We dug in and explored the challenges of educating a large media company, dealing with multiple content management systems, syndicating content and query deserves freshness, all viewed through the eyes of ABC!

Interview Transcript

Eric Enge: Can you provide a brief outline of your responsibilities at ABC News?

John Shehata: I’m the Executive Director of Search for ABC News, and recently heading SEO for media properties like ABC Family, ABC.com, and Oscars. Everything SEO goes through our department.

Eric Enge: How many people are in the department?

John Shehata: We’re a small centralized team of 2. We also have “SEO Ambassadors,” which is a group of people that work within different groups at ABC, different shows or properties that help others with SEO as the first line of defense and make sure that SEO is top of mind. They bring us in if they have bigger questions or bigger projects for us to work on.

Eric Enge: The word “ambassador” is a lead-in to the next question. One of the issues with a larger media organization (in my experience, a traditional media organization) is there’s a constant need to educate on what role SEO can play in the growth and promotion of the business. Can you talk a little bit about some of the challenges and some of the wins, perhaps, that you have had along those lines within ABC?

John Shehata: Sure thing. I believe that 40% to 50% of our job is to educate and inform people on SEO best practices, and we actually spend a lot of time doing this. We create different presentations and tailor them to different types of audiences. Presentations to C-level are different from presentations to development or tech teams, which also differ from the editorial team.

We try to help them see why SEO is important to them, and what they will gain from doing it. We’re not telling people how to do their job, but really helping them to become more successful at what they do by applying the best practices of search engine optimization. We also stress that we really don’t optimize for search engines, but instead, we optimize for people who use search engines to find our content. Education is a big part of our job.

We educate, train, and we create best practices guides, documents, and one-sheets that go around, answer emails with information, send weekly and monthly newsletters and create wikis. If we get the question more than once, we create an FAQ. We have a lot of material out there for the different users, the different departments and the different properties.

Eric Enge: I imagine that the adoption and acceptance of what you’re trying to teach is like any other human process: some people get it faster than others, some slower. You’ll have people who aren’t officially your SEO Ambassadors that actually become like an ambassador, which must be exciting when it happens. Conversely, you’ll have others where you have to proactively keep getting back in front of them saying, “Hey. Don’t forget about this. This is how it impacts you.” Is that a fair characterization?

John Shehata: Absolutely. There are different ways of approaching this. One of the things we do is highlight the success stories. When you work on a big project with the dev or tech team or on a story with the editorial team, you highlight the results. People love to see their successes shared and to get credit for their hard work, People, in general, when they see successful projects, want to jump into the bandwagon and say great job; how can we do the same?

You will find people coming to you who in the past maybe didn’t acknowledge SEO as something important within their workflow. Seeing someone else’s success makes them want to try it out themselves. You’re not going to be in every department at every single moment, so this is a very good way of letting other people actually come to you asking for help. Highlighting the successes with different departments works very well.

Eric Enge: Do you use the same content management system for all the websites you’re involved in?

John Shehata: No. it’s a good question. When I first started with Enterprise SEO, 8-10 years ago, I thought everyone was using WordPress or MovableType, and I was surprised.

Most of enterprise level companies/sites have different CMSs within the same company and sometimes within the same site. In addition, each CMS uses a different technology; some proprietary platforms and a mix of small and enterprise technologies. You’re working, most of the time, with legacy code; sometimes languages that you never worked on before.

Many companies in the past used to create their own development languages, so you’re dealing with legacy code and legacy CMSs. You have to adapt because you can’t just go to an enterprise-level site and say, “Hey. Let’s change the CMS.” You have to adapt and find your way around it.

Eric Enge: I bet that each CMS comes with its own configuration challenges.

John Shehata: Absolutely, yes. Some CMSs don’t even allow you to change a page title or add new tags into the head section. There are a lot of challenges. The biggest challenge is to convince stakeholders and explain why they need to improve the CMS, code or even upgrade the CMS platform. Another big challenge is that many big companies customized their CMS to the point that they can’t do any upgrades because it will break all their customization.

Eric Enge: I imagine that some of these content management systems freely create duplicate content. Would that be an accurate assessment?

John Shehata: I can say that every single enterprise site I worked on in my career had major duplicate content issues. But, it is not only the CMS, there are many other practices that cause this:  inconsistent linking, sharing content on multiple sections for traffic credit, and sponsorship opportunities that require different integration and designs so you need to replicate the content in different ways to satisfy the sponsorships.

Sometimes, you also see internal tracking done incorrectly causing many duplicate content issues. People want to track what happens on the sites, so they add tracking parameters to URLs for links within the site. There are so many things that can cause duplicate content.

Eric Enge: There must be lots of situations where simply eliminating the duplicate content is not so straightforward.

John Shehata: True! Sometimes, there are strong reasons for the duplication that you can’t simply fix or change, especially the ones tied to sponsorship opportunities. So, while eliminating the duplicate is the best SEO solution, we often have to rely on alternative fixes like rel=canonical tags and XML site maps. You must work with the tech team to figure out the correct permalink URLs for each of these pages so you don’t end up duplicating URLs in your XML site maps or in the source code. This can fix a little bit of the issue.

You can also work in Google Webmaster Tools to eliminate certain parameters that create duplicate content. If you have some kind of sorting or tracking parameters, you can identify those in Google and Bring Webmaster Tools to be ignored; just be very careful so you don’t end up blocking a big chunk of the site.

Eric Enge: Are you confident, in your experience, that the rel=”canonical” really does a good job of dealing with the duplicate content? Have you seen measurable benefits from that?

John Shehata: We have seen some benefits, but it’s really a Band-Aid on the problem. In many cases, I believe the XML site map actually is a much stronger signal than the canonical tag. I think a combination, of different signals, or different places giving the same signal, will hopefully get Google to accept the page you designate as the main page.

In many instances Google continues to index and rank other pages, not the canonical ones, so the best solution is still to fix the site architecture. If this is not doable, doing rel=canonical, and XML site maps are good Band-Aids.

Eric Enge: Are you also involved in syndicating content?

John Shehata: Yes. Most of the major publishers either syndicate content on their site or syndicate their own content on other sites.

Eric Enge: What kind of challenges does that bring for you?

John Shehata: There are so many questions you need to ask before you syndicate your content. Are you syndicating content to other sites that are more authoritative, or have less authority than your site? When should you syndicate the content? Will they accept or implement certain tags on their pages to tell search engines where this original content comes from?

A couple of years ago, Google told publishers to use syndication source tags. But, most of the sites that syndicate the content on their sites don’t have the capabilities, or sometimes the resources, to add these tags dynamically, linking back to the original content.

Later, Google said to use rel=canonical tags. They told us that they would consider canonical tags to be more authoritative than the syndication source tags. So, even if the site syndicating your content has the syndication source tag, but the page has canonical tags pointing to their site, this will override the syndication source tag.

Some sites do delayed syndication, where they wait a couple of hours or more before they syndicate content to other sites, so Google can verify that this content actually belongs to their site, and it was published there first, before they see it everywhere else.

What we have seen is that even if you have all the tags pointing to your site but you syndicate to a site that has a much higher authority than yours, Google might end up actually ranking the higher-authority site.

Eric Enge: I’m no Googler, but when you syndicate content to a higher-authority site, I believe that from their point of view, the user might prefer to read it from the higher-authority site, and feel more comfortable there, than they do reading it on your site, even if you published it first. Since their goal is to maximize the satisfaction of users, while it may seem wrong to us, it might actually be the best answer to them to do that.

John Shehata: I totally agree with you. A lot of times, SEOs think of one specific tag, or one specific tactic to optimize, and then they don’t get the results expected. Google has hundreds of different factors and signals that go into their ranking formula. Syndication/Canonical tags are one of those factors. As you mentioned, if the higher-authority site has a higher click-through rate, more clicks into the site, users spend more time into the site; then Google may consider it as a better experience for the user versus the original site.

Eric Enge: Sometimes media companies let other media companies syndicate a synopsis of the article, rather than the whole article. Is that something that you use at ABC, as well?

John Shehata: We don’t do it this way; we do full syndication. ABC News, in general, is a high-authority site, and we make sure tags are implemented correctly. Even though we don’t use it, I think it’s another good tactic. Delayed syndication, partial syndication, and having a rel=canonical link back are all good tactics.

Eric Enge: As the Head of the SEO team, you have to integrate your efforts into a larger online marketing effort. Can you talk about that, as well?

John Shehata: We believe SEO is just one of the marketing channels every site has. There is paid marketing, social media, email marketing and so on. The SEO team needs first to understand the objectives and goals of the marketing department, of the site, and of the company as a whole, and then start working with all the various channels to meet those objectives.

For example, SEOs need to know what keywords the paid marketing team is targeting, such as what are the 10 most expensive keywords. Can SEO help there and save the company money? When you make other people successful, they absolutely work better with you. For example, money we saved can go to SEO or be directed to tools or development.

Social media is another marketing channel where I believe collaboration is important as SEO and social are becoming more and more integrated. Social media is becoming, or is about to become, an important SEO signal. SEO is just a part of the bigger picture of online marketing.

Eric Enge: With a large media company, I imagine that you get a tremendous amount of attention without having to do much outreach.

John Shehata: In general, that’s true. But, it differs from one brand to the other. For example, if you have a very big e-commerce brand, you will probably need to do a lot of outreach; product reviews, trying to get links, etc. For media publishers in general, people like to share, link, discuss and engage with the kind of content we produce already. In all cases, social media is very important for all types of brands to utilize effectively.

In addition, one social platform is completely different from the next. What you communicate in Google+ may not and shouldn’t be the same for Facebook or Twitter; the frequency, the timing, and other factors also need to be taken into consideration.

Eric Enge: I would picture, even for large media companies, the speed at which you can get the word out about the news you publish would have something to do with whether you win what I’ll call the ‘query-deserved freshness battle.’ There’s still some incentive to do outreach. It’s not like you need help getting in front of people, but it’s not a bad idea to get in front of people faster.

John Shehata: Absolutely. I think speed is very important factor, especially for breaking news, especially with QDF, query-deserves freshness. Taking that into consideration we never sacrifice accuracy for speed. Social Media is great help getting the word out there faster, for sure.

I believe that search engines are absolutely looking at the social rates for new Tweets, Re-Tweets, Likes, Shares, Pluses on Google+, or for whatever platforms that are opened for them.

I believe that search engines are looking at that, and some already have it as a part of their algos. If a story gets 200 Re-Tweets in 5 minutes, versus another story that got just 10 Re-Tweets, it makes a difference. I think search engines do consider these factors. Google also has sharing buttons now in Google News, so they also consider how people Like and +1 the news on Google News, as well.

Eric Enge: Thanks for joining us today.

John Shehata: Thank you so much for having me, Eric.

Closing Notes

There are many things that make Enterprise SEO particularly challenging at this scale. Some of the biggest ones are the inherent complexity of the organization, the many different business objectives, and the number of different properties involved.

John faces all of these on a day-to-day basis. But, on the other hand, there is tremendous leverage to be gained. For example, at ABC, the massive marketing machine is a big advantage; once you get it pointed in the right direction, look out!

You can see more of John Shehata’s thinking on Enterprise SEO here. Below is an example of what you will see there:

In closing, I hope this interview has offered plenty of insights for enterprise SEOs. If you have any comments, please post below.

SEO Challenges: New Site Launch Or Major Updates

1. Too many decision makers. This is one of the more basic headaches. Product marketing, engineering, PR and the executive branch can all weigh-in in ways that can have you ripping your hair out.
2. Many development priorities. This is often the most frustrating challenge, because it is the hardest one to combat. Other priorities may well be more important in the short term than what you are pushing for. Or, worse, it just makes it harder for you to get key players to buy into your idealistic view of the best way to build the new website.
3. Ignorance of SEO. You still run into those that think that SEO is unimportant. Yes, they are still out there. I dug into this a bit in my article, Why Do Brands Overlook the SEO Opportunity for Non-Branded Keywords?
4. Misinformation About SEO. This one is worse. The exec who thinks they know something, but the information is wrong, just might drive you mad. Just last week at SMX West, I was talking to one exec that told me they had just finished updating the meta keywords on 1800 to their pages. To them, this was SEO. The lost time is certainly one bad part of this, but worse still, to this exec, the SEO for the site was done. (I took it upon myself to set them straight.)
5. Misconceptions. The belief that SEO is Hopeless.

The above misinformation deserves its own brief discussion. Many believe that Google changes the rules at its whim, and hence, no SEO effort is sustainable. They have not yet bought into the basic concept that modern SEO is not about tricking Google; it is, in fact, a form of branding activity that uses many tactics that support the overall brand in ways well beyond pure ranking in the SERPs.

There are many other types of problems, but you notice that none in my above list have anything to do with the actual execution of the new website project. Let me hit you now with the stunning conclusion: the real opportunity for you occurs long before the project planning starts.

Ideally, it starts months before. That is the timeframe in which the game will be won or lost. You want to get out in front of this before the people involved begin to generate any level of activity on the project at all. If you know that the plan is to begin working on a new site in July of 2013 (or even October of 2013), the time to begin doing your work is now.

Getting the best result depends on educating the people involved, and getting their incentives aligned properly. (You can read some thoughts on how to help with the education process in my articles: Selling the Benefits of SEO in a Large Enterprise, and Getting Top Management Buy-in for Enterprise SEO).

In addition to what you see in those two articles, below are two more ideas about speeding up the education process.

Show Them Examples Of Failure

I wrote about one example of this in my last Enterprise SEO column. If you search on the generic term [diapers], Pampers and Huggies are not part of the first 6 organic SERPs:

As you can see, Pampers and Huggies are far and away the biggest brands in this space but do not rank prominently. The same thing happens if you search on another generic term such as [aspirin]:

This happens with major brands for all kinds of generic search queries. In addition, if you dig into it hard enough, you can find interesting case studies such as this one Bryson Meunier wrote about on a Mobile SEO audit. Launching a new mobile site? You can use this article to talk the team out of using a transcoding approach.

Use Your Own Analytics

Your execs might point out that you are getting lots of traffic to your site. However, traffic is not the same as organic traffic on relevant non-branded keywords. The purpose of SEO is to get you traffic on these types of keywords, anyway.

As a first step, use analytics to show them what percentage of that traffic is organic. Then you can dig into it a little further to show them the search query mix. Hopefully, it does not look like this:

If it does, you are not doing well in your SEO! Your non-branded search query volume should dwarf your branded volume. Another way to get a look at this same problem is to look at the landing page mix. It is not good news if it looks like this:

Takeaways For SEO Project Planning

The key is to realize that you need to get way out in front of this. By the time people are trying to schedule the project planning meeting, it is way too late to make major changes in the way they perceive SEO, or perceive the priority of SEO.

Changing this type of mindset is something that takes many months, and lots of data. No matter how you start the exercise, and no matter how good you are as a teacher, different people get stuck in different places, or will raise different objections.

You need time to find out what these objections are, and then you will have to find ways to address them. When you are done with the first objection, you will run into the second one, and the process will repeat itself again. In addition, you are likely to have only limited time slots in a given month.

For example, if you get to meet with the development VP next Monday, you may get 10 minutes to insert a bit of your thoughts on SEO. That may be the only time you get with her/him until next month. If you have several objections to work through, it can take many months. Plan on this and work the process long before that new website project ever hits the drawing board.

The further out in front you can get started, the better. And, be prepared to exercise lots of patience!

For example, one brand we work with is only concerned about ranking for its brand names (both the company and individual products). To them, that is the sole purpose of their SEO. This is not quite as trivial as it sounds, as the company has a large scale affiliate program effort, where many of their affiliates are quite aggressive at SEO. If they do not put the effort in, they can get outranked for their product brand names.

A Practical Example

Let’s take a real world example. If I use the AdWords keyword tool, I can pull a list of the search volumes on the largest brand names in diapers:

click for larger image

As you can see, the top 2 are Pampers and Huggies. Both sport impressive search volume! This is great stuff.

Since these are the top brand names in the diaper biz, let’s take a look at how they fare in a search for their key non-branded search term, “diapers”:

The screen shot shows the top six search results, and they are nowhere to be found! In fact, Huggies.com comes in at #7, and Pampers.com does not show up until the 3rd page of results. Not a good thing! Interestingly enough, the story is reversed if you search on [diaper], where Pampers comes in at the #8 spot, and Huggies is not on the first page. Still not good.

What’s The Opportunity Cost?

One way we can estimate the opportunity cost is by looking at the Google AdWords tool phrase match search volume for the major brand terms vs. the non-brand search terms:

The phrase match volume shows us the total volume of all the search queries that include the word (or words) inside the quotes. This can give us a rough estimate of the long tail search volume associated with the word or phrase. Clearly, there is a lot more long tail volume associated with the non-branded terms than there is with the branded terms.

You could argue, and you would likely be right, that the conversion rate on the branded terms will clearly be higher, and that a significant percentage of the non-branded terms long tail volume will come from things that do not apply as directly, such as [diaper bags].

However, all of that non-branded long tail search volume represents a branding opportunity that is being missed in a big way. Note that if someone searches on your brand name, they already know it, and they already have developed a certain level of interest in that brand.

If they search on a non-brand name, chances are that they don’t have that same level of commitment. What we see in the above data is 2.5 million branding opportunities. You could also argue with me that you can get more impressions per month by other media. While that would be non-trivial, you could do that, but you would be missing the point.

Nearly 100% of the search impressions are from people who have a direct and immediate interest in diapers. Granted that some of the people who search on phrases like [Diaper cakes] may be researching cakes for a baby shower, but even some of them may buy diapers for the event. Nowhere else are you going to get anywhere near the same volume of people with such a focused interest.

In addition, there is a lot of business to be had among these keywords. This is all incremental business since the search query started with a non-branded phrase.

What Can We Learn From This?

The basic branding campaigns that major brands pursue bring huge benefits. I remember when I was first learning about marketing and branding. One of the key concepts I learned was the notion that it took 7 branding impressions to create a sale. Seems to me that there are a ton of those impressions still available in search. Right now, diapers.com is reaping far more of those benefits than the major brands.

Algorithm Update Regularity Increasing

I have been known to say that 2011 and 2012 marked a major transition in Google’s spam fighting capabilities and mindset. Panda, Penguin, EMD, DMCA/Pirate, & Page Layout / Top Heavy Ads all had one thing in common: Google used offline analysis to identify potential ranking adjustments and then fed these results into the main algorithm. This is why those updates happen periodically, causing Danny Sullivan to tell us that The Google Dance is Back.

In my view, these few algorithms represent the tip of the iceberg. By my count, there were 9 such algorithm updates in 2011, and 21 in 2012 — more than double the 2011 total. It is more than reasonable to expect that we will see much more in 2013.

While I went public with a stance against link buying in 2008 (I actually bought my last link in 2004), this notion is much more commonly accepted today than it was then. But, we need to think more deeply than that today. Google will continue to improve its capabilities in fighting bad link building practices.

To illustrate, let’s look at the number of Panda releases over time vs. the number of Penguin releases:

As you can see, the pace of Panda releases had accelerated over time, and we can expect the same thing to happen with Penguin. This means more and more updates targeted at punishing or devaluing poor link building practices.

Google is using its expanded spam fighting capabilities to push consistently toward one goal for its organic results: include in the SERPs the webpages most likely to make the user happy with the search results.

Search engines need to evaluate two types of metrics to make these decisions: the relevance of a page to the query, and of all the pages relevant to the query, which ones are the most important/valuable to the users conducting the queries? This type of analysis is performed on a query-by-query basis.

If the search engines evaluate on a query-by-query basis, it stands to reason that they evaluate the authority of a website in response to a query in a manner that is specific to that topic area. In other words, high level website authority is not sufficient to cause a site to rank for each product area that they cover.

Enterprise Branding & Link Building

A lot of major brands cover many types of different products. Proctor & Gamble lists 47 different brands on the All Brands page on its website. Even Ford Motor Company lists five major categories of product lines on its home page. Digging in a bit further, they list six major lines of cars, ranging from the inexpensive Fiesta, to the sporty Mustang, to the pricier Taurus.

(Did you notice how those two companies each got a “free link” from me just because they are major brands?) However, let’s look at these links, and the pages they point to. Would a user, or a search engine, use those links to decide that Ford is the best company to buy a sports car from, or that P&G makes the best laundry detergent?

Of course not.

We can learn from the example of the offline advertising world. Does P&G run ads to promote that they are a large consumer goods company? Again, not really. What do their brands do instead?

1. Old Spice runs ads to tell you that they are the best deodorant.
2. Tide runs ads to tell you that they are the best detergent.
3. Ivory runs ads to tell you that they are the best soap,
4. … and so forth …

Ford does the same thing with its line of products. They have one set of ads for trucks, a different set for SUVs, and another for economy cars. These ad campaigns communicate to consumers both the relevance and importance of a product line. This type of campaigning is the logical equivalent of deep link building. You establish relevance and importance on a product line by product line basis.

For fun, I decided to see how many ads were posted on YouTube for each of their car product lines. You can see it here:

My technique of using these search queries is admittedly crude, but the point is that they run ads for each of their product lines. Your online strategy needs to reflect this, too. Search engines try to look at signals the same way that users do.

One set of such signals is the links that your product category pages get. This goes back to the traditional model of links being seen as academic-type citations for your content.

So when you look at one of your product lines, some questions you can ask yourself are:

1. Is my product covered in reviews of similar products in major magazines?
2. Do major websites covering similar products write about your product at other times (not just reviews)?
3. Do lots of bloggers who fit your target demographic write about you?
4. Is there any evidence online that your company is an expert in your product’s topic area?

There are plenty of these basic types of questions. Let me enhance them one step further. Do the places that write about you online include links to the relevant pages on your site? Note that someone writing about you without a link could represent some level of endorsement, but implementing a link, which offers the user the opportunity to leave the site publishing a link is, by definition, a stronger endorsement.

The need for deep link building, or product line by product line promotion, is not going away. For enterprise sites, this will require a similar commitment to online promotion on an item-by-item basis that mirrors what they do (or have done in the past) in promotion offline.

The Complexity Of Topics

It has been stated many times by the search engines — there are hundreds of different factors involved in determining rankings for search queries. The complexity has only grown over time, as the search engines constantly seek out new signals for improving rankings. There are many ways that this happens, but one that people often forget is that there are many ranking algorithms which are scenario based.

Here are just a few simple examples of scenario based algorithms:

1. If you search for [Pizza], the search engine may show you local results. This is true for all kinds of locally based searches.
2. If you search on [hotels] and your prior search was for [Rome], you may see some results for [rome hotels].
3. If you are passionate about music, and you search on [jaguar], you may see Jaguar guitars in the search results.
4. If you conduct a search while on a Smartphone, you may get more action-oriented results.
5. If you follow someone on Google+, something they have shared or +1′ed may be raised in rankings on Google.

Scenario-based algorithms are just one aspect of the complexity of search. New signals become available all the time. One of the most notable ones is rel=author, which I predict will become the new ranking signal story for 2013.

It is a high-quality signal indicating which authors write stuff that people like. Bing is conducting its own major experiment with its social sidebar, which offers an entirely new paradigm for integrating social media into the search experience. You can expect a lot more from Schema.org in 2013, too. These types of changes are taking place all the time.

There Is Great Value In Working With Many Sites

As readers of my writings know, I am not at all a fan of chasing the algorithm. I believe people should focus on producing great content and promoting it effectively.

That said, it is still a great value to having the experience of publishing on many different sites. This offers both SEO and traditional marketing benefits. You get to see what works!

One client might be inclined to try one type of promotional strategy, and a different organization may do something else. One company may be aggressive about testing a new rich snippet markup and be willing to invest in it long before other clients are.

One client might be struggling with implementing something on their website that someone else has already mastered.

When you try to do everything in-house, you are forced to deal with all of these challenges on your own. You need to learn everything. You need to do everything. You need to be right nearly all the time. You don’t have a lot of time or chances to get it right. Ouch.

Don’t get me wrong. In my view, any large enterprise should have an in-house SEO expert. It just makes sense. At a minimum, this will provide you with someone in a great position to be an internal SEO evangelist. This is worth his or her weight in gold. Seriously. They can also interact with internal teams and answer a great deal of questions without any outside help. They can also help build a culture of SEO awareness.

But, if your in-house SEO is a good one, he or she will recognize the value of outside help. They know that one person, or one team, working on one website, can’t do it all by themselves. There is simply too much complexity in SEO today, and it is not going to get better or simpler. The sophistication of the search engines is only going to increase, and so is the challenge for the average Web publisher.

In my view, the ideal enterprise SEO team includes both an in-house person or team, complemented by an external team. Hire the right external team, and they will help you understand the current best-of-breed practices and keep you one step ahead of the curve.

Like any other company, the management team in a large enterprise prefers that everyone work on projects that potentially offer a large scale impact. This can be very challenging on e-commerce sites that cover dozens, or even hundreds, of categories of products.

Here are two typical business case scenarios for a given e-commerce site:

In this example, I have compared a very aggressive plan of doubling the traffic to one category with the much more modest goal of a 5% traffic gain from a horizontally focused project applied to the site as a whole. You can see why the management team is going to want to focus on site-wide projects — the potential impact is nearly always greater.

A common scalable design approach for large scale e-commerce sites is to roll up the most popular products and show those on the main page of the category. However, this does not lend itself to either SEO or landing-page optimization of the experience. Note that if you believe that landing-page optimization is an SEO factor, as I do, then that is also SEO at work.

The challenge, in a nutshell, is to sell management that approving a project to redesigning individual category pages is a worthwhile effort. There is a lot of custom work involved for each category so you can’t do them all at once, and each individual project is not going to move the overall traffic needle by a large percentage.

To succeed, you have to sell the concept of going through the major sections of the site one at a time.

Demonstrate With An Example

As is so often the case in enterprise organizations, there is great power in selling a test project to prove a concept. Management teams respond well to real test data. If you can sell one test project to show that you can impact the traffic for a single category in a significant way then you can prove that the design concept works.

That leaves you with the follow on challenge of selling more projects over time, but because you have shown with a test case that the redesign does work for an individual category, at least one potential source of risk will be off the table — that being that management will question whether the redesign concept even works.

Of course, there is a process to selling the initial project. Here are some key aspects of how you do that:

• Manage Expectations: the goal is to show a concept of what can be done over time with a series of projects. Strong category experiences can create substantial leverage for those horizontal projects.
• User Experience: testing and surveying our users can help build the argument for creating a better browse path and discoverability for customers.
• Brand Value: better user experiences on key landing pages for the site can create a different perception of the quality of the site.
• Search Engine Value: search engines try to reward sites with better user experiences.
• Horizontal Projects are Finite: there are only so many of those you can do before you start getting more marginal returns.

I had the opportunity to discuss this with one large e-tailer recently. Unfortunately, I can’t disclose who that is, but the data below is real:

1. Category Re-Launch June 2012
2. Doubled traffic to the category within 2 weeks and doubled again within 1 month
3. Increased rankings for head terms within 1 month:
   • Top Head Term: moved from #22 to #8
   • Key Secondary Term: moved from unranked to #2
   • Another Head Term: moved from #26 to #8

Summary

With results like these, you now have proof that the changes can help at the category level. The results from this example were spectacular (4x traffic growth to the category), and you probably should not expect that to be the norm.

By itself, the the total incremental traffic for one category may not move the overall needle at a corporate level; but, now it becomes a lot easier to sell the concept of repeating the project.

Imagine if over a 2 year period you updated all of your top categories? Strong traffic growth at that point will move the needle in a big way at that point. In addition, you will have impacted the perception of your brand in the process.

There are always reasons why they want to continue to do what they are doing. That the current methods still work for them is usually the most basic one. They go out and place a bunch of low-quality blog posts on low-quality sites with low relevance, and with rich anchor text, and they see rankings for the related terms rise.

The reason why it may still work for them is that larger brands have enough overall link equity to protect them from the current implementation of the Penguin algorithm. On top of that, the people who are making the decisions have short-term financial goals (often the companies in question are public), and the key people up the management chain don’t have enough visibility to understand why they shouldn’t keep doing what they are doing.

For them, falling behind aggressive competition while trying a new tactic is a bigger risk than being penalized by Google. New, less manipulative tactics seem like they might bring about new traffic slower. At least, that is the way it seems now. The unknown is always scary. So how do you create the mindshift? Great question! Let’s dig in and talk about ways to do just that.

Recent Google Algorithm Update History

I am not going to recount the entire Google algorithm change history for the past two years, just the initial release of the major new types of algorithms that came out:

1. Page Layout / Top Heavy — January 19, 2012
2. Panda 1.0 — February 23, 2011
3. Penguin 1.0 — April 24, 2012
4. Pirate — August 10, 2012
5. Exact-Match Domain (EMD) Update — September 27, 2012

These major updates (the “New Algorithms”) all have one thing in common that distinguishes them from the updates that Google has made over the past few years — they all occur at periodic intervals. Danny Sullivan refers to this as the return of the Google Dance (all Google’s algo updates used to be periodic in nature, and the industry used to refer to this as the “Google Dance”).

The Panda algorithm has been through numerous updates, and Penguin has had two total updates since the initial release. Each of these appears to require calculations that are separate and distinct from what is going on in the main Google algo, which is dynamic in nature. We can visualize this process simply as shown here:

I did ask Matt Cutts about this new structure representing an increase in capabilities for Google in my recent interview with him, and here is what he said:

“… you are right that these algorithms do represent new types of capabilities for us.”

We have seen five major, new types of algorithms in the past 20 months — how many more to come? How will the existing ones be updated? I think that it is reasonable to assume that these new types of algo updates will keep coming, and perhaps with increasing swiftness. As long as there is a manipulative SEO practice that is causing the Google algorithm to do things they don’t like, the updates will keep rolling.

OK, but what do I do?

That’s today’s million dollar question. Here are some ideas:

Strike The Fear Of Google Into Management’s Hearts

It is well known that people respond more to the fear of losing something they already have then obtaining something new. Examples of victims of other Google actions are good. After all, no one wants this headline:

Of course, the naysayers will point out that history shows that JCPenney is the exception, but certainly not the only victim of these types of moves by Google:

However, the more important point to make is that these articles show Google in a bad light, and they end up putting public pressure on Google to prevent these things from happening. This means that Google needs to do more of this type of cleanup work.

This is what the “New Algorithms” are all about. They represent a new wave of attacks on these practices. Over time, they will use these new capabilities to eliminate more and more of the SEO tactics they don’t like.

Show Management The Upside

Having shown them the risks and made your case that these tactics can be harmful to your business over time, now it is time to show them the upside. How do you do that? Start talking up the positive traditional branding impact of a high quality link building campaign.

There are many ways to do this, but consider the value of obtaining a guest post placement in a high quality side such as Techcrunch:

Or the New York Times:

That should get people’s juices flowing. You mean I can get SEO value and visibility in high-profile journals such as those? You sure can.

Also, consider pitching them on the value of a social media campaign where you deliberately build an engaged audience that interacts with your brand. Then further, take that idea down the patch of using social media to build and reinforce relationships with major influencers in your industry (or closely related areas).

You just might end up creating the type of synergy shown by this graphic:

Many types of ideas like these are possible with a little bit of brainstorming. Generate some ideas, model them, and then use them to pitch the upside of pure, white hat promotional methods.

Wrangle A Test Budget & Run A Proof Of Concept Program

Unfortunately, in a lot of cases, this will still not be enough. The fear of changing something that is working now is strong. Worse still is the fear that your competitor will continue to cheat and get away with it and get an edge on you. That is a hard one to combat, especially when it remains possible to find sites for search results relevant to your business that are driven by questionable practices.

But, there is still a way to move forward. Nearly any senior executive is going to understand the arguments made above. Most of them are going to have an interest in doing things in a long-term, sustainable way.

Consider pushing them for a test budget. For six months, run a test where you run a true white hat campaign. For example, perhaps the current plan involves placing low-quality guest posts on low-quality, low-relevance sites, and jamming them with highly targeted anchor text.

Take a portion of that budget, and reallocate it to doing some high-quality guest posts on high-quality, high-relevance sites. If they let you use some rich anchor text, that is fine, but if they don’t, don’t sweat it. Other types of test campaigns might be better for your business, but you get the idea.

In the work we do at STC, these types of campaigns show consistently positive results. Chances are that it will for you, too. And once you can show that doing it the right way can lead to success, you should be off to the races. Then, your competitors who don’t change their methods will be carrying all the risks, and you will be competing effectively in the short term without carrying those same risks.

Isn’t that exactly the way you want it to be?

Some common scenarios where this happens are:

1. Different product lines which are owned and operated by different business units.
2. International sites operated by local country based divisions / subsidiaries.
3. Marketing sites designed to run separately from the main company sites.

This brings up a host of problems, including:

• Some internal groups may resort to questionable SEO practices to get traffic, and to make matters worse, they may even be unaware they are doing risky things.
• There may be a lack of consistency in the approach used to SEO.
• Lessons learned are not shared across the various site owners.
• You may face lots of overhead in trying to prevent the first three items on this list from happening.

It goes without saying that you don’t want to wake up to this nightmare:

Avoiding this type of catastrophe is reason enough to get some controls in place, but there are many other reasons which we will get into a bit later in this article. First, let’s dig into how you structure a centralized group.

Setting Up Centralized Practices

I am a fan of having a corporate SEO team that helps establish policies in a way that is both understood and agreed on by executive management. This is where, hopefully, clear guidelines can be set on policy items, such as what types of practices for obtaining links is considered spammy, or how SEO teams work with engineering or PR.

The corporate SEO team can then work with vertical SEO teams specific to each business unit / local country subsidiary / marketing group (we will call them business units for simplicity in the rest of this column) that owns a website.

It is important to not carry the notion of centralization too far. There are valid reasons for a certain amount of local control. If your corporate SEO team is based in California and you have a country-based office in Taiwan, there is a need to be able to make some decisions locally simply so actions can be taken quickly.

In addition, there are some specific things that will apply to search in Taiwan that are not at all intuitive to someone based in California.

The same is true across product lines. Certainly, the keyword space for each product line will be different, and the business unit that owns the site needs a certain amount of independence.

To ensure long-term success, the corporate SEO team must work at building constructive relationships with the vertical SEO teams. I can tell you from direct experience that the vertical SEO teams will rapidly develop disdain for a corporate SEO team that simply tries to mandate policy without their input.

One key step is to take into account the unique challenges the individual teams face and act as a resource to those local teams.

To help build these relationships, the corporate SEO team should work together with the vertical SEO teams in a variety of areas. Some examples include:

• Get input on corporate SEO policies.
• Be available as a resource to answer questions on best practices.
• Provide access to SEO tools to facilitate the efforts of the local teams.

Once this atmosphere of internal collaboration is established, there are a variety of areas that the corporate SEO team can be involved in.

Some example areas include:

1. Driving the selection of enterprise SEO tools that are used corporate wide, thereby enabling standardized reporting on SEO results.
2. Creating and maintaining a corporate-wide data repository for all the SEO data collected over time (the “SEO Database”).
3. Potentially implementing applications to come up with reports drawing on the information in the SEO Database.
4. Working with the executive management team to set policies to avoid risky SEO practices, and publishing those policies within the company.
5. Defining a standard for implementation of title tags and meta description tags. You can take this further and create a automated default metatag template as advocated by Yahoo’s Markus Renstrom.
6. Establishing a standard to ignore keyword metatags (except for the news keyword metatag that Google just announced).
7. Evaluating new announcements by the search engines and providing information to the vertical SEO teams on the potential impact.
8. Creating standard page templates for common page types that are to be used by each vertical SEO team.
9. Outlining standards for link building practices, both to minimize the risk of manipulative link building and  also to provide leverage.
10. Assisting in the integration of the SEO efforts with the social media efforts of the company. These areas should not be seen as independent activities!

The Takeaway

A properly structured central SEO team can bring a lot of leverage to the diverse business units across the enterprise. It is wasteful to have each independent vertical team try to become an expert in everything, and chaos can ensue if you don’t have some basic, agreed upon controls and policies. Once you have established your corporate SEO team as a trusted resource, you can effectively establish policies that will bring you many benefits.

Many enterprises have difficulty implementing a strategy to build up the visibility and expertise of key individuals on their staff. So today I am going to walk through the types of problems that come in the execution phase.

Why An Authority Strategy Is Important

There are many major reasons to pursue an authority strategy. These include:

It is great branding. When your organization is seen as having people who are recognized experts in their field, it reflects well on the entire organization.

The rel=author tag is a way for authors (and your business) to get recognized in the search results for the content they produce. It certainly increases your chance of getting a click through when a searcher sees the author’s smiling face next to their article in the search results:

Many have also speculated that over time the concept of author authority could become an important SEO ranking factor.

For example, an individual who writes a regular column for the New York Times might start to be seen as an authority. If that same person starts to write articles for a lesser known site, they may carry some of that authority with them. This could impact the visibility that Google gives to the articles they write for the lesser known site, and may even lend some authority to that site on an overall basis.

In addition, building author authority is an important part of a much broader link building strategy involving developing relationships with influencers and your industry community. By developing those relationships and actively sharing high quality content, people are more likely to share and link to your stuff.

Organizational Steps Required To Execute

Understanding how an authority strategy can help you is great, but successful execution is non-trivial. At STC, we have worked on helping many enterprise organizations set these up. What we have learned is that there are many places where it can fall apart.

Here are some of the key steps you need to go through to make sure you can start building such a process within your enterprise.

1.  Identifying an author. Your author may or may not actually create the content, but you need to have the name of one or more people who will actively lend their identity to the cause. You are dead in the water if you don’t nail this one, yet many organizations have trouble with it.

A key point is that even if the person is not the original creator of the content, they need to be actively engaged in the process. Don’t view this as a completely outsourced activity. People like to engage with people, and your named author is going to have to commit some real time to this process, even if you give them lots of help to make it easier.

2.  Content team identification. You need one or more people to create content that is up to the quality standards required, and of course, that fits the content plan. Bear in mind that you can’t have your content creator be someone who knows nothing about the topic, or without passion for it. This is true even if they are not your named author. The more your content creation team is engaged, the better.

3.  Understand quality requirements. I always discuss this a separate step, as the organization really needs to understand that this is a brand building activity with huge SEO benefits, not some activity that takes place on the “dark corners of the Web” where no one goes. High quality is an absolute must. The quality is a key driver in establishing authority.

4.  Develop a content plan. What type of content can be created that is brand building, drives links, and establishes authority? Writing a great article on a topic that has been well covered a dozen times before won’t cut it. If you are not prepared to create really innovative content, don’t bother starting the effort. Also, how often do you want to publish?

5.  Rel=author mechanics. This has a modest amount of impact on the development team, so you do need to get the mechanics right. Refer to the aforementioned article by Rick DeJarnette for a great guide to this.

6.  Content creation. Once the plan is in place, go out and execute the plan.

7.  Content review and approval. Have this setup to make it painless. Simpler if the author in step 1 is also the writer, but if not, be prepared for some heartache and tuning in the early going. Make sure the named author has some time allocated to do this properly. You need this as your final quality control step.

8.  The promotional plan. Now that you have everything in place to create killer content from an authority, how do you let the world know? I won’t try to outline how to do that here, but simply identify the need and that someone needs to own that. One tidbit I will include is to define a role for guest posting on authoritative third party sites to help develop the author’s authority (and get those sites to implement rel=author for your author as well).

Building Authority Is A Process

Not every company is ready to buy into all this, but this is what it takes to be successful. For enterprises, budgeting up front is usually a big deal, so the earlier you walk people thorough these steps the better! Make sure they know what is involved before they commit, and success in execution becomes a lot more likely.

In a year or two, as tracking author authority becomes more mainstream, we believe that it is going to be a serious ranking factor. Combining rel=author with a frequent publishing schedule on authoritative sites and a strong social media program is going to be a fabulous way to build visibility and traffic for your business. It will bring great benefits to you well beyond any SEO implications. The time to get going is now!

Add to that the frustrating reality that the person/people setting the priorities often do not have nearly enough information to decide which projects could have the biggest impact for the organization.

This can leave you in stunned shock at a conference room table when that killer project you were pushing for did not make the cut in the development schedule for the next 6 months.

A few key steps can help you minimize the risk of this happening to you. Of course, there is the usual tactics of making friends with key people in the decision making process, and buying the developers pizza and beer. Those are excellent ideas, and you should do them, but let’s dig a little deeper into the process.

Determine Your Non Branded Organic Search Traffic

Being definitive about the potential economic gains is always impossible to do precisely in SEO, but that does not relieve you of the responsibility of doing the best you can. The first step is to get a handle on the current level of non-branded organic search engine traffic (“NBOST”) that your site gets.

The reason we focus on NBOST is that your site should not need any help from your SEO efforts to rank for your company brand name. Google and Bing are very good at ranking your for your brand even with no real SEO efforts, while your SEO is best measured using NBOST. Obtaining an accurate NBOST total is a little tricky in the era of “not provided”, but it is still doable.

Follow these steps and it will get you there:

• Determine your total organic search traffic (“Total Organic Search Traffic”) in your analytics tool. In Google Analytics the steps are: Traffic Sources -> Sources -> Search -> Organic

• Look at the incoming search phrases and see how much of your Total Organic Search Traffic for which the keyword reported is “not provided” (“Total Not Provided Traffic”).
• Calculate the traffic for which you have keyword data by subtracting Total Not Provided Traffic from Total Organic Search Traffic. We will call this “Total Keyword Specified Organic Search Traffic”.
• Find out what portion of Total Keyword Specified Organic Search Traffic comes from brand terms (“Brand Specified Organic Search Traffic”) by reviewing the 100 most popular keywords and counting it up manually or by setting up a filter in your analytics tool.
• The Total Keyword Specified Organic Search Traffic which is non-branded (“Non-Branded Specified Organic Search Traffic”) = Total Keyword Specified Organic Search Traffic minus Brand Specified Organic Search Traffic.
• The % of your overall traffic that is non-branded (“Percent Non-Branded”) is just Non-Branded Specified Organic Search Traffic divided by Total Keyword Specified Organic Search Traffic
• NBOST = Non-Branded Specified Organic Search Traffic + Percent Non-Branded times Total Not Provided Traffic

This formula uses the percentage of non-branded traffic you do get (in the portion of your traffic where you can see the keywords) to estimate what percentage of your Total Not Provided Traffic is likely to be non-branded.

At Stone Temple Consulting, we did some validation of this methodology using several methods. In one example, we saw that the mix of landing pages for the not provided traffic seems to be very similar to the mix for pages for which we can get keyword data.

The above text based formulas may seem complex, but it is easier than it looks in print, as shown in this sample calcluation:

Estimate Growth Potential For Your Project

This is where some magic occurs. There is no accurate way to forecast SEO gains. And yet, this is a crucial step.

Let’s say you convince your team that NBOST will double based on a given project, and in the next step, you show them the fabulous wealth that will come in once that happens. So they do the project and 6 months later NBOST is up just 5%. You don’t want to be that person, because they will never listen to you again. Your credibility will be shot.

Of course, you need to put enough of a forecast on the table that it is motivating to them.

Note that if your prior forecasts of growth were all in range with, or underestimates of, how you set expectations, and they did the project based on those forecasts you job is much easier now – they will trust you are not just blowing smoke.

That said, how do you some up with a forecast?

Here are some tips on how to go about it:

1. First determine what portion of the site is affected. If the project concerns one section of the site, and that section gets 35% of the NBOST, that means your project will have little to no impact on 65% of the NBOST. If you double that section’s traffic, you are increasing overall site NBOST by 35%. Breaking down the problem this way will help you better understand how to scale the possible impact.
2. If you are adding new pages to the site, do the keyword research to determine the potential scale of traffic that could be added. Then scale that back by the time it will take to create the pages and for the search engines to rank that content. Also, do not assume position 1 placement, but assume something more conservative, like position 3 or 4 for less competitive keywords, and lower positions for hyper-competitive keywords. Use industry data to determing CTR by SERP position. You should also take into account the long tail, which I usually do by taking the traffic I see for the major head terms and multiplying by 10.
3. If you are adding new content to existing pages, such as UGC or newly generated written content by staff writers, understand the schedule by which this will occur. UGC can take many months to have a real impact, and your staff writers need time to generate content too.

These are just a few examples, and there are way too many types of SEO projects for me to model each type. But, the key realization to start with is that you must do this.

If you can’t come up with a way to make reasonable guesstimates of the impact, you can’t expect long term success in selling your projects.

A final tip here is to provide a range. Hopefully your team will know that there is a large dose of guessing in SEO forecasting, so aim to give them a range of possible results.

Ideally, the low end of your forecast is high enough to still be interesting to do, and also covers your downside risk on the project potential. If it doesn’t, then the question is: why are you doing the project?

Of course, sometimes you have to do one project to enable others, so one specific project may offer little in terms of direct gains, but then it needs to be setting you up for future NBOST and conversion gains down the road.

Estimate Revenue Impact

This is a little bit easier to do, and just as important. Your website exists for a reason, and that presumably includes trying to get visitors to do something (a “conversion”).

It may be as simple as visiting a particular webpage, requesting to be contacted, downloading a white paper, completing a purchase, or any one of the above. Whatever it is, the action you are looking for them to complete has value to you as an organization. Otherwise, you would not have a website.

Make sure you understand what the potential conversions are, and the value for each. Some transactions may not have an obvious value, such as downloading a whitepaper, but if that is what a conversion is for you, work internally to agree on what the value is per download.

The other number you need to get your hands on is the conversion rate for your site. Ideally, you want this number specific to the conversion rate that results from NBOST.

If you can’t get this number because analytics is not setup to track it, the next development project you push for needs to be to get analytics setup to track it. You can’t improve what you don’t measure, and this is the basis on which you sell that particular project.

Once you have the conversion rate and conversion value in hand, coming up with an estimate of the revenue impact is easy. It’s just the projected growth in NBOST times the conversion rate times the average value per conversion.

Simple, Clear Presentation

The entire discussion so far has been about assembling the raw material for the real task … presenting this and selling it. Here is where your personal skills are key. As before, I am not going to focus on how to make people friends, or buy pizza, or the like, but the end goal. And the goal is to learn what motivates the stakeholders you need to sell.

This might be understanding how they are measured. Or it may be learning with they are passionate about that relates to their job. The better you can do in helping them see how this feeds into their own goals, the better.

Making this happen also requires that you keep it simple. Don’t offer up arcane explanations of things that will require work for people to understand. Part of this is speaking their language, and part of this is keeping it short, and to the point.

You may have 10 sets of details that you could spend time explaining to make sure everyone understands all the intricacies of what is involved, but my advice is that you skip them all and stay at the top level.

There is a direct relationship between the simplicity of your presentation plus your personal credibility, and your chances of success. Have the detailed information ready for when the questions come, but keep the initial presentation of the case for your project straighforward and simple.

In Summary

In an enterprise environment, your success often depends on getting your fair share of development time allocated to your projects. Invest the time required to be able to sell your projects effectively. If you choose your projects well, this will put you in a position to bring SEO success to your company.