lunes, 27 de agosto de 2012

Answers to 7 Questions About the State of SEO in 2012

seo

1) If I'm just getting started in SEO, where do I begin?

If you're just getting started with SEO, you're probably going to want some quick wins. And the easiest way to get those quick wins is to target long-tail keywords -- typically classified as phrases that are three or more words in length. Long-tail keywords typically have less search volume than head terms, but they are also less competitive. Think about it ... it's probably harder to rank for "lawyers" than for "reliable family lawyers in boston," right? Here, take a look at this image to understand the value of long-tail search a little more:

long tail seo strategy resized 600

See that callout in the chart above that says long-tail search yields an average 36% conversion rate? That's the other benefit of targeting long-tail keywords -- because they're more specific in nature, the traffic you're driving by ranking for them is more targeted. And more targeted traffic means higher conversions! Piggy-backing off of our lawyer example, there are millions of people that could be looking for a lawyer; that doesn't help the family lawyers in Boston looking for new clients. So even if they drove traffic to their website for the term "lawyer," chances are, most of that traffic isn't going to turn into a new client ... hence the benefit of targeting long-tail keywords, instead! Less competition. Quick wins. Higher conversions. Good stuff.

2) What's the deal with rel=Author? What does it mean?

When you see rel=Author, it simply denotes the author of a page -- think of it as a way to connect authors with their content on the web. Google is using it to help surface content from great authors in the SERPs. They know that the best content comes from awesome authors, and Google obviously wants to return the highest quality content possible in their search results. They hope this tag will highlight those authors, and help their content rank in search results without having to rely so heavily on inbound links to achieve high rankings.
rel author use for seoWhen this markup language is employed, it also links to your Google+ profile and through that, pulls in your profile image. This should help make that content more engaging in the SERPs, improving click through rates, showing a new focus on not just ranking in SEO, but actual engagement. To start taking advantage of rel=Author, just follow these steps:

1) Have a Google+ account with a full profile. Make sure the email address for the profile matches the author email and name.

2) Use this link to connect your website to your profile: https://plus.google.com/authorship

3) Wait for the content to be indexed -- this may take up to a week or longer.

Follow the guidance from Google Webmaster Tools for additional best practices.

3) What kind of content should marketers create for optimal SEO results?

Search engines are built to mimic human behavior. What's good for your audience is good for them. If people like your content because they find it helpful or enjoyable, they'll read it, bookmark it, and share it. That makes search engines happy.

That means the content you're creating should make readers happy first. And you can do that by making your content specialized and focused. You know, written with your target audience in mind. When you're creating content with passion and the intent to help someone, you'll naturally be creating great SEO content. So be thorough, think of topics that your audience wants to read, and offer it up in a palatable way -- well written, well formatted, and sometimes even in visual formats. If you do this, you won't have to un-naturally force a certain number of keywords into your content just to try to rank in the SERPs. In fact, doing that will make the crawlers and your readers really, really unhappy.

4) Do you have any tips on local SEO?

Why yes, I do. Before I get into some nitty gritty tips, let's get one more general best practice out of the way.

Google has been consistently supplementing search results with immediate answers when conducting local searches. It's their attempt to get you answers more quickly and improve your experience. As a local business, you need to leverage your knowledge and expertise to appear in these supplementary results. To do that, offer much more depth of content, and cover topics that Google may never offer up content on (simply because they don't know). Be as specific and detailed as possible, showing a depth of knowledge that cannot be replaced by a machine.

Now that we have that out of the way, here are some golden nuggets for great local SEO:

  • Add yourself to local directories like your local Chamber of Commerce.
  • Write about local events, history, and the people who work at your company, if possible.
  • Get reviews from sites like Yelp!.
  • Have a Google Places Account -- which is now managed through Google+. Bing and Yahoo offer their own versions, too.
  • Have a Facebook Account so people can perform local check-ins.
  • Have a true "Contact Us" page with a working phone number, physical address, and map.
  • Exercise superior customer service. What people experience in the real world, they talk about online.

5) It seems like larger websites need a slightly different approach to SEO. Do you have any enterprise SEO tips?

Enterprise SEO doesn't need to be radically different, it's just that enterprise businesses typically have more resources. So use them! Here are some tips on how you can leverage your resources for better SEO:

  • Vary your content. In other words, think of content outside of the realm of just text-based content. Make use of videos, surveys, visual content, and industry trends backed up by data.
  • Segment your content. Google really likes segmentation. It drives better search results if you segment content by, say, industry, because it lets them deliver more specific results.
  • Leverage multiple contributors and authors, especially when those contributors may have segment specialization and want to build thought leadership online.
  • Continue improving your overall website speed and uptime. Nothing is worse -- for users and search engines -- than a slow site, dead ends (404s), or worse, a site that's frequently down.

6) What best practices do you have for pay-it-forward link building?

Link building is like relationship building. Come at it with a long-term, mutually beneficial approach. Think about it ... if you were looking to foster a good relationship with a new co-worker, what would you do? Probably slowly get to know them on a more personal level, try to be helpful with projects, do nice things for them periodically, and do things to make them more successful at work. You know, just general, good-person things.

That's what link building is like. Give websites links to their content because it's awesome, and you think they deserve the credit and boost in the SERPs. Connect with them non-aggressively, like on social media. When they post a fantastic blog post, or you share a snippet of their content in your own, tweet at them to let them know you think they rock. As time goes on (and you continute to create excellent content, of course), you'll appear on people's radar naturally. You'll have built a relationship with high quality webmasters, and you'll be one yourself. Birds of a feather flock together!

7) What are considered to be "sketchy" SEO tactics?

A good rule of thumb is if it doesn't help the searcher's experience, it's probably considered sketchy by Google's crawlers. Here are a few more concrete things not to employ, though:

  • Keyword stuffing in your content and URLs
  • Overuse of tags on your blog
  • Buying links
  • Aggressively adding yourself to directories
  • Pumping out extreme amounts of low quality content
  • White on white text -- for the purpose of keyword stuffing or receiving inbound links

Original Article

Facebook’s New Retargeted Ads

Facebook Retargeting Done

Facebook Exchange could pull in a wealth of new advertising dollars by letting businesses retarget ads at Facebook users who’ve visited their websites. Now I’ve learned and confirmed that the first Exchange ads are running, and Facebook has nearly doubled the companies with access to the program. Plus one source close to Exchange tells me “initial results are good, they’re going very, very well.”

Announced 10 weeks ago, Exchange brings the first cookie-based retargeted ads to Facebook — a format common around the web but that Facebook must handle delicately because of privacy. The expansion and comments from Sheryl Sandberg show Facebook has high hopes that Exchange will bolster its bottom line.

Why Exchange Is Critical To Facebook

Here’s a quick primer on FBX. A Facebook user visits a third-party website, such as a travel site where they view prices for a hotel in Hawaii. That website drops a cookie on the user’s browser, and then hands a demand side platform (DSP) the complementary cookie. The DSP identifies the user to Facebook and tells it it wants to advertise to them the next time they visit Facebook.

The user visits Facebook, which pings the DSP asking for a real-time bid for how much they’d pay per impression to show that user a Facebook sidebar ad. If the bid is high enough, the ad gets shown, Facebook charges the DSP, and the DSP charges the website owner.

This is all so important for Facebook because it has classically been seen by advertisers as stronger for institutional brand advertising that improves awareness and perception than direct advertising whose click-throughs translate to sales.

Essentially Facebook is at the top of the funnel at demand generation, opposed to the bottom at demand fulfillment where measurement of return on investment is easier, ad rates can be higher, and Google Search rules. If Facebook wants to boost revenues and also make its ads more relevant, it needs to break into the bottom of the funnel, and retargeted Exchange ads can help.

And they are. Direct advertisers may flock to the DSPs because my source tells me Exchange results are “in-line with performance of retargeted ads [across the web] that have really high ROI. It’s going quite well.” However, there’s currently only a small volume of ads being run and I’ve heard many big travel advertisers are still getting their cookie-syncing set up to work with FBX.

New FBX Partners Signal What’s Next For Facebook Ads

On the social network’s first public earnings call a month ago, Sandberg said regarding Facebook Exchange, “We’re at a very early alpha stage test… but we’re really encouraged by how interested our advertisers and customers are.” And Sandberg’s not blowing steam. We hear AdRoll has assigned a huge chunk of its engineering resources towards working to perfect its integration with Facebook Exchange. Apparently it sees a pot of gold at the end of the rainbow if it can become the premier platform for advertisers to retarget Facebook users.

Some time between the announcement of Exchange and now, Facebook struck deals with six more DSPs to join the program. I’d learned of a few and Facebook agreed to give me the full list of Exchange partners. The newly added ones are in bold:

  • AdRoll
  • AppNexus
  • Criteo
  • DataXu
  • MediaMath
  • Nanigans
  • Optimal
  • RocketFuel
  • Tellapart
  • TheTradeDesk
  • Triggit
  • Turn
  • Xaxis
  • X+1

With the first wave, Facebook curiously sidestepped all its existing partners — the tools and services big businesses use to efficiently buy standard Facebook ad campaigns and Sponsored Stories. There are two possible reasons for this. First, retargeting and real-time bidding required for Exchange are specific technologies that many Ads API partners didn’t have built.

Second, and more interestingly, Facebook is not currently allowing Exchange partners to layer its standard biographical ad targeting parameters on top of retargeting. That means a DSP can’t say it only wants to target users who’ve received a cookie and are 25 to 45 year old males. This might be because Facebook refuses to give personally identifiable data on its users to advertisers. If it allowed the coupling of standard targeting and retargeting by admitting partners who do both, those ad platforms might be able to deduce characteristics of Facebook users and combine them with cookie-data to create stunningly accurate profiles of visitors to their site.

Facebook doesn’t want those DSPs to be able to turn around the Facebook bio data to improve their targeting of users outside of Facebook.com. But a different source says Facebook does want to let advertisers combine retargeting and bio data so they can serve more relevant ads on Facebook.com as long as it’s in a way that doesn’t jeopardize user privacy.

It may be setting itself up for once it’s figured out how to do that, as now Facebook has allowed Nanigans and Optim.al, two existing Ads API partners into the Exchange program. The presence of Nanigans and Optimal in the Exchange program may indicate that when Facebook is ready to combine targeting options, it wants experienced Ads API partners ready to test it. However, my sources say even if Facebook can’t securely merge the targeting options, there’s a very healthy business tapping into existing retargeted ad budgets.

So with FBX ads now running, Sandberg citing strong interest, and more partners signed on, the next step is for Facebook to prove retargeting can earn advertisers money. While the whispers I’m hearing indicate they work well, Facebook will need to compile conclusive studies demonstrating that Exchange ads sell products. Luckily, I recently came in to Facebook to talk to Sean Bruich, head of measurement and monetization analytics. He’s building up a team of specialists whose sole mission is showing advertisers that when they invest in Facebook, they get even more back.


How to see Google's "not provided" keywords?

A client wants to know which traffic from which organic keywords is converting (say, branded keywords like the company name compared to non-branded ones like "purple widgets"). However, the vast majority of organic traffic comes from the "not provided" keywords since the website visitors are logged into Google. 

Organic search keywords are not shown for logged in users to protect their privacy. Now if those logged in users click on an AdWords ad, their privacy no longer matters, and you will see what keyword they entered. So the workaround might be to run a small AdWords campaign to see the converting keywords. It's not quite apples to apples, but it may help inform your organic keyword efforts.

Here are a few ideas: 

- Filter GA to include just organic traffic and then look at the Landing Page report with keyword as secondary dimension. 

If you have done a good job of matching different pages to different keywords you should see that all of the known keywords have comparable metrics to the unknown ones and you can assume that the unknown keywords fall within a comparable keyword bucket. Then do your analysis using landing page/related keyword buckets. 

-If the site is big enough, use Bing data and extrapolate. 

Remember that Bing's index isn't nearly as fresh as Google's so wait at least a month after making changes to do any analysis. 

-Use Google Webmaster Tools to do your analysis. 

GWT shows how often the site appeared and how often the result was clicked. As a result, you should see higher numbers than GA which only tracks when the page loads, but you will still get complete keyword data for head terms.

What do you think?

jueves, 23 de agosto de 2012

Qestions & Answers From E-Commerce SEO

Q: In terms of bounce rate, what do you do if a product is temporarily out of stock?

A: This is a great question because it ties into what to do with out of stock products. My recommendation is to keep the page up and serve the 200 status code with an “out of stock” message instead of removing the page. However, if the product is being discontinued, then you want to redirect that page to the parent category. Given that user-experience is so important (especially when it comes to keeping visitors from clicking back to the SERPs), you should do more than just show an “out of stock” message. Some ideas include:

  • An alert signup form that sends an email when the product comes back in stock.
  • An estimated date when the product will be back in stock, and the ability to back-order it.
  • Show related products (e.g. This is out of stock, but you may also want to check out these other items).

Q: We are planning on migrating from a Zen Cart site to a Magento Cart. Are there any other tips you have before developing a new site?

A: Develop a plan on what you’re going to do about the layered navigation. A work-around I linked to in the slides is Amasty’s “Improved Navigation for Magento” which you can learn more about here.

You can also use the SEO Suite Ultimate Magento Extension or the SEO Suite Pro Magento Extension extensions to handle rel canonical tags and robots meta tags, which aren’t currently default features for Magento. I also recommend looking into this extension as a workaround for tracking multiple steps in a single-page checkout process.

There are a ton of general e-commerce tips included in the webinar, like content on category pages, that would apply to a new site.

Q: Everything Google does is designed to keep people on Google and clicking more of their ads. Do you have any proof of pogosticking and rankings, other than the ability to block a site in google if you bounce back to Google quickly? Why should we get penalized for referring traffic back to them?

A: No, I do not have any proof that someone clicking to your site and then clicking the back button without viewing any other pages is bad for rankings. It would be great if someone has some empirical data on this, but it is an obvious signal that they didn’t like what they found there, so I feel confident in advising my clients to keep that to a minimum. I might rephrase that as: “Everything Google does is designed to get the searcher to their desired goal with as few steps as possible.” Referring someone back to them so they can then click on the next result down from you adds two extra steps, which is why Google may not think that is a desired outcome.

Q: Can you clarify whether or not category names should be in the URL?

A: I think categories should be in a category folder and products should be in a product folder, but you shouldn’t include categories in the product URL. In other words, the set up might look something like this:

http://www.domain.com/category/widgets/ (top level category)
http://www.domain.com/category/widgets/blue-widgets/ (sub category)
http://www.domain.com/product/super-blue-widget/ (product page)

If you are to put a category in the product URL, you often end up with multiple non-canonical URLs for every product, such as:

http://www.domain.com/widgets/super-blue-widget/
http://www.domain.com/widgets/blue-widgets/super-blue-widget/
http://www.domain.com/widgets/super-widgets/super-blue-widget/

In this case, all three of the above URLs would go to the same product page.

Some people prefer a totally flat architecture in which everything is off of the root, but I don’t think that is logical. For one thing, it precludes you from doing some useful diagnostic searches on Google, such as (site:domain.com inurl:product) to find out how many product pages are indexed. Others feel that having the category keywords in the product URL gives them an advantage. Perhaps there is a slight advantage in terms of keywords in the URL, but I feel that is eclipsed by the disadvantages, including those above and the fact that you would be making product pages several more directories deep in the architecture.

Q: How do e-commerce sites optimize for countries? (For example: a Canadian company selling a product to Canadian visitors, when American sites still outrank them.)

If your business is located in Canada, the site is hosted in Canada. you are using a .ca TLD, your customers are located in Canada, and the spelling is Canadian – you should have a big advantage over U.S. sites for searches from within Canada whether on Google.com or Google.ca if all other factors are equal. The problem is that other factors are never equal. If you have all of the above based covered and are still being outranked by U.S. sites, it may come down to their having a better site, better links, more brand recognition and authority in the space, better on-page optimization, and so on.

In terms of having a site that sells to customers in multiple countries, especially if you are using the same content but change certain things like prices, shipping info, and possibly some spellings, you should look into the “rel="alternate" hreflang="x” tag in combination with geographically targeted pages, as discussed here.

Q: Is it wrong to use the format http://www.site.com/category-x/product-x/ rather than just using http://www.site.com/product-x/?

Yes and no. I do not recommend putting categories in product URLs for reasons discussed above, but I also recommend putting products inside a product folder, though I am much less adamant about that than I am about the issue of categories in the product URL.

Q: What is your recommendation of shopping carts? Do you have any preference like Magento, bigCommerce, or Volusion?

I don’t know a lot about bigCommerce specifically, but I would recommend Volusion if you’re OK with being on a Microsoft-IIS server (which many SEOs are not because they are more comfortable working in a Linux, Apache, MySQL, PHP (LAMP) environment). Magento is also a great platform once you implement the fixes discussed in an earlier answer above.

It really is tough to recommend one e-commerce product over another. They all have their pros and cons, some of which are going to be more or less important to different people. To some, a Software as a Service (SAS) arrangement is great because they don’t have developers of their own and need the support. For others, it is just tying their hands when they want to customize their store. For some, cost is the major factor, while others might see features or flexibility as the most important things to consider.

Q: I have reviews on my product pages, but these reviews are also contained in a separate location on the site. What would be the best way to handles these?

Great question! On one hand, you want that unique, useful content on your product pages, especially on a site with manufacturer supplied content. On the other hand, product review pages are a great way to get long-tail traffic from earlier on in the buying cycle. Depending on how much control you have over the review system, you could approach this in several different ways.

Preferably you would look to Amazon.com as an example of how to handle this. Put the first few comments on the product page and, once they reach the max (say, 3 to 10) they would start moving over to the reviews page, which could be in a reviews folder or on a sub-domain. You could decide which reviews show up on the product page by those that are found most useful by other visitors if you have that kind of engagement. Otherwise, you could base it on the most recent, or a variety of star ratings, or some other factor. I wouldn’t just show the highest rated ones because shoppers may be turned off if they see what you’re doing.

I wouldn’t show the reviews that are on the product page on the review page as well. The link on the product page should say “see more reviews” rather than “see all reviews” for that reason.

You may not be able to do the “best” thing for all sorts of reasons. I had to robots.txt block the reviews.domain.com subdomain on a site once because there was no way to keep that content from not duplicating what was on the product pages, and I couldn’t even put a robots follow,noindex meta tag on them. We left a good chunk of reviews on the product page and still allowed visitors to access the “see all reviews” page even though it wasn’t indexed. That situation worked out really well.

Q: Could you go over the point you make about creating absolute links versus dynamic links? How do you not make absolute links?

There is a scalability problem of “hard coding” links to product pages from other pages on the site, such as from other product descriptions. For a few of your “main” products that aren’t getting discontinued any time soon it can be manageable and helpful. But if you make a habit of hard coding links like that and then a year later lots of those products start getting discontinued, you’re now in the business of monitoring 404 reports or spidering your site to find and fix broken links all the time. By doing your product interlinking dynamically, as in “automatically” like with a Related Products area, you can typically avoid the time-consuming hassle of removing links because those products will no longer show as “related” or “popular” if they are taken off the site. If you only have a couple dozen or fewer products in the first place, you can pretty much ignore this advice.

Q: What is your recommendation for surplus paginated pages that once contained products, but for whatever reason there are fewer products now?

My recommendation is to let them 404. So, if you had 66 pages of pagination on Tuesday and then some products were taken off the site and now you only have 65 paginated pages then …/?page=66 would return a 404. The example site I used for this was showing the same content on page 66, 67, 68… as it showed on page 65. Other common issues with this type of thing is a “soft 404” page that doesn’t show anything, or has an error message for the user, but gives a 200 status code in the http header response, which tells search engines that the page is still there.

However, someone contacted me after the webinar and said he handles these with a 302 (temporary redirect) to the last existing page in the set. So in the example above page 66 would 302 redirect to page 65 until more products were added. I think this is a fantastic solution, but my only concern would be if page 66 never came back. That’s ok with one or two pages, but if you’re talking thousands of them across the entire site you may be better off just having them removed from the index.

Q: Do you recommend linking to products from the homepage, or just linking to the category?

Your category links will probably be contained in the site-wide navigation, though I think links contained within the main body section of the page are given more weight. I would advise linking to some of your top products from the homepage, as well as a few category links if they are more important than others. As you probably know, given the nature of your question, the homepage is usually your strongest page and has the most page-rank / authority to pass on to other page. Thus you should link to whichever pages on your site to which you want to pass on the most page-rank. I don’t believe in any “rule” that says you can only put this many or that many links on a page, but I do try to remember that every new link you add reduces the amount of page-rank that can be passed on through the other links.

Q: Regarding "not provided": does that mean you think there is enough randomness in the data that we know to conclude that the same conditions exist in the data that we don't know?

The data you have the ability to see can tell you some things about the data you can’t see. The more data you have, the more you can trust it to tell you something about the data you can’t see because the greater the sample size the more reliable your deductions are likely to be.

For more on this read, I recommend reading AJ Kohn’s post on Blind Five Year Old.

Q: If we have a small amount of products, would it be okay to flatten completely?

Yes. It is “ok” for anyone, but preferable for diagnostic reasons (or if you want to easily segment XML sitemaps, etc…) to have products in a /products/ folder. However, if we’re talking fewer than a dozen or two products, I see no reason why you couldn’t just have them off the root.

Q: Would it be harmful to put the content below the products in the category page as long as it's still useful for the user in order to not push the products below the fold?

I won’t say it would be “harmful” but I don’t see how it could be really “useful” to the user way down there either. I do understand your concern, however, and want to stress that I don’t recommend putting much content on category pages. We’re talking a sentence or two, not a big paragraph. Going back to your question though, I haven’t see any evidence of the “below the fold algorithm” affecting content on e-commerce sites. It seems more to be aimed at advertising that pushes all of the content below the fold.

Q: Should I use a 404 redirect, or just send the 404's to the homepage in Magento?

I don’t like to redirect users to a 404 page. I like to show the 404 response code and message on the URL they were trying to access. Redirecting them to the home page is especially troublesome because 1) it is a poor user experience that could confuse a lot of people, and 2) you lose the ability to capture 404 data that will help you fix broken links and salvaged page-rank from external links that might be going to pages that no longer exist. When you have a page that no longer exists that 301 redirects to a page that exists, which returns a 200 status code that is still a form of a “soft 404.”

Q: I have a buyer's guide page (with great content) ranking much better than a product category page for the keyword of our product category. We want our product category to show first. What can we do to make this happen?

Having some content on your category page should help. Even two short lines of text will go a long way to improve category page rankings. However, it still might not be enough to outrank a content-rich buyer’s guide if they are both targeting the same keywords. Then you might run into the issue that someone reading a buyer’s guide and someone shopping on a category page are at two different points in the buying cycle and probably wouldn’t be typing the same keywords anyway.

Let’s look at it this way: Google wants to return results that put the fewest amount of clicks between the searcher and their goal. A category page is by definition an intermediate step. Google would MUCH rather just send them straight to the product page. However, if the user isn’t performing a query with the clear intent that they want to see a specific product (e.g. best lawn mowers, lawn mower store, buy lawn mowers, zero-turn mowers, heavy-duty weed wackers…) then a buying guide or category page IS the destination most suitable to the query. In this case, “best lawn mowers” would probably be a great query/intent match for a buyer’s guide, but the others would be good matches for a category page.

Q: What number of products are manageable for a view all page? 100? 200? What do you suggest for steps to plan, test, and then implement a view all?

I don’t know the answer to this question, though I suspect that more than 100 is pushing it. I think it might depend a little on what your users prefer (every site will have different users who like to shop in different ways) and how fast the page loads. Test the load time of your View All pages (all categories, as some have more products than others) and make sure they are relatively fast compared to the average site. Google provides that info here. Of course, post-implementation, pay attention to rankings and conversion rates.

As I mentioned in the presentation, I haven’t had much luck with View All Canonical, though I’ve heard other shave. I prefer just using rel next/prev.

Q: Are press releases a good option for building backlinks to my product pages?

It depends on how you distribute them and whether your products are worth press coverage. If you are asking if distributing a press release on PRNewsWire, PRWeb, or some other distribution service saying that you have a new product, then the answer is probably not.

If you want to learn how to do PR the right way check out this post from Chris Winfield.

Q: Can you 301 redirect page 67 to page 66, if page 66 is the last page in the paginated content?

Great question! I would rather 302 that since it would probably be a temporary redirect.

Q: What would you consider to be too many categories? 2? 3?

I don’t have any rule or opinion about how many categories is too many. It all depends on the nature of your business. A site that sells everything like Amazon is going to have a lot of categories. A site that sells only bedroom furniture probably wouldn’t have more than five or six top level categories.

If you are referring to how many levels deep the taxonomy should go, I’d try to keep it at two or three subcategories and rely on filtering or facets to further whittle down the results.

Q: The inclusion of something like /products/ in the URLs allows you to easily isolate product pages in Analytics, which is great for product-specific performance tracking. Is it not worth using a single generic category for product pages?

I think we’re on the same page here, but we might be using different terminology. I wouldn’t call the /products/ directory a “category” but I agree that having it allows you to more easily isolate product pages for many different purposes.

Q: In your Volusion example, I noticed the URL had abc124.htm which seemed to be a product id as opposed to name. Do you like product ID in URL's or does that pollute the URL?

I would prefer just to have the product name in the URL, but if you do put the product ID in there it would be good to separate it from the name by a dash (e.g. productname-1234 instead of productname1234).

Q: How big of a problem is it if you have similar content on your product pages but the content is unique to your site (different variations of your own custom product)?

The more unique you can make that content, the better. If it is just color or size variations, I would use product variant selectors instead of a totally new product page. If they are different enough to require their own page and SKU, then you will just have to be creative in your copywriting. Sometimes it helps to have multiple copywriters work on it since one person could end up being repetitive after awhile if they are writing about similar products.

But to answer your question, I think it is a significant problem if you have a lot of content that seems very similar, even if it is unique to your site.

Q: How do you feel about redirecting (301) expired products?

I feel like that is a good idea. The only problem is when huge enterprise level e-commerce sites have thousands of products expiring every quarter and the volume of redirects gets into the hundreds of thousands, which could potentially cause some site performance issues in terms of load time. In that case, I would rather let the ones without external links just 404 since you wouldn’t be losing any page-rank and wouldn’t be linking to them internally.

Also, I try to avoid redirecting them to other product pages since you’ll end up with multiple redirect hops after awhile when the new product expires. I’d rather send them to the parent category page.

Q: Can the crawlers see contents behind the javascript in the collapsible example provided?

The content isn’t behind javascript in those examples. The content is in the source code on the page. The javascript just controls the page display.

If you put the content in the js file, or or some other file and had the js file or inline javascript call in that content to display then you could have some issues, though Google specifically has gotten pretty good at seeing even that.

Q: Suggestions to increase crawlability for a site with AJAX rendered search results?

I would not rely on search results for crawlability. I would rely on a logical taxonomy and solid architecture using category pages. Google is a search engine and they don’t want to send their users to another set of search results because that is not a good user experience. I know some e-commerce platforms use search pages as categories, but I typically advise their clients to create category/department pages instead.

For a deeper explanation, check out this great post about creating craw-friendly AJAX sites using pushState().

Q: Do you have any tips for generating product page meta data for thousands of products in a way that's more flexible that just plugging in "product name" for the title, description, etc.? I know some platforms use Tags, but that seems a little clunky as well. Have you seen anything work well for this?

It all depends on what information you have available in the database. Get an output of all of the data fields available for products along with maybe 100 – 1,000 examples to start. Then you can just begin mixing and matching like people do with those refrigerator magnets to create something that makes sense and is useful to searchers. I like including prices, manufacturers, model numbers, size, color, etc… But every site is going to be different so I can’t really give a single answer here.

One thing to look out for is the odd description that doesn’t make sense. Your Frankenstein meta description may work for “most” products but if you have a significant number that don’t work well with that automation, perhaps because the fields you chose aren’t applicable to all products, then it could be a problem. Combining things like is/are with singulars and plurals could be an issue too. For example:

Our [PRODUCTNAME] is just [$]! …
Makes sense with: “Our Sony Camcorder is just $19.95!...
But not with: “Our Rewritable DVDs is just $10.99!...

I also like to have lots of variations so they don’t seem repetitive. You wouldn’t use the same format like that on ten-thousand products. When in doubt, the first couple of sentences from the product description is usually a safe bet.

Q: Have you seen any penalties from hiding a large percentage of a page's content with collapsible jQuery functionality? I've been cautious with this and have tried to keep the total amount of content that could be collapsed below a certain threshold percentage.

I haven’t seen any, but I suppose it could happen if you go overboard.

Q: What is the best way to prevent the same product from having multiple categories? Also, would it be beneficial to remove category name from the URL when you click on the product from a specific category? How about choosing one main category for each product?

This is why I suggest not putting categories in product URLs. Only choosing one category per product could work on some sites, but could severely limit you on others and may not be a great user experience or a good way to keep your average order volume high.

Q: With large e-commerce sites, do you find that using xml sitemaps can have a large benefit for SEO? e.g., more pages indexed, more traffic etc

Yes, more pages indexed, especially if there is a deep taxonomy.

Q: Are you saying linking to the same page twice on a page does pass more link juice, or that the second link doesn't get counted?

The second link doesn’t get counted. BUT… I was suggesting a way around this by using a named-anchor link, as discussed here

Q: How do you deal with different models of the same product (different color ipods for example)?

I’d prefer to deal with these using product variants that are selected on a single product page instead of having separate product pages if possible.

Q: For a site that has over 15,000 products, wouldn't it be better to have some type of granular URL (instead of everything being dumped in one tree, i.e. product/)? Would that allow for segmented site maps and provide a more accurate categorization?

Great question. If you can do this without having more than one URL per product, and still having the freedom to add a product to more than one category, then maybe I would consider it.

Q: I am currently dealing with a client with around 30,000 products and he has an algorithm that uses same description but places the appropriate item keyword where needed. What are your thoughts regarding duplicate content and e commerce products? Do you have any recommendations or do I simply need to create 30,000 unique descriptions?

See the notes on slide 17. One thing you could try is having dozens of different types of descriptions applied randomly to the products so at least you don’t have 30,000 that follow the same pattern. Just make sure that once the pattern is applied to a product it stays that way, rather than being randomly assigned one for each time it is called up from the database.

Posted by Everett Sizemore

Agencia 365

PPC Ads in Facebook Search

Watch out, Google AdWords! Today, Facebook has officially launched its own version of PPC search ads, called Sponsored Results. While Facebook hasn't yet made any public mention of the launch, the social network did send announcements to some marketers today officially mentioning the release of the new advertising feature, .

How Facebook's Search Ads Work

According to Facebook's message to marketers:

Sponsored Results gives brands the ability to buy ads in search results, bringing more awareness and traffic to your App, Page, or Place. Sponsored Results give you:

1. Expanded distribution of your message, with placement in one of the most used features on Facebook.

2. The ability to target what people are looking for, by appearing alongside related Apps, Pages, and Places.

3. Flexibility to customize your landing destination, such as a specific tab on your Page.

Brands are not able to drive offsite using this product, but you will have the ability to drive to specific Pages, Tabs, and Canvas Applications.

Okay, so let's talk about this in more detail ...

Advertisers taking advantage of Facebook's new search ads can target searchers who are looking for specific apps, Facebook pages, and places, and they have the ability to include a link in the search results that directs searchers to their own app, page, custom page tab, or page post. These PPC search results are clearly marked as 'Sponsored' at the top of the ad, as you can see from the example below:

 

facebook sponsored results zynga

 

Furthermore, users also have the option of hiding certain ads by clicking the 'X' in the right-hand corner of the ad. Doing so launches a pop-up with a question inquiring why the user is choosing to hide the ad (options include "uninteresting," "misleading," "offensive," "repetitive,” etc.).

 

facebook sponsored results hide the ad

 

Although advertisers don't yet have the ability to direct searchers off-site from Facebook to an external web page, their ads can appear higher than the organic results for a given search. This gives Sponsored Results ads the potential to steal traffic from competitors, as you can see by the sponsored Match.com result appearing before the organic OkCupid results for the search "OKCupid."

 

facebook sponsored search results first resized 600

 

How to Purchase Sponsored Results

While Facebook started rolling out Sponsored Results to bigger app developers like Zynga and Match.com, TechCrunch reports that Facebook is now rolling out the functionality more broadly.

We're told that Facebook advertisers will soon have the option of purchasing Sponsored Results through Facebook's ads API tool, the Power Editor. In addition, the  already exists.

So ... Will Sponsored Results Be an Effective Form of Facebook Advertising?

As with any new offering, we'll have to wait and see. However, we're skeptical about it really challenging Google AdWords. As TechCrunch highlights, Facebook users rarely use Facebook search for anything other than an attempt to find a very specific Facebook presence. For example, users would use Facebook search if they were trying to find the official Facebook page for a particular restaurant they already know about. But if they were searching for different Mexican restaurant options in their geographic region, they probably wouldn't turn to Facebook search -- they'd go to Google!

Original Article.

Agencia 365

martes, 21 de agosto de 2012

Twitter nutre a la creatividad porque es una constante fuente de estímulos.

1.  RECETAS QUE SE DESAFÍAN: las recetas del “buen tuitero” se van desafiando a medida que pasa el tiempo.

Ejemplo: uno de los tips más comunes acerca del comportamiento en las redes sociales es aquel que dice “ser amable y contestar con cortesía”. Especialmente para quienes están a cargo de cuentas de empresas o marcas masivas.

Esta especie de “regla” nos parece acertada. Pero hay quienes la rompen. Y con éxito. 

Las respuestas de la cuenta de Taco Bell, con más de 200 mil seguidores), se ha convertido en una atracción en sí misma con sus respuestas ácidas y políticamente incorrectas. Algunos ejemplos, aquí.

Cuando las reglas se reinventan, las opciones no se agotan.

2. FRONTERAS QUE SE TRANSFORMAN: la creatividad en Twitter no termina en el contenido de los 140 caracteres.

Las “Bio” ya no son un “curriculum abreviado”. Son espacios para comunicar. Algunos las cambian constantemente. Otros las utilizan como “cartel diario” de novedades…

Los “Hashtags” permiten expresar una idea sin necesidad de desarrollarla. Ese es su poder. A veces se sobre-utilizan, es verdad. Pero esto no quiere decir que se hayan agotado. Al contrario, permiten un flujo constante de creatividad.

Los mismos 140 caracteres son relativos. Algunos tuiteros los aprovechan al máximo con tuits cuidados y contundentes (@DelCorrector). Otros tuiteros, rarísimamente tuitean con más de la mitad (@Gauyo).

Al mismo tiempo, Twitter también desafía las fronteras entre especialidades: un meteorólogo en Twitter es algo muy diferente a lo que entendíamos por “meteorólogo” hace algunos años. Lo mismo con un arquitecto (ver Apartment Therapy, por ejemplo). O un urbanista (ver Richard y Rana Florida, por ejemplo).

Las fronteras se van modificando. Progresivamente, sí, pero se modifican.

3. ESTILOS QUE EVOLUCIONAN: más allá de lo que pase con la herramienta, los mismos protagonistas  van evolucionando en sus estilos.

La interacción con Pinterest, la fuerza de Instagram, la estética de Flipboard, la dinámica de Foursquare…

La versatilidad de Twitter para completar y ser completado por distintas aplicaciones, redes y conceptos permite que, quienes tuitean, vayan probando otros recursos y evolucionen en sus estilos.

Maria Popova y sus diversas plataformas (Blog, Lore, Tumblr, Pinterest) es un excelente ejemplo.

4. ESTÍMULOS QUE SE RE-DISEÑAN: la versatilidad de Twitter permite diseñar nuestro propio time line (TL) siguiendo y dejando de seguir a las cuentas que deseemos. Esto se hace más versátil, aún, si recurrimos al uso de listas.

Las personas que tuitean suelen tener ciclos. Tal vez luego de un mes de publicar material de excelente calidad, pasan un tiempo sin enviar casi ningún tuit.

Esto sucede, por ejemplo, con varias cuentas creadas para cubrir eventos anuales o periódicos. En los momentos previos y posteriores a estos eventos, es interesante seguirlas.

Si estamos atentos y no dependemos sólo de la dinámica de las cuentas que seguimos, podemos tener un TL siempre provisto de nuevos estímulos e ideas.

5. PÚBLICO QUE SE RENUEVA: algo que  nos maravilla de Twitter es su posibilidad de ser una suerte de escenario siempre abierto a un público que sabrá aplaudir, abuchear o ignorar nuestras ideas.

Para quien tiene algo que decir, Twitter es un gran aliado. Sabemos que, a toda hora y en cualquier lugar, puede haber alguien interesado en prestarnos su atención.

Por supuesto, la calidad de la atención que nos prestan se cultiva a base de coherencia, calidad de contenidos, imaginación y constancia. Pero, en reglas generales, siempre algún público tendremos.

Y este público se renueva. No sólo porque aparecen nuevas cuentas. También porque la conducta cíclica de muchos usuarios hace que, luego de un tiempo de “no darle bolilla a Twitter”, reaparezcan transformados en usuarios con intereses que han cambiado.

Por eso es que, desde nuestro punto de vista, no hay una forma “avanzada” o “no avanzada”, “mejor” o “peor” de estar en Twitter.

Twitter no es una herramienta unilateral en donde los “expertos” saben más o saben menos. Cada quien sigue su ritmo. Cada quien puede encontrar alguien que valore lo que propone.

Hay que aprender a usar la herramienta, claro está. Pero, al poco tiempo, vale mucho más la osadía para experimentar que la obediencia para buscar en los lugares que todos buscan.

martes, 14 de agosto de 2012

11 Reasons a 23-Year-Old Shouldn't Run Your Social Media

1.  They're not mature enough. Compared with young people 50 years ago, who were eager to enter adulthood and settle down, today's youth are not only not eager to do so, but most do not feel that they've reached adulthood until late into their 20s or early 30s, according to research from Clark University. Instead, they tend to feel unstable and self-focused and would rather explore who they are and how they can transform their lives. This is great for them but not so great for you, their employer--particularly because social media is all about communicating with your audience in mature and accountable ways.

2.  They may be focused on their own social-media activity. Because of the above, if you hire a young person to manage your social media, you may also need to need to worry about how he or she is actually spending his or her time. Will you need to be monitoring the person?

3. They may not have the same etiquette--or experience. Your recent college grad may have experience with Facebook and Instagram, but make sure you check out the substance of his or her updates and posts. You need to make sure your posts reflect your brand--and that you don't wind up with a late-night smartphone photo landing in the wrong account. At the very least, ensure you have a social-media policy in place.

4.  You can't control their friends.  This isn't exclusive to recent grads, of course, but it's a risk to consider: Even if you hire a real winner, be sure that his or her friends won't post inappropriate content to your company's social-media accounts.

5.  No class can replace on-the-job training.  Social media for business is really so many things wrapped into one: marketing, customer service, public relations, crisis management, branding. How deep is the experience of a young person in delivering any of these things?

6.  They may not understand your business.  You are handing the keys to your social-media kingdom to a newcomer, but there's plenty that he or she needs to understand beyond the social tools themselves. What are the nuances of your products or services? What makes you stand out in the marketplace? What are the typical expectations of your customers? How do you troubleshoot issues or cajole customers into working a bit more with you? What does your company stand for? No new hire will be able to absorb these issues overnight, of course--but a brand-new graduate will have an even steeper learning curve.

7.  Communication skills are critical. Communication is critical to solid social-media execution. Before you let a young hire take over your company blog posts, take stock of his or her writing skills. Also: Many young people have not yet learned the "art" of communicating. Make sure they know how to read between the lines, rather than taking things too literally.

8.  Humor is tricky business.  People like to be entertained, on social media as well as elsewhere. Will a young hire understand the boundaries of humor and entertainment appropriate to your target audience, or could your audience wind up being offended?

9.  Social-media savvy is not the same as technical savvy. Good social media requires a combination of both. Successful social-media management involves production requirements, tools, analytics, and other aspects of work.

10. Social-media management can become crisis management. The real-time nature of social media can quickly turn fun engagement and conversation into a public relations disaster, especially if the person behind the wheel isn't thinking a few steps ahead. Are you really willing to take that risk?

11. You need to keep the keys. If you do go ahead and hire a new grad, make sure he or she sets up the social-media accounts using your company's email and shares the passwords with you. Otherwise, you could wind up with no access to these social-media accounts--and no way to take them over.

Social media is not the be-all and end-all. It's a marketing tool--part of an ever-growing arsenal of ways to bring your company to your prospective customers' attention.