Ramblings About SEO

The interview for this week is with Yahoo!’s Dennis Mortensen. Now that IndexTools has been part of Yahoo! for more than a year, and the product is now called Yahoo! Web Analytics, I thought it would be great to catch up with Dennis again. One of the very cool things that now makes up the current product is integrated demographic data. Yahoo! Web Analytics now integrates data from the rest of the Yahoo! network of web sites, so site owners can get a view into the age, gender, types of interests, and other Yahoo! properties that they visit.

Now that is significant stuff! To learn more, check out the interview. To see musings from Dennis on the world of analytics you can visit the Visualrevenue blog.

I recently spoke with Richard Zwicky about the Enquisite and the newly emerging field of search analytics. Search analytics is a close cousin to web analytics, but with some key differences. Both use a Javascript tag to collect data about what is taking place on your site, and both use that data to build reports and analyze the data.

Where the difference comes in is that search analytics are focused specifically on the needs of search marketers, and this specialized focus allows it to go much deeper in helping search marketers analyze and optimize their SEO of PPC campaigns. Check out the interview if you want to learn more.

Sometime last week Google introduced a fascinating new feature into one of the Local Business Center accounts I manage for a client.  I haven’t seen anything written about this among the Local blogger community, or on the Google blogs, so this appears to be a bit of a stealth feature that Google is testing quietly.

The following links showed up in this GLBC account (and notably not in any others I use) around the middle of last week:

Clicking on one of the “View Report” links leads to a detailed set of statistics. To keep the identify of my client private, I’ve sanitized the report and broken it into several pieces. The first, and possibly most interesting, piece is the Activity report, which shows “Impressions” and “Actions” for this particular listing graphed over time:

You can float over the data points in the chart and get a small information “bubble” displayed on the chart showing you the date and the data value for the point. You can adjust the time scale — either using the pre-defined past 7-day or past 30-day window, or using your own custom date range.

Google on-page help defines impressions as follows: “We add 1 to your total count of impressions each time your business listing is shown as a local search result on Google or Google Maps”. So impressions count both 1, 3, and 10-packs shown as part of Universal Search Results, as well as searches performed on maps.google.com.

Actions include:

• Clicks for More Info
• Clicks for Driving Directions
• Clicks through to the Web Site

Overall, Google is providing several major improvements to the very simple clicks and impressions data it has provided in the past: historical trending, and a breakout of the kind of clicks/actions taken by users.   I’m particularly pleased with the historical trending, as this should allow one to carefully monitor the performance of listings over time based on optimization efforts, seasonality changes, market changes, etc.

We can also begin to understand actions taken from click-throughs at a much finer level of granularity.  Clearly clicks through to the web site are very desirable, but we can also begin to understand our geographic market by looking at the volume and distribution of requests for driving directions.

Indeed, Google is providing a wealth of useful information in the bottom section of the report, labeled “Where driving directions requests come from”, but we’ll get back to that in a moment.

Visually, the next two sections of the report (again, I’ve broken this up for formatting and discussion, but all of these sections appear on a single integrated web page for each location in the GLBC), appearing just below the Activity and Totals section, contain information about keywords driving impressions and the driving directions section.  These sections are shown below:

In language similar to the top queries report found in Google Webmaster Tools, Google defines “Top search queries” as: “The top Google search queries for which your business listing appeared, along with the number of times users saw your business listing in the search results for those queries.”. In the screen shot above, I’ve sanitized the search queries, but envision this as a list of 10 keyword phrases.  Next to each phrase is the number of impressions that phrase drew in local queries (i.e., impressions, as defined above), along with a horizontal bar proportional to this value. This, of course, is quite valuable keyword intelligence.

Finally, below that, is yet more business intelligence in the section titled “Where driving directions requests come from”. Here, you can see what appear to be a count, aggregated by zip code, of the location of users requesting directions. In a nice bit of Google Maps eye candy, the city/zip phrase in the ranked list turns out to be a link that, when clicked on, causes the map to pan and display the region containing the zip code. Further, when you float your mouse over the map marker with the count number displayed for that zip code on the map, Google visually highlights the zip code. Altogether, this is a stunning little bit of wizardry that would actually seem pretty useful for visualizing your geographic market.

In addition to the statistical data, the report page includes a nice pane displaying most all of your data - much like the main data entry page in the “Add new listing” (or “Edit”) panel of the GLBC. I’ve omitted this pane as, with one exception, there’s nothing new here and since it has so much of my client’s identifiable data it was difficult to sanitize. However, there was one intriguing tidbit worth pointing out. Above the info pane on the right, the following indicator appeared on my listings:

Note the “86% complete” indicator. I’m a little unsure how this is being calculated. For this particular listing, the only GLBC data we haven’t provided is a “Mobile phone”, “Fax” and “TTY/TDD” phone number. We’ve included everything else, including photos, videos, hours of operation, categories, and “Additional Details” (i.e., custom attributes). I’m pretty curious whether Google is asserting that the lack of the 3 phone numbers above is what constitutes my “14% missing data”, or if there’s something else I’m missing (unrealized opportunities?!?!). Guess I can test this and report back.

Well, there’s lots more analysis and discussion of the various data elements of the report, but in the interest of getting this out, and getting a bit more insight into the mystery of “how widespread is the ‘preview’ of this feature?”, I’m going to go ahead and close out for today. I’d love to get peoples’ comments and questions, and if my client’s account is, indeed, a rarity at the moment, I’d be glad to provide more observations and feedback on what I’m seeing.

Imagine you want to establish yourself as a true leader in your market space. However, the market space has been around for a few years, and there are other sites that have already established themselves as leaders. A natural thing to do is to backlink those leaders (i.e. use a tools such as LinkScape to get a map of who links to your competitor).

The next thing that people often do is to start contacting people at the best sites (with the most link juice) that link to the competitor. No question that this is a good idea, but the analysis should not stop there. Here are the two major reasons why:

1. The first problem is that even if you proceed to contact everyone that links to your competitor you are likely to only get about 10% of those people to link to you (10% is actually a very good result). However, we have already stipulated that you want to be a leader, not 10% of a leader. So while this is a good strategy, it should not be the only strategy. You need a strategy that will help you catch up with (and pass) the competitor.
2. Most people recognize that contacting every single site that links to a competitor will get a little bit tedious. So they target the obvious high end links. This is OK, but there is a real danger that your links will become one dimensional. What I am getting at here is that your competitor defines your challenge by more than just their top links. The scope and breadth of those links is equally important to understand.

Expanding upon the 2nd point above, the search engines already know how to assess the relevance of a link. They also recognize sub-sectors within each market space, such as news sites, blog sites, sites selling related commercial products, academic sites, industry organizations, not to mention the scores of categories of sites that are tangentially related to the space, and therefore refer to link to sites in it from time to time.

A true leader in a market space will most likely have links from all of these sectors. A single focused link building campaign will not give you this type of breadth. To think about it in non-SEO terms (or in a non-search engine world), a single successful marketing campaign does not make you a broadly accepted leader in a space.

You need breadth too. You need to implement campaigns that will touch all of the various segments of the market. Note that you don’t have to do these all at the same time, but over time you want to touch as many different relevant segments as possible.

One great way to do this is with PR related strategies (there are other ways to achieve breadth in links other than via PR, we are just using it for an example here). Create some truly unique content or tool, and then work hard to let the world know about it using PR. Fuel the fire by implementing more great content and tools. This may seem like a hard road, but you can’t get to a leadership position in a market space without doing these types of things.

But, the focus of this post is not to tell you to build great content and promote it well (although you should do that). The focus is on putting together a map of what defines breadth for your market space and implementing a link building and promotional strategy to achieve breadth in your link building. To help do this, there are two questions you should ask:

1. What does a well distributed link graph in your market space look like?
2. How did your market leading competitors achieve their breadth, i.e., what were there promotional strategies?

To answer these questions you will need to do more than look at the best links your competitor has received. You will need to map out a broader picture of different market sectors that have linked to them along the way. I don’t have a simple formula for doing this, other than getting detailed backlink reports on your leading competitors, and analyzing hundreds of the links they have and classifying them. What categories to use will vary for each market, but you can certainly start with the list of categories I mentioned above.

The next step will be to figure out what types of content, tools, and marketing you will need to build a similarly broad profile. With this plan in hand you will be well on your way to being able to design a market leadership strategy for your business.

This week I am publishing a recent interview with Microsoft’s Dr. Scott Prevost. Scott is the principal development manager in the Powerset group at Microsoft. I used this discussion to get some insight into what Powerset is about, and how Microsoft intends to use it as a part of their core search technology in the future.

Powerset focuses on natural language search, which is a concept that certainly has a lot of potential. At the moment, the amount of this technology that has been integrated into Live Search is realtively small. You can see a test platform running the Powerset technology on Wikipedia at http://www.powerset.com.

One of my recent interviews was with Tracy Chan and Matthew Liu of YouTube. Some of the areas discussed include:

• YouTube Insight’s Popularity feature
• YouTube Insight’s Discovery functionality
• YouTube Insight’s Demographics tool
• The ability to export data from YouTube Insight
• How Promoted Videos can help you launch a video
• Other tactics for driving video views
• YouTube video optimization basics

All in all, a pretty good primer for learning about how to get started with YouTube.

This week I get to post an interview with well known Googler John Mueller. John works in the Google Zurich office and has a lot of insight into the inner workings of the Google algorithms as well as their webmaster guidelines and policies.

Our interview was laser focused on the issue of cloaking, as well as First Click Free. We discuss many different aspects of cloaking, such as:

• Replace bad URLs (with session IDs and superflous URLs) with SE friendly URLS.
• Multivariate testing
• Rendering content within a graphic to a search engine in text, including sIFR.
• Fix up complex javascript/coding, or Flash messes, including SWFObject.
• Language/country/region cloaking
• Serving up different content based on cookies.

Check it out!

One of the things that is evolving in my philosophy of SEO is how I look at the role of trust. Trust was not something that was important during AltaVista’s hey day, when keyword density was king, or even during the early days of Google when PageRank in its purest form ruled the day.

Another big thing in my mind these days is that the number of factors, and the importance of each factor, involved in ranking algorithms has changed. When the PageRank paper was published, you pretty much had a blueprint for how it all worked. However, knowledge is power, and in this case the power was in the hands of the spammers.

As a result of these factors, numerous patents have been published by each major search engine, on a variety of topics related to ranking, yet these patents no longer provide a clear roadmap to ranking algorithms. They provide hints as to what the search engines could choose to use a ranking signals, but they don’t tell us what they do use. For this we have to rely on intuition, judgment, and testing. In evaluating ranking signals I believe there are two major factors:

1. Noisiness of the signal. Does a strong positive always, or nearly always, mean a good, relevant site? Does a strong negative always, or nearly always, mean a poor, or less relevant site?
2. Importance of the signal. Assuming that we have a signal that is not noisy, how significant an indicator is it when compared to other signals? What made links such a powerful element is that they were, and still are, a powerful indicator of relevance and quality.

One example of a noisy signal is Bounce Rate. In principle, the idea is that when a user goes to a site, and returns to the SERPS after a relatively short period of time, that this is an indicator that the result was not a good one. But the problem with it is that on a reference search (e.g. zip code for Charlotte) the user may have gotten what they want in just a few seconds.

One of the signals that I think has low noisiness and a high degree of importance is trust. One important paper on this topic was published in 2004 by Yahoo! and Stanford University. The paper was titled Combatting WebSpam with TrustRank. The paper proposes that the search engines use human editors to identify a site of highly trusted seed pages. Then, “once we manually identify the reputable seed pages, we use the link structure of the web to discover other pages that are likely to be good”.

The general notion is that the closer a web page is to a highly trusted page (closer as mentioned in number of link hops) the most likely it is to be a trustworthy page. You can think of a Trust Rank factor that reduces the overall trust level of a page based on the number of hops from the human reviewed seed sites. The paper also suggests that links placed on pages with lots of links (even if they are on one of the human selected seed pages) tend to be placed with less care than links on pages with very few links. As a result the trust communicated by two selected seed pages can differ.

The researchers who wrote the paper on TrustRank also authored an interesting paper on a concept they call spam mass. This paper lays out a method for calculating the percentage of a web sites total PageRank that results from being linked to by spam pages. The higher this ratio is the greater the likelihood that the site is itself a spam site. There are some obvious problems with this idea - as your competitor could buy links to your site from thousands of poor quality sites and potentialls trash your rankings by giving you a high spam mass. Nonetheless, the concept is an interesting one.

Once you start thinking about these things, it is easy to come up with fresh ideas. For example, the trust damping factor you might apply for each link between a given page and the highly trusted seed pages could vary depending on the TrustRank level of the intervening web pages. It could also vary depending on the trust level of the domains on which those pages reside. For example, the home page of http://www.usa.gov may be deemed to be a highly trusted seed page.

In three hops you may be able to find yourself on links fairly well removed from the seed page. But if those three hops take you to a different page on USA.gov, do you lower the TrustRank as much as you would if you have transitioned to a completely different domain? Maybe not. You can also think about the notion of “Reverse TrustRank”. This is the notion that if your site links to spammy sites that this should lower its TrustRank. This thought should provide ample motivation to make sure that you take care to not link to any bad sites. Better still, you should screen your site for this on a regular basis. After all, the quality domain you link to today may have a different owner tomorrow, and that new owner may have poor intentions. Don’t let this happen to you!

Summary

So we don’t have as clear a roadmap to ranking factors as we used to, but we can still use common sense. Take the time to learn the basic tactics that search engines can use to do their job. Don’t get overly hungup on any one factor. Most of the time, if you build a great site and promote it properly you should achieve good results. But, knowing what things search engines are likely to rely on can really help you understand what you need to do to improve your rankings.

In addition, the way the search engines use and measure trust most likely varies significantly from the papers reference above, but my own experimentation convinces me that they are measuring it, and using it in a significant way. We factor this into our thinking at STC on a day to day basis.

After a brief break from posting regular interviews, I am back at it! This week I have a great one, with Sarah Bird of SEOmoz. As a general counsel, Sarah handles a variety of different legal issues, and there are some issues that face both those who work as SEO professionals and the companies that engage them. For example, we talk a bit about performance based SEO contracts, fair use, DMCA requests and more. Good stuff!

Jakob Nielsen recently put up a nice article about link text, titled First 2 Words: A Signal for the Scanning Eye. In it, he covers some guidelines for creating anchor text, reviews some specific examples, and reports on usability testing they did of various link implementations. The article is worth a read for any SEO.

Here is how he summarizes the main things you should do:

• Use plain language
• Use specific terminology
• Follow conventions for naming common features
• Front-load user- and action-oriented terms

Basically, the lesson is don’t force users to think hard. They are busy, have other things they need to do. Making them work to find what they want is just a bad idea.

In addition, users have been trained to look at the top left of pages all over the web, so that is what they do. What it means is that you have to left justify the action or benefits keywords in your link text.

This is all great advice and the article is well worth a read. In addition to link text, I would argue that all this advice applies equally well to page titles. After all the page title often becomes the link text for your page when it shows up in the SERPs. Having a title that will entice a click through starts to sound like a really good thing. In addition to these usability benefits, it is a commonly held belief among SEOs that the search engines also weight the words at the start of a page title more than the words at the end.

So to put it into my own words for do’s and don’ts for page titles and link text:

1. Get the benefit of a page in the first 2 (or 3) words of its title and links to that page. Among other things, this means don’t lead with your brand or site name.
2. Keep it simple. Remember, people don’t want to think. They want to get something done.
3. Don’t invent new buzzwords. People don’t care. They want what they want, and they want it quickly. New buzzwords, or unique new trademark phrases do not help your cause.

Let me illustrate this with a made up example. Consider an Atlanta based car rental company called “In-Car-Nation”. Chances are that someone there might pick a home page title such as:

“In-Car-Nation provides superior quality service and car rentals”

Contrast to:

“Atlanta Car Rentals from In-Car-Nation”

I know which one of those two I would pick.