Search Engine Marketing Case Studies Articles by SEO Speedwagon
January 16, 2008
Less Sponsored Ads = More PPC Revenue? Que Pasa, Google? 
One of my favorite clients of all time, with us now going on 8 years and powered mightily by the rare, dual client-side SEO strengths of search understanding and inter-departmental implementation influence, recently noticed the same thing Mark Jackson saw in Google's most recent round of Universal Search peekaboo:
Notice the incredibly disappearing PPC Ads? My immediate explication was that surely this must be to prove, in a small test sample, that someone's bad idea from above would be a disaster, indeed.
Mark, though, has made me think again:
Google may succeed in encouraging companies to bid more ferociously for the top two positions. If universal search leads to more searches because it's fun, this could be a win for Google (higher revenues) and users (better experience).
Sometimes it's hard for us to imagine that there is a finite set of clicks on any given day. The business model in a closed set like this, then, must discover what to do to increase the value of the average click within the set on a given day. Mark's point about less ads likelier driving up value per is on target, I believe, but thanks to him getting me to think again I think the test layout in question has less to do about increasing searches "because it's fun" and much ado about that map, an image mind you, kissing the PPC ads at the right corner of the screen and making your eye immediately jump there to focus.
Take a look yourself and see where your eye is drawn, and then check out what eye tracking heat maps are telling us about how pictures affect focus on a search page.
Less Sponsored Ads = More PPC Revenue? Que Pasa, Google?
Posted by john at 04:27 PM
| Comments (0)
| TrackBacks (0)
Printer-friendly version
September 18, 2007
Search Tearing Down Walls Like It's 1989
We knew it was coming and we tried to bake a cake for Maureen Dowd more than a Month ago, yet we are still surprised at how search-friendly they are being in their explanation today:
What changed, The Times said, was that many more readers started coming to the site from search engines and links on other sites instead of coming directly to NYTimes.com. These indirect readers, unable to get access to articles behind the pay wall and less likely to pay subscription fees than the more loyal direct users, were seen as opportunities for more page views and increased advertising revenue.
If you have any doubt that this is the SEO equivalent of 1989 scroll a bit further down the page for this money quote:
The Wall Street Journal, published by Dow Jones & Company, is the only major newspaper in the country to charge for access to most of its Web site, which it began doing in 1996. The Journal has nearly one million paying online readers, generating about $65 million in revenue.Dow Jones and the company that is about to take it over, the News Corporation, are discussing whether to continue that practice, according to people briefed on those talks. Rupert Murdoch, the News Corporation chairman, has talked of the possibility of making access to The Journal free online.
Mr. Murdoch, tear down that wall!
Search Tearing Down Walls Like It's 1989
Posted by john at 03:53 PM
| Comments (0)
| TrackBacks (0)
Printer-friendly version
September 17, 2007
PPC vs. Yellow Pages vs. Direct Mail CPA
Via Chris Zaharias via MediaPost via Piper Jaffray, we get this stark contrast:
Search advertising has proven to be fertile ground for customer acquisition. A recent study by Piper Jaffray & Co. entitled, “The New eCommerce Decade: The Age of Micro Targeting,” indicated that the average CPA for search was $8.50, considerably lower than the CPA for the Yellow Pages ($20), online display ads ($50) and direct mail ($70).
Could you imagine how low the Organic CPA would have been in comparison, had they found a way to incorporate that into the study?
PPC vs. Yellow Pages vs. Direct Mail CPA
Posted by john at 04:03 PM
| Comments (0)
| TrackBacks (0)
Printer-friendly version
March 07, 2007
Over Half of 2008 GOP Presediential Candidates Outranked by Wikipedia in Google for Own Name
Here and there The Wagon has been known to get political in its analysis, often to illustrate pols know not what they do on the internet. Today we have a new honey of a rankings scandal courtesy of techPresident:
In a recent survey, I found that Wikipedia has an expansive influence in organic Google search results for 2008 presidential candidates. For each candidate, their Wikipedia entry is ranked no lower than 5th place by Google. In addition, the Wikipedia entry ranks higher than the election web presence of that particular candidate for 25% of Democrats and 60% of Republicans.
Now first, lest the uninitiated, casual SEO observer not fully grasp the above search incompetence, it is quite difficult for a major brand not to rank first for its own brand name. You almost have to be doing something wrong at the site level, and most competent SEOs will be able to discover the reason for the glitch and remedy the error fairly quickly. The higher the brand recognition the greater the ease, if for no other reason than Google understands that a pure brand search will almost always signal an intent to find the brand site itself. Google's product is relevancy, as we like to say here.
Is there a more recognized brand on the techPresident list than America's Mayor, [Rudy Giuliani]? Yet at second he languishes, behind the Wikipedia entry replete with detailed analysis of the controversies not broached on the site he would like for you to rather visit instead.
The difference between these first and second positions? We know from the massive AOL search data leak that on that engine, at least, about half of all searchers click on #1 and south of 15% on #2, at least for the 20 million searches performed by 658,000 subscribers in that data sample.
If you are losing half of all searches on your brand that should be visiting your site uncontested, you should try and do something about it. Let's keep an eye on Rudy and see if he does.
Over Half of 2008 GOP Presediential Candidates Outranked by Wikipedia in Google for Own Name
Posted by john at 05:15 PM
| Comments (0)
| TrackBacks (0)
Printer-friendly version
October 25, 2006
A Cure For The Summer Time Traffic Blues 
“Well I'm gonna raise a fuss
And I'm gonna raise a holler
About workin' all summer
Just tryin' to earn a dollar?
Eddie Cochran, “Summertime Blues?
There ain’t no cure for the Summertime blues?
One of the clients I work with has some seasonality to their business and traffic to their web site either levels out or dips during the summer months. This summer was not any different, traffic-wise. Their search traffic numbers in the Spring months averaged 482,000 visits per month while their Summer average was 410,000 monthly visitors.
That doesn’t sound like it’s on the road to cure anything right?
A deeper trek into their analytics, though, raises the eyebrows. Their conversion rate during the higher traffic Spring months from visitors coming to their site from search engines was .825% which calculates out to approximately $13,918 in online sales per month. Their conversion rate during the “Summertime blues? months was 1.23% which is 45% higher than the Spring and calculates out to $17,538 in monthly online revenue.
Nothing like an increase in revenue to melt those blues away. But still, the higher revenue isn’t the real cure nor is the higher conversion rate.
The cure is in the answer to the question: Why is their conversion rate 45% higher?
The Cure
I have a fever, and the only prescription is a higher quality web site visitor.
I love best practices SEO.
Higher quality visitors are a direct byproduct of improved search positions for SEO-targeted keywords and phrases.
It’s no surprise to also see that this client had a Summertime increase of over 200 positions at Google for their optimized phrases.
A Cure For The Summer Time Traffic Blues
Posted by doug at 06:45 PM
| Comments (1)
| TrackBacks (0)
Printer-friendly version
April 18, 2006
SEO: The New Establishment in Online Marketing 
I'll get the disclaimer out of the way: I think the Million Dollar Homepage was a brilliant and cunning concept, and like many, I rooted for founder Alex Tew as his little boxes filled up with garish, cryptic signs.
The brilliance lay in the site's ability to run the gamut from creation to supernova in only a few short months - and making Tew a millionaire in the process. The cunning lay in the inventor's keen method of creating a concept that simultaneously raids the bank vault and boobytraps it so others will never enjoy the same haul.
As many predicted, wannabes (here are two examples, although there are thousands) learned that "first-mover" status really does mean something.
But wait. Even for the first movers, what exactly did they get? Here's what they saw, and what they have to look forward to:
Do any of the pixel buyers expect that line to shoot up again? Maybe a new PR campaign - something like "Hey, we're still sold out!"? I'm sure that many of the sites that spent $100, $500, even $1000 could have eeked out enough conversions to make it worthwhile, but 25 businesses spent $3500 or more, and some spent as much as $10-15,000 - some even after the initial spike. Ouch.
As SEO celebrates its 10th birthday, it's worth noting that online marketing dollars can be spent to give you a chart more like the following:
In online marketing, old-school, organic SEO isn't the only game in town, but it's one of the surest. It's like a solid investment in a fund with a proven track record. Pixel advertising, and whatever other flavors of the month that follow, are like scratch-offs. A few will win big. The great majority will lose.
SEO: The New Establishment in Online Marketing
Posted by erik at 11:49 PM
| Comments (0)
| TrackBacks (0)
Printer-friendly version
March 22, 2006
Technical Book Publishers: A Search Report Card
SEO Speedwagon is launching a series of "Search Report Card" articles that explore the state of search in various verticals. This is the first such article, and it examines how well technical book publishers are utilizing search and user behavior within their sites.
The purpose of this article is to explore, using a representative (but unscientific) sampling of industry-leading publishers' web sites, the level of focus currently being applied to search.
[Some history: I spent about 10 years in technical book publishing, both in-house and as a freelance editor. When I left the industry, search was just warming up, so it's been interesting to apply some current SEO benchmarks to some former-life web sites.]
Background: Challenges in the Computer Book Industry
Offline technical publishing currently faces many potential obstacles, but that's nothing new. Historically, one of the industry's greatest threats has been the contrast of the paper-publishing timeline versus the shelf life of the technology covered in the book. In other words, publish an 800-page book on June 1st about "version 4.5" of any given program, and 28 days later, "version 5.0" goes into beta testing.
Today, the old threats remain, and they're bolstered by new ones. Even Alan Meckler takes time off from buying companies to weigh in on potential risks to the model (although his post focuses more on trade mags). And then there's Amazon, the Silvio Dante of distribution partners: Nice hair, but do you really want to get into the Cadillac with him?
Publisher Sites Examined
I picked seven major computer book publishers at random:
Note: Wiley purchased Sybex in 2005, but the two maintain separate web sites, and thus both seemed suitable candidates. Peachpit and Que are both owned by Pearson Technology Group, but again, their products appear on separate web sites.
Methodology Used
I took a two-pronged approach to evaluating the sites for their respective search awareness.
First, I checked over some basics. The following criteria, taken individually, don't mean a lot, because lack of one aspect can often be accounted for by presence of another. But as a whole, they show signs of which publishers are really paying attention to search.
- Sitemap. This represents a check for a Google Sitemap file - either sitemap.xml or sitemap.xml.gz. (All sites had a traditional sitemap page.) Absence of a Google sitemap file isn't necessarily damning, especially if the site is already well indexed. Results: Only O'Reilly appeared to have a sitemap XML page.
- Custom 404. A simple test for a error-trapping 404 page.
Results: All sites had a custom 404 page. Que and Peachpit have a 404-style page in place, but it doesn't produce the correct HTTP header code. Instead, it uses a series of 302 redirects before settling on an error page with a 200 code. As we've discussed before (here and here), incorrect header codes in error pages can produce problems.
I also picked one backlist title from each publisher to perform some random on-page checks. The title was published in late 2005 or early 2006 - plenty of time to be naturally indexed, barring any obstacles.
- Unique Title. Does the page devoted to the randomly selected backlist book have a unique title page? Second, does the title contain the book's ISBN?
Results: Each publisher offers a unique page title for the pages devoted to a specific book. Strangely, however, none inserts the ISBN number in the title. ISBN searches make up a small (but not insignificant) percentage of book searches. Because the publisher is the first party to receive ISBNs as they are issued, they should own the query. - Meta Desc. Does the page have a unique meta description?
Results: Only O'Reilly and Microsoft supply their pages with unique meta descriptions (or meta descriptions at all, for that matter). Meta descriptions are a very easy place to put a book's category and description - especially if the book page template is item/list-based and doesn't contain much space for narrative text. - Deep URL. Is the URL of the specific book indexed at all three major engines (GYM)?
Results: None of the publishers appeared to have indexing problems, although due to the dynamic nature of some of the sites, I could easily navigate to a version of the book page whose URL differed from the one indexed at the engine. (Example: URL obtained by site nav and via SERP) - Robots.txt. Check for the presence of robots.txt file.
Results: All sites contain a robots.txt file.
While O'Reilly seems to hit on all cylinders, no publisher is completely ignorant to search. As I suggested earlier, these could be considered fairly superficial criteria for search awareness. They're a series of "little picture" elements that are important, but not all-important.
Second, I tried to consider the bigger picture.
Generally speaking, for most tech publishers, the current search strategy appears to be targeting their specific products. "Computer books." "C# programming books." "isbn 0470009241." And so on. And for the most part, they're doing a decent job of achieving those goals, even though sites like Amazon and smaller, one-off sites are taking a bite too.
Most publishers, however, aren't even coming close to fully exploiting the search traffic they could be. Why? Because they're busy fighting for long-tail crumbs of "book" related searches while in addition, they could be grabbing the valuable keyword torso as well.
Here's an example. For every searcher querying [c programming book], 60 search for [c programming]. For every searcher querying [windows xp book], 240,000 search for [windows xp]. For every searcher querying [xhtml book], 8000 search for [xhtml].
In other words, publishers need to become known for their subject areas - not books about their subject areas.
Is it harder? Sure. But can it be done? Yes. Who is more qualified to rank for a subject area term than a publisher with a backlist of a dozen titles or more in that subject area? Very few sites, providing the publisher uses its assets efficiently.
So what's the secret? There isn't just one. It requires an entirely new look at the structure of the publisher's web site, external sites, link-building strategy, and the role of the books themselves in the grand scheme. Following are some ways that publishers can leverage or generate the content necessary to improve traffic for their book topics.
- Bring your authors home. Computer book authors have a lot to say. I worked with dozens of them, and 90% of the time, if a book had "page count issues" during development, it was too many pages, not too few. These people live to write, and blogging has been a godsend for many of them. Publishers must provide a blogging platform for each author, and strongly encourage its use. Give them the main and right columns, and keep the left one for yourself.
Amazon is already offering a blogging space to authors (known as (gasp) "plogs"), and good authors are taking them up on it. The issue here isn't that Amazon might sell your book instead of your own shopping cart selling it. Instead, as I pointed out a couple weeks ago, Amazon doesn't care whose book they sell. With every click on Amazon, the odds that your book will be purchased are diluted - even if the odds of a book purchase increase.
- Make all roads lead to (and spring from) specific titles. Publishers have at least three "brands" in each title: The author, the series, and the topic. The URL for each unique book is the cross-section of all three:
As such, the book title URL should be the fountainhead for all types of content, from comments and reviews to coding samples to related announcements to the aforementioned author blog.
- Treat your backlist like a blog. When it comes to site architecture, think of the similarities between a publisher's site and a typical weblog. New titles (posts) emerge and are filed under specific topic headers (categories). Just as Engadget's "Gaming" category receives regular updates via new posts, so should your ".NET" category receive regular updates via each new published title. Reviews (comments) are added regularly, and all hope that the author (poster) will engage the audience in a dialogue.
- Cultivate new backlinks based on your content. Without getting too specific, if you're wondering why you can rank well for [.net programming book] but not [.net programming], one smart place to look is your backlinks. Do a query for your topic, both with and without "book," and take a look at the sites linking into the above-the-fold placeholders. Would any of the new content described earlier lead to incoming links from those different types of sites? If all your backlinks are coming from publishing-based sites, do you realize what Google considers synonymous with [~publishing]?
Who Gets It?
While none of the publishers appears to have fully grappled with search from a holistic perspective, many are approaching search intelligently, mostly via consumer-generated methods like blogging:
- It's hard not to recognize O'Reilly's Radar blog as a dominant force in the publisher-as-tech-blogger circles, but the tie-in from blog to backlist is a little lacking. Still, the book site rocks, and smart programs like Rough Cuts help to propel titles like Ruby on Rails Up and Running into SERPs for queries like [ruby on rails programming] - several months before the final book is even pubbed.
- At the acquisitions end of the product chain, Jim Minatel's Wrox Blog takes a microscopic look at the publisher-author relationship, and Joe Wikert's simply-but-smartly titled Book Publisher Blog looks at the bigger picture, while at the same time owning [publisher blog].
- While somewhat lacking in text, the Que individual title pages - via their breadcrumb nav and title-specific list - offer nice links to each of the three axes (author/subject/series) referred to in the graphic above.
Conclusion
Outside the comfortable surroundings of SEO/SEM, we're sometimes shocked by a vertical for whom search isn't at the forefront, and as SEOs, it's good for us to gain that perspective. Search should be a significant part of a bigger picture, not the picture itself. But in the face of ever-growing threats to their business model, publishers would be wise to better capitalize on searcher behavior. In a way, publishers are lucky. Many verticals, as we'll show in the coming months, need to start creating assets by the bucketload. Fortunately for publishers, all they really need to do is start utilizing the assets they already have.
Technical Book Publishers: A Search Report Card
Posted by erik at 10:23 AM
| Comments (0)
| TrackBacks (0)
Printer-friendly version