It seems obvious that to drive ecommerce success, you need a website. But in many cases that is probably the wrong strategy.

Consider an article in The Hollywood Reporter about Viacom MTV’s decision to abandon the mega-site and create dedicated sites for shows like The Daily Show, The Sarah Silverman Program, and SouthPark. As reported:

The sites join a growing list of targeted Web sites that the Viacom property has launched in the past year in conjunction with its TV shows. Other sites include Comedy Central's Indecision2008.com, MTV's YoMomma.TV and VH1's BestWeekEver.TV.

The sites are the latest in MTVN's strategy in the online arena, which establishes individual destinations for shows and related subject matters instead of a centralized site. The new portals will bring the worldwide total of MTVN Web sites to more than 300 destinations by year's end.

Traffic across MTVN's properties has increased from 76 million unique visitors in January to 91 million visitors in July, according to comScore Media Metrix.

It makes a lot of sense. Fans of The Daily Show really don’t care that it’s part of ComedyCentral or is part of MTV and owned by Viacom. And a dedicated site almost certainly gets more resources, more content, and earns dramatically better organic search results.

Most ecommerce sites offer a range of goods from different manufacturers across a number of categories aimed at different user groups or usage purposes. A large sporting goods retailer like Dick’s sells tons of Nike, aims at Golf and Soccer and Running, and services weekend, hardcore, and armchair athletes. Targeted sites, which have been sparingly used under the name ‘microsites’ for a long time, would give Dick’s the same advantages as MTV.

A look at the Dick’s site reveals a highly attractive and professional site, with advanced features like guided navigation and user ratings. But managing such a huge range of goods has resulted in very little depth beyond core product information – the golf section for example is 100% ‘picture-price-paragraph’ without any supporting information beyond lovely glamour photography. A quick check reveals that the site doesn’t gain organic rankings on the first page for any of the main phrases of that section – golf bags, golf clubs, golf balls, etc.

Who does rank for Golf Balls? The answer is golfballs.com, golfballs101.com, golfsmith.com, onlygolfballs.com, etc. For Golf Clubs the answers are similar. How many millions of dollars is Golf worth to Dick’s? How many more if they ranked well. (They also don’t seem to be competitive in related PPC, fyi).

What would it cost to develop 50-100 pages of content and more informative and interactive features for the golf section, or better yet the new golf site? Relatively little. From my experience the technical issues of their CMS systems, commerce platforms, and in-house IT depts. are far larger barriors. Whatever these struggles are, they’re going to need to get resolved.

This isn’t picking on Dick’s, I chose them as an example of broad inventory, but all of the above could be said about Wal-Mart, Musician’s Friend, Advance Auto Parts, and many others. Of course, windshieldwipers.com, harmonicasandstuff.com, http://shitbegone.com, have already demonstrated the benefits of targeted sites, gaining top 3 rankings ahead of these merchants.

As the revenue potential, competitive realities, and ‘glass ceiling’ in paid search become clearer to the major (and smart minor) ecommerce companies I believe we’ll see a lot more MTV Retailing, with companies running a network of focused sites instead of (or in addition to) one central mega-store.

Posted by Craig Danuloff at 9:55 AM | Permalink

Search Engine Optimization sure seems to be a controversial topic. It’s Easy. It’s hard. It’s dead. It’s alive. It's BS. It's Real. It should be done in-house. It should be outsourced. All that plus hundreds of blog posts every day, thousands of forum messages every hour, and a conference just about every week.

But who really understands this somewhat obscure art and science? Do the marketing VPs or online marketing directors who decide to allocate budgets and resources?

Apparently not. How else can you explain the fact that paid search gets 8X the budget allocation of organic search, and yet generates only about one quarter of the clicks?

And within those organizations that do make the decision to invest in organic optimization, do the insiders that need to support the execution of an SEO initiative know how it works and why it’s important?

Certainly not in my experience or that of the other professional SEO’s I recently talked to about this topic. In fact, just about everyone you’d need to assist in search optimization is pretty likely to work against it. Consider the following profiles - caricatures to be sure - but not far off the mark in my experience.

• Marketing Manager – Generally supports SEO or the project wouldn’t get started, but usually knows next-to-nothing about what it really entails in terms of the breadth and depth of technical, content, and design changes to the website that will be requested or required. When the magnitude of these change requests appear, they often focus on negotiating down to what’s feasible given the internal politics instead of managing the politics to get the necessary changes done. In the end, they get a sub-optimal result – hoping they can somehow get 80% of the benefit from the 20% of the changes that made it through. Of course they won’t.

• Web Developers– Doesn’t know much (or anything) about search optimization, doesn’t really care, and sees all the SEO requests as changes to what they’re planning or the ways they’re used to building site/pages. Frequently have an easy-going ‘I’ll do whatever they want me to’ attitude going in but as development moves along somehow the SEO requests often get shaved in order to ‘make the schedule’. In other words, they were all seen as ‘nice to have’ requests and not ‘must have’ requests.

• Product Managers / Merchandising Managers – This one is the most surprising. You would expect these folks to be naturally included to present clear and complete information about their products. They’ve been limited picture-price-paragraph in catalogs and brochures for years that suddenly in a medium with free incremental space you’d expect them to go wild – extolling the virtues of their products at length. Doesn’t usually happen. More commonly see the need for deep rich content and extra work (we don’t have long descriptions or up-to-date feature lists on all these products!). Seem to think prospects will buy their goods or services just because they’re selling them. Forget SEO and better rankings, shouldn’t they want to deliver more information in pursuit of higher conversion rates and more revenue?

• Graphic designers / Web designers – While it’s almost possible to understand how web developers have made it to 2007 without any idea about the methods of SEO, how did so many people who design websites for a living make it this far without becoming at least minimally aware of this stuff? The battle here of course comes down to form over function, with ‘text’ being a four-letter word for ‘aesthetically unappealing’ and anti-aliased .gif files somehow frequently ending up the lynchpin of the designs.

• CEOs / Senior Managers – These folks shouldn’t have to know about SEO, because those functionally closer to the projects should have it covered. But since they don’t most are left to notice (or be told) that their sites don’t show up on the first page in Google for some keyword they decide is critical (like their core product name or category) and then they spring into action. Calling in the Marketing Manager (described above) they then hear a long and confusing list of reasons and finger-points as to why the site/page doesn’t rank: the developer had technical limitations due to the platform and the design that graphics and marketing wanted doesn’t index well for some reason, and the product managers are working on generating more copy, and it’s basically an intractable situation but next year they plan a site redesign and they’ll hopefully get it right then. They’re left believing that SEO is an intractable problem and with the view that paid search is at least ‘actionable and manageable’.

Herein lies the real problem with SEO.

While the potential benefits are easy to articulate (who doesn’t want high volumes of free traffic from the search engines) as a marketing discipline it has something of a branding and positioning problem and even when projects are approved and begun the real-world implementation very frequently turns into a protracted and very difficult process.

A lot of the reason is the perspectives and motivations of the players, as described above. But there’s another massive problem too.

More about that in the next post.

Posted by Craig Danuloff at 10:26 PM | Permalink | Comments (2)

The question really isn’t whether SEO is dead or not. As long as people want to get their pages and content listed and linked in Google and the other engines, the process of increasing the odds of that happening will exist and it will be called Search Engine Optimization.

Never. Gonna. Change.

The question is whether or not the algorithms (and their human assistants) are so darn good and in fact measuring the actual level of satisfaction that your pages and content bring to real users so well that by optimizing your site for the users you have in fact done everything you could possibly do to increase your chances of the search engines ranking you highly for all of the search phrases relevant to your content.

That is the idea that Mike Grehan, and then responsively Bryan Eisenberg are getting at. Make the people happy and the engines will take care of themselves. It’s also the ‘Don’t Worry Make Them Happy’ song the Google Webmaster Guidelines have been singing for years.

Trouble is, it just isn’t true. The search engines don’t rank based on customer satisfaction. They develop indicators and proxies that they hope/believe should highly correlate to customer satisfaction, but keyword placement, word context, link structures and the dozens or hundreds of other variables they track don’t always hit the mark. The current results prove that.

So until Google starts surveying visitor satisfaction and using that data to rank sites in their engine (not a bad idea, but one that brings its own issues), it’s not yet time to cross SEO off your to-do list.

NOTE: I want to make clear that I completely agree with both Mike and Bryan on the importance of working hard to improve the core quality and satisfaction delivery of websites. I’d even strongly agree that it should be the most important thing you focus on. But we aren’t yet at the time or place where it’s the only thing you need to do.

Posted by Craig Danuloff at 11:57 AM | Permalink | Comments (2)

It will probably only last a second or two. In the scheme of things it’s a relatively small shift of weight. There is no danger they’re going down, let alone for the count. But let’s enjoy it.

The New York Times and Jason Calacanis have found the Achilles’ heel and with some justice it’s deep in the heart. The Algorithm.

As fast and as smart as it is, it can and is gamed on a very consistent basis. Although the larger miracle of general quality and accuracy has created a brand image that overcome reality time and time again, and despite the irony that there has probably been more progress made to improve the results in the toughest spots in the past twelve months than in the previous five years, it still remains true that the more likely there is money attached to your search, or significant diversity surrounding its concept, the higher the probability that your results will NOT be the 10 most relevant in the world.

The fact that humans can do a better job isn’t really even an argument. The issue is speed and scale. The brilliance of what Jason is doing at Mahalo is the idea of taking on the top XXX thousand most popular searches and letting the algo’s have the rest. It’s still a bear of a task, in terms of coordinating and avoiding conflicts of interest (real or imagined), but the rise and domination of Wikipedia clearly says it’s not impossible.

A lot of the response from Matt Cutts sounds very defensive. ‘People turn on our computers, and people wrote and loaded the software…’ but in the end he seems genuinely open to having people edit results. That of course would be the smart thing to do, both functionally and competitively. Google could certainly arrange to have the top 10K results human edited in very short order. They don’t even have to make a big switch, just offer a tab at the top, or preference or alt search button (‘algo search’ vs ‘editor says’). Let users opt-in to these results when available, or switch to them when the machine is being out-witted by 100K spammers just out for their commissions.

It’s amazing how competition drives better results. And how a little press (mainstream or blog-o-based) magnifies the significance and pressure generated by anything. I have no doubt that the ‘human-assisted’ arguments inside of Google will heat up and make more headway tomorrow than they did last week. I’m also sure that Mahalo and lots of other competitors will be equally spurred on to leverage their progress into even greater success. And that’s good for all of us.

Posted by Craig Danuloff at 9:36 AM | Permalink

When Mike Grehan says SEO is Dead it’s a little like Martha Stewart telling you that Twinkies with candles make a fine wedding cake. Yet that’s what he said. The release of Google Universal Search Results and the new Ask had so much impact on him that a man who’s probably spent as much time as anyone trying to grok the engine’s algorithms seems to have thrown up his hands and cried uncle.

His argument, if I understand it correctly, is that these newfangled results are so darn good, and so attractive with all the images and videos and maps and pricing and other useful stuff, that finally ‘better marketing counts’ and SEO techniques are now just plain uninteresting. He implies that getting a text listing on these multimedia pages isn’t even worth the effort.

This is where I lose him. Don’t you have to ‘optimize’ to get a listing near the top of that first page regardless of the type of content you hope to have listed? Don’t you still have to be selected as one of the 5-10 most relevant results in the world to appear on that page? Isn’t that selection going to be done based on some scoring criteria and isn’t SEO the process of figuring out that scoring criteria and then putting those points on the board?

It’s impossible to argue that if it’s relevant to the search term anyone wouldn’t prefer to have an image, or video, or other piece of visually interesting content appear. But when that’s not relevant or possible I’m sure most would be happy to have a top text link with good title copy and a great descriptive snippet. Isn’t search engine optimization still the only way to get that spot? What exactly am I missing?

Posted by Craig Danuloff at 6:36 AM | Permalink | Comments (3)

Long ago I heard somewhere that half of all the searches done on Google every day were brand new – Google had never (ever never ever) seen them before. I’ve repeated that stat a lot over the last two or three years, never having been able to attribute the source.

So imagine my pleasure and surprise to find a brand new reference to Udi Manber, Google's VP of Engineering saying that “20 to 25% of the queries that Google sees today, have never seen before.”

Wow. It doesn’t even matter that the number is ‘only’ 25% - a few years later it would have to be lower – that is an amazing number. Google does between 2 and 10 billion searches per month. It’s been at that for several years now. And every day those slippery humans ask for 50 million or more things that nobody has ever thought to ask about in that exact way before. Wow.

I’m not exactly sure what this means for search marketing, other than showing that along with the algorithms and the competition and the way results are grouped and presented the actual inventory is a moving target as well. In paid search ‘broad’ and ‘phrase’ match probably take care of most of these new searches (or at least those in your area of interest). For organic optimization it certainly favors the ongoing development and growth of content – the larger the site and the richer the language the better the odds of having an exact match of all or part of these new search phrases.

What do you think it means?

Posted by Craig Danuloff at 9:23 AM | Permalink | Comments (1)

Bounce rates are public enemy #1 in the world of conversion, because they represent the number of times visitors slam the door in your face without even getting to know you. The good news is that bounce rates could be dropping considerably in the near future.

The bad news is that the reason isn't that web sites are improving so dramatically that more visitors will more immediately recognize a path to their goal and therefore continue clicking deeper into the site. It's because web-page previews are starting to appear in search engines (and other places) so visitors can reject you without the hassle of having to even visit your site in the first place.

Instead of slamming the door in your face they'll just be driving in and out of the parking lot. The trouble is you won't even know they came because since they didn't load your page the drive-by won't show up in your website analytics. Conversion rates will skyrocket (yeah!) but this could give the false impression that your home page (or any other web page for that matter) isn't really bad - when in reality it probably is.

The new ASK 3D is the example I saw today that got me thinking about this. There have been other tools like this include SnapShots that can be added to any site. As these tools become prevalent and more full featured, it's another major shot across the bow of website analytics as we now know it.

Of course, there is a simple solution. The preview technology guys (Ask, SNAP, others) could be kind enough to send us a tagged ping when our previews are viewed. Maybe even include the keyword (in the case of search, the URL in the case of SNAP) where the preview occurred. Quick, someone call the WAA or our Analytics Lobbyists in Washington...

Posted by Craig Danuloff at 8:37 AM | Permalink | Comments (1)

I’m talking of course of the ever-expanding list of Webmaster Guidelines, a supposedly friendly set of suggestion for how you should create and code websites if you want to be respected on the Google Internet. Of course, it’s really just a Webmaster Wishlist, describing the way sites and pages and links would be created in a world where everyone was trying to serve Google’s best interests.

In other words: “Those with the gold, make the rules.”

The existence of the list implies an arrogance, but there’s a practicality too. We’re free to ignore their suggestions, and they’re free to ignore our websites. In theory, a balance can be found between what they ask of the world and what they provide to the world, and we can all just get along.

When the ‘Guidelines’ specify clear actions that will result in pages or sites being removed from the Google index, I find it a useful document and think it operates on a fair premise: Don’t stuff keywords in white type on a white background. Don’t create 700 versions of the same page with different title tags. Don’t show one page to the GoogleBot and another page to Joe Consumer.

When the ‘Guidelines’ get vague and try to make water flow up-hill in pursuit of some supposed karmic-good, I think they’re laughable and should be treated as such: Don’t create additional pages for search engines (just think about your users). Don’t separate sites into different domains or sub-domains because you believe it will result in better organic rankings (again, pretend the engines aren't there). Don’t exchange links or otherwise lift an un-natural finger to get more links to your site (pretend this is not predominantly what you're being measured on.)

Which brings us to today’s broo-haa-haa. The trouble with links. Votes. Recommendations. References. Google Juice. The bots at Google have asked the carbon-life-forms for help. You too can squeal in the name of Google.

They need a clean world where all links are truly organic (naturally forming) and since we’re all cursed with the knowledge that more links improve our rankings (and thereby results and generally bottom lines) we just can’t help but bite that forbidden fruit. Thus paid links are original sin.

Here’s where the ‘Guidelines’ really go crazy, actually requesting that you not share in the value that has been jointly created by your content, your link equity, your rankings. Your website in co-operation with Google and its algorithm has been given a certain status, and the collective value of that status is why people search Google and they make all that money. There is no doubt they own the collective value.

But shouldn’t you own your sliver of the value, and be able to monetize it? Both legally and morally in terms of the relationships between websites and the Google index? Why should they keep 100% of the value?

And if sharing in that value is confusing to their algorithms, is that really our problem?

As a side note. I’ve never sold a link and very rarely recommended or participated in their purchase. Neither decision was based on what Google may or may not think.

Posted by Craig Danuloff at 10:46 AM | Permalink

The stakes in local search engine optimization went up today, as Google announced their free 411 service. Now anyone call dial 1-800-GOOG-411 and get not only the phone number but also connected for free to local businesses - or a list of available businesses in a category.

And guess what - the listing results are provided in the exact sequence of the results for the same search at local.google.com. Local search has been growing, and I know in some cases having huge impacts for very small businesses smart enough to optimize.

But as rank-based free-411 catches on the benefits of a #1 ranking for a category in local are going to skyrocket. And so will the efforts made to win the local search race.

Bonus: Tips for the service below.

Continue reading "Google 411 - Optimizing For Local Voice" »

Posted by Craig Danuloff at 4:40 PM | Permalink

Today Jason Calacanis helps us take another step towards honest language for online marketing. He does this by saying that some sites don’t deserve traffic.

He says it as part of an insult meant for the ‘snake oil SEOs’ and their clients, and in a post otherwise filled with ignorance regarding search engines let alone optimization.

But never-the-less, he’s exactly on target with the notion that the first step towards ranking well in the search engines is deserving to be ranked above everyone else.

What a great way to start a conversation about improving search engine rankings: “Do your pages deserve to be ranked #1 in Google?” “Is yours really the most relevant and important site on the entire internet relating to the word _________?”

The perfect ensuing discussion would review the merits and shortcomings of the website, define the editorial and structural changes necessary in order for both the client and the SEO to agree that the site deserves to rank highly, and plan the allocation of time and resources required to make the needed changes.

But even in this idealized situation the need for genuine Search Engine Optimization isn’t really reduced very much. Jason’s contention that “‘if you make great content, keep your page design clean, and stick with it you're gonna do just fine in the rankings” is simply wrong – if by ‘fine’ you mean ‘achieve the majority of your potential in terms of organic search engine traffic’.

Continue reading "The SEO Results You Deserve" »

Posted by Craig Danuloff at 11:12 PM | Permalink

I really try to avoid posts about new Google products or features. But today they put one out that is actually related to search and it’s terrific. It’s an upgrade to Google Webmaster Tools that provides clear and useful information about who links to the pages in your website.

Until now the tool available for finding the inbound links to your site were very poor (like all other SEO tools). And while Google is clear that this new information isn’t complete, it is significantly better than what we’ve had before in terms of both quantity (there are way more links listed than before) and accuracy (we’re now shown how many links point to each page and can easily see and download the linking URLs).

This is important because inbound links play a critical role in how pages from your website rank in Google and the other search engines. (Skip this part if you already know how it works.) The search engines essentially measure relevance and reputation – and the only clues they have are what you say about yourself (the content on your web pages) and what others say about you (the links into your site and who/where they’re from). As such, optimization is the process of managing website content (which words go where on what pages) and cultivating relevant links.

So half (or more) of our organic search engine fate is in the hands of inbound links, and for the first 10 years or so of the internet age we’ve had to do without very good visibility into these links. Technically speaking, it’s been like getting ready for a date without a mirror. We’ve had only a vague idea of where we stood going in and just enough information to be confused afterwards.

One tab over in Google Webmaster Tools, under Statistics and Page Query, Google shows a seemingly random sampling of the link text they found pointing to our site. This another piece of the puzzle, so while they're feeding us peasants why doesn't Google link these two pieces of data so we can see the link text for each inbound link?

If they don’t, let’s hope someone puts up a simple tool where you can upload the .cvs file of your links (which you can now download) and then see a nice anchor text analysis. I bit someone is probably coding it as I type this request. If you're the one doing this, please also grab the PageRank of the linking page, and let me upload a list of target keywords so you can tell me how many of them are found on the linking page too. Thanks.

In any case, thanks to the good folks at Google for this tool. Here's hoping it's the first of many Webmaster Tool enhancements we see this year.

For more details and helpful analysis of the new features, visit SearchEngineLand)

Posted by Craig Danuloff at 9:47 PM | Permalink

First Search Engine Optimization was easy, now it’s something you really shouldn’t do at all. In a column on ClickZ this week, Shari Thurow argues that optimizations for the sake of the search engines aren’t that good of an idea.

Miss Thurow apparently thinks search engine optimization is something that happens to you while you’re busy doing other things – like building ‘user centric websites’.

Newsflash: It’s called Search Engine Optimization. It’s about getting your pages to rank highly in the search engines. It shouldn’t then be controversial to actually try to figure out what the search engine algorithms measure and then build or modify sites in ways specifically designed to rank highly.

Yes the algorithms are secret, and so the process requires study, testing, and some guessing. Results are not guaranteed, the algorithms are certain to continue to change over time, and mileage may vary. And of course there are methods and tactics that are ‘over the line’ which I’d reject only partially on moral grounds but largely because those specific methods are likely to only have short term benefits. But there’s a lot of ground between ‘be nice’ and ‘you’re evil’.

I think her distinction of ‘short term vs long term' seo is bogus. There is intentional SEO and there is coincidental SEO.

• Intentional SEO recognizes how the engines work and take that into account throughout site planning and content development. Without defining the boundaries between algorithm chasing and algorithm awareness, it just doesn't' make sense to avoid optimizations that are purely designed to provide the engines what they want to enable a page to rank more highly.

• Coincidental SEO is what happens to most people who build websites without sufficient algorithm awareness – they happen to rank at some level for some number of related terms. Building a website with a strong focus on satisfying the user would likely mean more text, more pages, and therefore more coincidental ranking results. But it would also miss a massive amount of traffic that actually optimizing the website (for the engines) would generate.

In a recent post I wrote that sites need to serve three masters – marketers, customers, and search engines. It’s possible to fully satisfy all three and online marketers should strive to do so.

Posted by Craig Danuloff at 8:02 PM | Permalink | Comments (1)

Links built the internet. It was the ability to connect two pages or sites with just a link that separated the web from all that came before it. It was the inferred value of that link that created Google.

Today one of the most important and link-rich sites on the internet opted-out of the linking business, at least the Google side of it. They’ll probably reverse that decision within a few days, and they should.

But we are a long way from those innocent days when a link could be trusted. Why do they have to ruin everything?

Posted by Craig Danuloff at 1:10 AM | Permalink

Who are you trying to satisfy with your web site? Don't answer too quickly, it's a trick question.

• If you read too much about search engine optimization, you'd quickly get the idea that every word on (and underneath) your website should be designed to satisfy Google and those who chase it.

• Think about it for a while, and you're likely to say that serving the needs of your visitors is most important - answer their questions, solve their problems, give 'em what they want.

• Be completely honest, and the answer might be 'I just want the boss to sign off on it'. The boss might be the marketing dept, the CEO, a committee, or whomever gets to approve website content.

The real answer should be all three. Your site needs to satisfy the needs of your company, the desires of your visitors, and the realities of the search engines. This will mean some compromise, but it's the path to long term success.

(Inspired by another great post by Nicholas Carr, discussing how newspapers are writing stories with Google in mind.)

Posted by Craig Danuloff at 8:40 AM | Permalink

Was there ever a simpler and more controversial topic? Phillip over at Google Blogscoped answered the call of some non-savvy relatives and wrote a generally great primer that nails the three primary directives:

1. Create good content.
2. Make the content accessible.
3. Tell others (hope for links).

His advice is clear, honest, and will produce good to great results for many who today are lost in the dark woods without a flashlight or a map. But then Phillip himself wanders off the trail. After all the good advice he says:

“take a break for a while. And then get back to continue to grow your site. But don’t worry about Google results for the first couple of months, in fact, don’t worry about Google results at all. Your site might not appear in search engines in the beginning, and maybe once it does, your competition will rank higher than you... but these things take time….What you can do, though, after a couple of months, is…”

At which point he gets back to good stuff about checking analytics for keywords and avoiding SEO trickery.

My argument is with the ‘don’t worry about your results, go back to growing your business, in a few months check back if you have time, you’ll win some and you’ll lose some’ stuff. Telling someone who has created a site, has expectations for their online business, and took all the trouble to learn and then optimize, to ‘don’t worry about it’ just isn’t realistic, or good business.

I think at that point he should have told them to immediately get analytics set up and start learning to understand how visitors use their site and where they came from, and pay close attention to new visitors who start coming in from natural search.

Common wisdom is that a) you can’t rush Google and b) results don’t start for a couple of months. Both are dead wrong.

There are things you can do to get a new site or new pages indexed, like turn on Google Sitemaps or get some links. And the more links you get the faster the site or any content is going to get notices. The reaction time for unique keywords like company names is often just a few days – and getting those first links is important and exciting. More importantly, the trickle will start at some point (if everything is going well) and you’ll want to know about it and work to amplify it.

The one other addition I’d make the great opening sections of the post is to explain the idea of keyword selection and covering the diversity of ways different people say the same thing.

Phillip does a great job of recommending and explaining page focus and basic tags, but just behind these concepts a newbie should learn that if some people call your ‘Blue Dog Coat’ a ‘Blue Dog Poncho’ then you’ll want to work that content in or add a separate page for it.

Covering the range of potential keywords for any site could be considered slightly advanced for what he was covering, but in the end it’s a basic point and very important – working in extra phrases could increase traffic by a few or many times – especially if some of the ‘alternate’ phrases are less competitive from an SEO viewpoint.

These suggestions aside, I’ll be bookmarking his post and recommending it.

Posted by Craig Danuloff at 10:48 PM | Permalink

Yesterday I posted some comments on the complexity of organic search engine optimization. I tried to make the case that completing a basic site optimization is (theoretically) easy, but delivering the intended results - the maximum possible amount of organic search engine traffic – is extremely difficult. Marketers need to stop buying search engine optimization and SEO’s need to stop selling or promising site optimization, and both need to start (respectively) buying and selling results.

A few related thoughts:

1. I believe that ‘SEO is easy’ in the way and context I described yesterday, but Danny Sullivan makes a cogent point in saying that ‘it is rocket science if you know nothing about it.’ While the basics of search engine optimization can be, as I am occasionally fond of saying, be written on the back of a business card, the painfully obvious fact is that the vast majority of web site developers and owners are clearly oblivious to these basics.
   
   Moreover, implementation of these principals once you do know them does seem to confound otherwise intelligent people more often than not. And there are certainly way more exceptions to the basic rules than basic rules themselves. Hence the addition of the parenthetical (theoretically) into today’s summary of yesterday’s point. Maybe it's not so easy.
   
   
2. The SEO industry is effectively crippled by a horrible lack of measurements and tools. Neither the engines nor the analytics companies treat organic ranking like an important metric. In fact, for reasons both good and bad Google and the others make it very difficult to track search engine result page rankings – doing so is in effect an illegal activity.
   
   Sure there are a half-dozen 'SERP reporting' utilities out there - but until the engines sanction, or sell, legitimate rights and methods to adequately check results for large numbers of keywords building such tools are just plain bad investments for serious developers. Notice that Adwords and Google Analytics will tell you dozens of things about how your paid ads are doing, and effectively nothing about your organic rankings. Until paid advertisers start request (or better yet demanding) vastly improved organic reporting we're probably not going to get anywhere.
   
   Additionally, I’ll renew my call for the engines to pass along in their referrer data the exact position number in the results where the clicked listing appeared. I just want them to tag the following on the end of the referring URL: ?serp=14 to tell us that the organic click was generated by a listing currently ranking 14th in their results. They could do this easily and it would be a great first step toward meaningful organic reporting. Think a report showing how much more money was being made off of top 5 results than page 5 results would motivate some SEO attention, effort, and spending?

Thoughts on paid search and the PPC vs SEO comparison tomorrow.

Posted by Craig Danuloff at 11:58 PM | Permalink

How hard is it to optimize a web site? Is it a one-time or an ongoing activity? Is paid search harder to manage or master than organic search? Is one ‘better’ than the other?

These are the meaty questions flying around within a somewhat silly and intentionally provoked debate going on within a number of blogs and forums in the search marketing community.

But they’re also real questions that we hear all the time from CEOs and Marketing VPs trying to select a path through the world of online marketing.

Optimizing a web site is easy. The same way that playing golf is easy. It doesn’t take long to get the hang of holding the club, whacking the ball, driving the cart, and counting your strokes.

But becoming a good golfer and producing great results (ie a low handicap) is very difficult. It takes oodles of practice, development of very fine motor-skills, the mastery of widely varying conditions, selection and control of equipment that best suits your individual characteristics and style, and sustained mental discipline. And of course there is no such thing as total mastery – the courses, the competition, and other variables make the pursuit of perfect an ongoing process.

A web site can fairly be considered ‘optimized’ if it has individual pages targeted at specific keywords, title tags and other content focused towards those selected keywords, and at least a few relevant inbound links. Anyone with a little time and at least half of a high-school education should be able to get it done for you.

And with a little luck you’ll start seeing an increase in traffic from Google and the other engines, and perhaps some ‘high ranking results’ for your company name and maybe even a few targeted keywords. But in the broader scheme of things (meaning measured against the potential), your results will not be significant. Completing a basic site optimization (even a very good one, which VERY VERY few companies accomplish) and then expecting significant results is a bit like learning to putt through the windmill and then expecting to pick up a Green Jacket at Augusta National.

Continue reading "Rocket Science For Dummies Pt. 1 (Or - Why You Need To Outsource SEO and SEM)" »

Posted by Craig Danuloff at 11:10 PM | Permalink | Comments (1)

MSNBC points out that GoogleBombing is alive and well in campaign ’06. Organized bands of partisans are working to drive negative articles higher in the rankings for their political opponents in a new form of participative democracy. The classic example of this technique took place for the ’04 Presidential campaign, when a search for the term ‘miserable failure’ started turning up what Google themselves call ‘odd results’. (more history here.)

The incident marked something of a turning point in the SEO world causing links to replace content (and tags) as the broadly accepted ranking powerhouse. More importantly, it was an early chink in the armor of Google’s public image, forcing most people to consider for the first time exactly how it was that Google decided who shows up at the top of search results.

Today, Google is a habit and a verb, but we’d all benefit from more scrutiny and skepticism of their results. The algorithm has made impressive progress in many areas over the past two years, but we’re still very far from any kind of true relevance meritocracy in far too many results. Towards that, I’m glad to see another round of publicity on how and why search results are manipulated.

I’ve heard that the original GoogleBombing to get the G.W. Bush page to the top rank for ‘miserable failure’ took only 60 inbound links. Of course, it’s hard to say at this point whether Google really was manipulated in this case or if something in their algorithm was actually able to predict the future.

Posted by Craig Danuloff at 2:05 PM | Permalink

You've probably noticed that Google recently revised the look of their #1 listings for many search results. Beyond the normal snippet and URL, they now include a sub-group of options for each page, enabling users to better locate the content they actually need.

What's interesting about this is that Google may be shortening the 'long tail' of search. In other words, if Google can provide sub-options with relevant results for more different search intents within the results provided for the generic search, it may actually dissuade people from getting more and more specific with their search phrases.

So the benefit of (and difference between) searching for ‘order sirus’ and ‘sirius customer support’ and ‘compare Sirius radios’ vs just searching for ‘sirus’ goes down with the new listing format. Over the past five years people have been learning to search on longer and longer phrases to try and find more specific and relevant content. General terms still get the bulk of searches, but the tail had been getting longer. The actual quality of the results for the longer searches varies greatly, sometimes it's better but often times the result produced for a highly specific search is worse, which is a lot more frustrating.

In a purely academic sense, it would be better if users continued to build search phrases with more precision, and if the engines also continued to improve their targeted results. But the reality is that habits change slowly and if the engines can just offer an ‘intent menu’ under each relevant listing, this may improve the process for everyone. SEO’s now have to build intent cues into pages not so that they can get found for longer searches but instead so that the engines can correctly classify the available information.

I assume Google is tracking the clicks on these sub-results very closely. It would be great to see these rolled out for more than just the top listing.

Posted by Craig Danuloff at 10:11 AM | Permalink | Comments (1)

It took only about 24 hours for Google to respond to my call for more statistics for search listings, such as the number of impressions and clicks. Well sort of. Yesterday Google Base, their database listing product began offering these details to its users.

Great. Can we have this info in the main search listings too now please?

Posted by Craig Danuloff at 5:49 PM | Permalink

Why is organic search the bastard stepchild of search marketing? I think there are two main reasons:

• Paid search is predictable, organic search is not. In paid search you control the keywords for which your listings appear, the copy that is displayed, the target landing page, and the date when listings first appear. With organic optimization you can attempt to influence each of these but the results are quite unpredictable and essentially uncontrollable.

• Paid search is measurable, organic search is very difficult to measure. The paid networks including Adwords and Yahoo Search provide easy access to impressions, click-through rates, average positions, costs, and conversion rates/values. Analytics packages will tell you how many organic clicks you got and provide conversion rates/values, but getting any more takes work or is just plain impossible.

Imagine if we had a clear picture of the current and potential value of organic search position rankings. Here’s a rough idea of what I’d like to see:

1. Each referring URL from an organic engine should provide the position number along with the keyword that drove the click.
2. Analytics software should be able to later look up (from each engine) the distribution pattern of clicks for each keyword (ie position 1 got 23.3% of the clicks, position 2 go 7.3% of the clicks, etc.).
3. Using these two numbers it would be possible to get a report of the potential traffic for each keyword if you achieved various positions in the search results. Summary reports should show how much traffic if you achieved #1 for every term, and then a bunch of other target distributions (20% of terms in the top 5 slots, 30% of terms in the next 5 slots, 20% on page two, etc.)
4. Using your actual conversion rates and gross / net revenue numbers, produce the same tables showing the revenue potential of various combinations of organic search success. In other words, tell me that I’m now in position #12 for ‘professional table saw’ but that if I earned position #3 it would be worth a 650% increase in volume and a 12% increase in conversion for a total economic value of {whatever the hell it would be}.

This all just one quick example. Nothing wished for here would be hard if the engines would provide a little data and the analytics companies would do a little work. The recently leaked AOL data, combined with some other publicly available info and a cool new tool from Black Hat SEO offers a sliver of this dream.

Continue reading "Dreaming of Better Organic Search Numbers" »

Posted by Craig Danuloff at 8:16 PM | Permalink | TrackBacks (1)

Earlier this week we were in a strategy session with a new client helping them to understand the value of dramatically expanding the scope of the content they planned to add to their new retail website. We want to encourage them to move beyond the picture-price-paragraph that still characterizes too many web retailers, and instead become a massive information warehouse on their unique market segment.

Yesterday I noticed Jacob Nielson's new Alertbox article, which provides interesting data and view, mostly supportive, on the idea of expanding site content and specifically the number of site pages. Jacob points out that a typical linear graph of page views shows how dominant a few pages are in the total traffic of most web sites, but that if you instead plot a double-logarithmic chart what you see is that in fact the less visited pages do contribute significantly when aggregated, but in the case of most sites there really aren't enough pages to fill the long tail. So it droops, as you can see below.

It would take a lot of pages to fully fill the tale in Jacob's example (259,000 to be exact) but the idea that there is traffic out there for more content on your topic is one that doesn't require these types of impossible numbers, nor does the tail have to be filled all at once. It is interesting however, that he shows the filling the tale would be worth a doubling in site traffic according to these projections.

Expanding the size of your site, if done with quality relevant content, almost certainly expands the range of search terms from which you'll draw traffic, and if wisely architected into your site should improve visitor satisfaction and potentially return and conversion rates. Blogs have helped some people turn at least one part of our websites into ongoing 'works-in-progress'. It would benefit all of us - marketers and site visitors - if more sites would continue to pick our tails up off the floor.

Posted by Craig Danuloff at 2:43 PM | Permalink

Slowly but surely the search engines are changing, and because of this and a number of other reasons, the principals and practices of search engine optimization are changing as well. (Or should be, I guess there will forever be folks out there worried about their meta tags...)

A huge aspect of the change going on in search optimization is the death of simplicity. The algorithms we're being judged by are lot more complicated. The level of competition has increased 1000-fold in just 2 or 3 years. And we no longer want a lot of visitors, we want qualified prospects with a defined intent at the precise time that our information matches where they are in their buying process. These changes and dozens more like them alter the concept of 'being optimized'.

Aaron Wall covers one of the issues forcing change in this thought provoking post, which includes the following:

I frequently get asked to look at a page to see if I think it is perfectly optimized. But I rarely think you can tell if a page is perfectly optimized just by looking at one page. Most of the optimized pages I am asked to look at have no clear goal at hand. Is the page meant to be link bait? Am I supposed to buy from the page? How is the page integrated into your site?

There's a long trail where this idea leads. I hope to explore it here in the near future.

Posted by Craig Danuloff at 12:39 AM | Permalink

HuckABuck is a new meta-search engine with the twist that it co-mingles results from Google, Yahoo, MSN, Technorati, Digg, and Delicio.us - and lets you adjust the priorities among those engines. One would assume other engines would be added eventually.

This is far from the first meta-search engine or twist on the idea. But it reminded me of a trend that I think is one of the most important in search, and yet doesn't get a lot of mindshare amidst the Google Algo Updates and link-spam discussions.

In the video world, the fall of the 'Big 3' has been exhaustively discussed under the 'tag' Exploding Television. In search, I think we're similarly coming out of the era of giants, but the fact has been obscured by the fact that the giants themselves are still growing unlike in the TV world where they were shrinking rapidly.

Google and the other engines themselves are contributors, with the 'invisible' tab splitting search into web|images|news|shopping|blogs etc. Vertical search engines such as those which have appeared in shopping, jobs, travel, weblogs, and other areas are another important component. Desktop widgets, toolbars (some of which aren't from the search giants), search built into the operating system, and a wide range of other solutions continue the fragmentation.

UPDATE: Blake from Huckabuck left a comment that pointed out perhaps the most dramatic explosion - personalized customized search. If each persons results are based on their past history, preferences, social network affiliations, or whatever, then the concept of being #1 on Google is really shot.

This is important to the online marketer because just as Tide can no longer 'talk to america' with 30-seconds on the evening news, you can no longer 'reach your market' by ranking #1 on Google. You've got to figure out everywhere that your category, brand, and company might be looked for and get yourself ranked and positioned in as many of those places as you can. This dramatically changes the nature of the task and of the resources you should put against it.

In practice many people don't even think about search in terms of the 'Big 3' but rather just focus on Google and take what they get from Yahoo and MSN. But it's time to re-define search to include all starting points and put the engines in proper perspective. Yes they drive the majority of traffic now, but as the shopping and travel verticals have seen it takes only 2-3 years for the 2ndary sources to become very significant.

There are several other important implications to this trend, which I'll comment on in future posts. Leave your thoughts on 'exploding search' in the comments.

[via Seth Godin]

UPDATE: The Huckabuck guys posted their own response and comments on the issues raised here over in their own blog.

Posted by Craig Danuloff at 10:39 AM | Permalink | Comments (1)

Ranking highly on Google for your top keywords is a priority for any search engine optimization effort. But what percentage of your efforts, budgets, and resources should you apply to that task?

Jakob Nielson provides strong anecdotal evidence in his new UseIt column, showing that his very popular website gets by far its largest chunk of traffic from Google, BUT the other sites that drove traffic combine to provide 35% more traffic than Google does.

Similarly, the top 10 search words and phrases account for a full 10% of his traffic, but 83% of the 110,399 terms that did drive traffic were only used once during eight weeks. And these single-time queries accounted for 3X the traffic of the top 10 terms.

The importance of multi-word terms and phrases is also highlighted, by a great chart that shows the distribution of 1-2-3-4-5 word searches. Short (one and two word) searches dominate the most active searches, but 'long tail' is filled with 3, 4, and 5 word phrases that in aggregate generate an enormous amount of traffic that would be missed if the site is not text-rich and at least somewhat optimized to rank for longer terms.

Look at your on analytics history to see if your results match those of UseIt.com. The lessons from his numbers are clear:

• Don't think about optimizing your site for just a small number of terms. That isn't how people search and it isn't where all of your traffic opportunities lie.

• Use extensive and expressive language to describe the subjects you cover. Searches will find ways to use even more varied language, and you have no chance to capture them if you stick with narrow 'approved' terminology lists.

• Work hard to rank well in Google, but also work hard to gain visibility and links at other search engines and at all kinds of other sites too.

I strongly recommend reading the entire column for more interesting and instructive information.

[via inside analytics]

Posted by Craig Danuloff at 3:05 PM | Permalink

There is no doubt that Google - the technology and the phenomenon - helped the internet. As the web grew they made it easier to manage, they took search seriously at a time when everyone else gave up on it, and for a while there in between web 1.0 and web 2.0, they almost single-handedly kept the hype of the internet going. But these days they have a lot to answer for.

Most significant is the fact that it's still far to easy to get lame or even completely fake pages into the index and the search results, and that they themselves help monitize these fake pages, which further encourages their creation. In other words, while Google says their goal is to provide access to the world's information, but what they are simultaneously doing is encouraging the creation of non-information on the largest scale in the history of humanity and then allowing that non-information to compromise their primary goal of clearly organizing the real information.

Nothing could prove this more clearly than this paragraph from a blog, prefectly enough, called Monitize:

Check out this site: search of eiqz2q.org — depending which datacentre you hit, you will see between 3.8 and 5.5 BILLION RESULTS. Even worse… the domain is EIGHTEEN DAYS OLD. That’s right, in under 3 weeks, one person has managed to get one domain 5 billion pages indexed in Google. And they are ranking, too. That particular domain has an Alexa ranking of under 7,000. Another domain owned by the same person, t1ps2see.com, has between 1.7 and 2.4 billion indexed pages and an Alexa ranking of under 2,000… after 4 weeks. Coincidentally, the sites also have 3 blocks of Adsense ads on each page. I wonder how much that one person is earning per day with billions and billions of pages indexed and ranking?

In english, this means that one person has created about 5 BILLION pages of essentially gibberish filled with nothing but text and links that make him money, and in less than a month become one of the top 10,000 most visited sites on the internet (according to the highly inaccurate Alexa, but it certainly means there is traffic being generated).

And he did it ONLY because Google (and I'm sure the other engines) take in his worthless pages and put them in the search results AND give him paid links to put on those pages to earn money off the people that Google sends him. You don't have to be a conspiracy theorist to wonder if Google may be encouraging their employees to spend 20% of their time on hobby-projects in order to slow down their progress in actually fixing these problems.

Of course, the Google Adsense monitization side of the equation isn't the only way the monitize junk pages. There are any number of networks that will pay sites for links generated from visitor clicks. And while the Monitize blog reports that Google Adense were on these pages, I can't see them now so perhaps the publicity has gotten this guy cut off. But rest assured there are hundreds if not thousands like him.

The problem has many dimension, and I'm sure there are many at Google who honestly try to police both sides of this equation. But as I'm sometimes fond of saying, the best you can do is not by definition good enough. It certainly isn't in this case.

BONUS LINK: Read about yet another way to bu