answer is not-a-whole-lot, and what we do have comes from just a couple sources.
A chart produced by SEOmoz shows that “the impact of Penguin was immediate and substantial”, with over 3% of Top Ten Rankings in their analysis changing (versus the ~2.7% change caused by the Panda update on April 19th).
Untold quantities of keyboards have rendered the phrases key-word-stuffing and over-optimized-anchor-text in hypotheses about what Penguin is doing. However, MicrositeMasters had unique access to a fabulously large data set and produced the only set of charts I could find showing detailed effects of Penguin.
Findings & Action Items from MicrositeMasters
1. Keyword-Stuffing in Anchor Text: Only sites with over 60% of anchor texts containing “money” keywords matching incoming links were negatively affected by Penguin.
A money keyword generates a lot of search traffic. Guzman defines it as a high-volume search term or phrase with a Google rank of roughly 5-15. In other words, the Penguin is looking for you if over half of your traffic is driven by searches using high-volume keywords that exactly match your anchor texts.
Action Item: Unless your incoming link anchor texts were totally dominated by high-volume search keywords, this component of the Penguin Algorithm is not hurting your site.

2. URLs in Anchor Text: Penguin is going to hit you for stuffing URLs in your anchor texts, but not near as much as money-keyword stuffing (far right in chart below).
Action Item: Reduce your URL-stuffing, but take care of the keyword-stuffing First!

3. Relevant (Niche) Inbound Links: Having links from websites outside of your niche is okay, as long as you have some from relevant websites, too.
I made this chart by eye-balling the numbers in a pair of charts from the MicrositeMasters study. The chart shows two things:
i) It doesn’t matter if you have a high or low percent of inbound links from websites with content that is relevant to yours, as long as you have SOME quality, niche inbound links: websites with low percentage (10%) of same-niche inbound links were penalized at the same rate as sites with high percentage (60-100%) of relevant inbound links.
ii) It matters a LOT if you have NO relevant inbound links – the red bars show that about half of websites with 0% of relevant inbound links were penalized. Once you have even 10% of your links coming from quality, relevant sites though, Penguin isn’t penalizing you for having non-niche links too.
Action Item: Make sure you have some inbound links from quality, relevant websites.
That’s not easy – and it’s part of the point of Penguin: good content creation and targeted community development around your site is harder than typing a keyword into 20 anchor texts. Countless blog posts offer advice on building a quality community (inbound links from websites with good content related to your site). Here are a few tidbits from Brownrigg’s insights on generating a community for your website.
i) Post quality articles relevant to your niche (that people will want to repost on their site).
ii) Compile some of your articles into an ebook.
iii) Publish your articles on sites like digg and submit them to article directories.
Post-Penguin Reality
Penguin seems to have a fairly narrow target – keyword-over-optimized text anchors and disproportionate inbound links from irrelevant websites: in other words, it has effectively removed shortcuts for generating website traffic. In general, however, and despite the apocalyptic rhetoric out there, the new Sheriff should leave the world with a lot of well-deserved winners – people creating genuinely good content and generating traffic with honest “white-hat” methods that ultimately benefit the 2,267,233,742 people searching for content on the World Wide Web.
The Familiarity Effect: PPC vs Organic Search
Dave Underwood of Topspot Internet Marketing spoke on “Improving Website Measurement & Analytics” at the HiMA luncheon last week in Houston, Texas.
He offered useful insight for successful marketing, some of it surprisingly simple. For example, recording and listening to phone calls: if your sales people give inaccurate information or fail in some other way to appropriately respond to potential customers, it really doesn’t matter if your marketing campaign generates 100 leads from Fortune 500 companies – the sales people are not converting the leads.
He also touched on the hot topic of Organic vs. Paid advertising. Agreeing with a study by Google showing that 89% of the clicks generated by paid ads would not be generated by organic search*, he recommends maintaining a PPC campaign even when your Organic rank is high.

The idea is explained by the basic psychological principle that humans prefer familiar things – the more times customers see or hear about a product or company the more likely they are to trust or prefer it.
Think of Billboard charts: the average song increases in popularity for about 6 or 7 weeks, meaning that people like it more as they become more familiar with it. For brands and companies familiarity drives brand equity: the recession resulted in a 30% earnings decline for companies overall, whereas familiar, trusted Best Global Brands saw only a 4% decline in earnings.
In other words:
People need to see your company lots of times, in lots of places (PPC and search lists), and they need to keep seeing your company show up over time.
* Clearly Google is incentivized to present their findings in a way that encourages companies to pay for advertising, and indeed analysts point out the limitations of Google’s interpretation of their results (Google presents an average rate of Incremental Ad Clicks unique to PPC, and doesn’t explicitly talk about the fact that lower IAC means your PPC is “cannibalizing” more of your organic clicks). In general though, there’s over half a century of robust research demonstrating the familiarity effect, so shelling out for those paid ads is probably worth it for your company.
HOWTO: Use the Dimensions Tab to Improve Your PPC Campaigns
Guest Blogger: Laura Rodnitzky is the Director of Production for PPC Associates, a search engine marketing agency with offices in San Mateo and Chicago.
If you’re running paid search campaigns in Google AdWords, and you don’t already use the Dimensions tab, you’re missing out on a ton of useful data. Rolled out in mid-2010, the Dimensions tab is one of the most important components of the AdWords UI. It allows you to view data for an entire account or specific campaign(s) that can be used to better target your customers, decrease wasteful spend, and improve conversions. At PPC Associates, our Production team relies heavily on the Dimensions tab to pull detailed reports on campaign behavior across different time periods or geos, and to better understand how and where our ads are being shown on both the search and content networks.
The Dimensions tab is located on the right-hand side of the tabs list in the AdWords UI. If you can’t already see it, click on the drop-down arrow at the end of the row to bring up the menu of available tabs.
Once you’re in the Dimensions tab, go to the “View” drop-down menu to see the types of data available. The screenshot below shows the main menu; for Time, Conversions, and Reach and frequency you have additional options, such as Day of the week, Day (date in time), Week, Month, Quarter, Year, or Hour of day.
So now that you know where to find this, how are you going to use it? There’s a ton of good stuff here, and obviously not all of it will be relevant to every campaign. If you’re not using track-able phone numbers in your ads, for example, the “Call details” option is not going to have any data for you. If your campaigns are only running on the search network, you won’t have any automatic placements to review. But take advantage of what you can. Here are a couple of examples of how we use the Dimensions tab at PPC Associates:
1) Day parting. We usually run two types of day parting reports: day of the week and hour of day. You can pull the data separately, or you can use the advanced segmentation feature in AdWords to break it down by hour of day for each day of the week. Once you have the data in Excel, use conditional formatting to easily spot trends in campaign performance. This can then be used to optimize the campaigns; for example, you may choose to increase bids during time periods with high conversion rate and low CPA, or decrease bids when the opposite is true. If there’s a clear drop in performance during specific days or hours of day, you may even want to turn campaigns off during these low-performing time periods. In the sample data below, it’s clear that the hours of 5 am – 8 am do not perform well, whereas the hours of 4 pm to 8 pm have high conversion rates and low CPAs.
2) Search query review. Being able to see the actual queries that cause your ads to show is powerful, both for finding terms you don’t want to show up on and getting new ideas for keywords. We all know that running keywords on broad match – or even modified broad match or phrase match – opens up a campaign to a wide range of search queries, many of which might not be relevant to what we’re advertising. The “Search terms” option in the Dimensions tab will let you see the queries triggering your ads. Use this data to promote high-performing queries by turning them into keywords with targeted ads, and also to scrub out unwanted terms. This is especially important when your campaigns include a lot of general keywords. Just imagine how many irrelevant queries you can get matched to when bidding on “will” (as in last will and testament) keywords. The screenshot below shows just a few out of thousands.
Clearly this is a very broad overview of the Dimensions tab and the ways you can use it to optimize your campaigns. There are a lot of different ways to use the data, and the “Customize columns” and advanced segmentation options let you slice and dice the data in innumerable ways. No time like the present to get in, start some tests, and figure out how to improve your campaigns with the options on hand.
- Opinions expressed in the article are those of the guest author and not necessarily of Schipul – The Web Marketing Company.







