What’s Best: Microformats, RDFa, or Micro Data?

In a recent post by Mike Blumenthal about Google’s announcement of supporting Microformats for local search, Andy Kuiper asked in the comments whether it would be best to go with Microdata versus RDFa or Microformat for marking up local business information. As the number of flavors of semantic markup have grown, I think Andy’s not the only one to wonder which markup protocol might be ideal. Here’s my opinion.

Microformats LogoWhen you’re asking “which is better?”, it’s important to know what we’re speaking-of, since there are a number of different goals that people could be pursuing. For some, this is a question of which is better from an elegance-of-coding perspective (if you’re interested in this, you might read Evan Prodromou’s great article, RDFa vs microformats). For yet others, the question should be focused on what’s best for their site — which solution is the simplest, most cost-effective to apply, and least likely to cause problems. Finally, the question could be seen from a perspective of what’s going to work best for the purposes of search marketing?

It’s this last orientation of the question that I’m focusing upon — which semantic protocol is going to work best for Search Engine Optimization (“SEO”)? Continue reading

The Associated Press’s News Microformat

The Associated Press (AP) recently announced a semantic markup standard they’d like to see adopted online for news articles – the “hNews Microformat“. The proposed microformat was announced simultaneously with their declaration of a news registry system to facilitate protection and paid licensing arrangements for quoting and using news article material. While the overall announcement and news registry system was widely ridiculed in the blogosphere (in part because of a confusingly inaccurate description which stated that the microformat would serve as a “wrapper” for news articles, and the overall business model and protection scheme seems both naively optimistic and out-of-touch with copyright “fair use” standards and actual technological constraints), but the hNews microformat part itself could potentially gain some traction.

So, if you’re an online marketer of a site which publishes large amounts of articles and news stories, is the hNews microformat worth adopting to improve your online optimizations?

AP Protect, Point & Pay Diagram
(AP's Diagram Illustrating "Protect, Point & Pay" System & hNews Microformat)

I’ve long been a proponent of incorporating microformats within webpages as a component of overall good usability and potentially valuable formatting for search engine optimization purposes. Microformats can provide some additional, enhanced usability for advanced users who are using devices which can read the information and store it for future use, and they can potentially improve search engines’ ability to understand the content within webpages which could lend a marginal increment more SEO value.

Both Yahoo! and Google have been sending signals for the past few years that they consider some of the microformats to be potentially useful as well. They’ve both marked up their own local search results with hCard microformatting for end users’ benefit, and they’re both starting to make use of microformatting to give certain types of data special treatment. In the case of Google, they announced that they’d begin displaying some microformat data with slightly different listing layouts in the search results, a treatment that they’ve dubbed “Rich Snippets”. And, they say they’ll be rolling out more treatments based on microformats in the future.

With this background in mind, it’s not surprising that the AP has jumped on the microformats bandwagon, but it also appears that they’re trying to influence the development of them where news articles are concerned, with a major agenda in mind. They wish to include some sort of webbug in each news story’s markup, so that publishers of the content can be tracked more easily by them – it will be clearer when sites are reprinting news stories, and how frequently those stories are visited and viewed by consumers online.

Other portions of the hNews microformat appear to be more useful from both a search engine viewpoint and publisher site aspect. Labelling of items including keyword tags, headlines, main content, geographic locations and including author’s vcard info all appear to be valuable standards.

(I could really criticize their “geo” tagging of the articles as quite inadequate, though. Merely adding a longitude and latitude to an article seems quite short-sighted, because there needs to be further definition of what is being geotagged. If an article is about multiple locations, it would be ideal to label each geotag to tell what item is being located. Further, it would be ideal to label the article with an assumption of the geographic region that the article should be expected to appeal to. Is it mainly of interest to people within a particular city, state/province, region, nation, or is it of international interest? Still, having some geotag is better than nothing.)

For any marketers out there considering adopting the hNews Microformat standard, I’d advise waiting until the dust settles on this one. Other microformats developed perhaps more objectively, and there’s a lot of distrust and disaffection with the heavy news industry influence that is involved in this proposed standard. Currently, I’m not convinced that it will be widely enough accepted to become valuable for use. While having AP partners all adopting the standard may be sufficient enough to reach a tipping point where many other sites and companies will make use of hNews, Google’s public response to it was unusually cold-sounding.

Blogger/reporter Matthew Goldstein quotes Google’s response on the matter: “Google welcomes all ideas for how publishers and search engines can better communicate about their content. We have had discussions with the Associated Press, as well as other publishers and organizations, about various formats for news. We look forward to continuing the conversation.” While sounding expectably neutral and noncommittal, Google is also stating that this has not been widely-accepted by everyone, even within the news industry itself. This in combination with widespread skepticism within the developer/microformat community and blogosphere signal that hNews may have a very long way to go before it becomes something worthwhile for optimizing articles on publisher sites.

So, for now I advise avoiding this proposed standard, sit back and see how the dust settles. If you’re already syndicating content via RSS and Atom feeds, then you’re already distributing your content in a manner that’s easily absorbable and readable by search engines.

Should You Use .TEL Top Level Domains (TLDs)?

.TEL domainsPeriodically, someone will launch a new, specialized Top Level Domain (“TLD”), claiming it’s the next big thing on the net. As we’ve seen time and again (such as with the .MOBI TLD), most of these efforts are never going to achieve the same level of recognition or adoption as the .COM and .NET standards, and businesses which muck about with them are likely to expend valuable resources resulting in zero ROI.

Such is likely to be the case with the .TEL top level domain which launched in March. .TEL, operated by Telnic Limited, is intended to be a sort of domain-based authoritative location for contact information – a sort of grand new evolution of phone directories, white pages, and yellow pages. When you obtain a .TEL domain, you don’t manage it on the servers of your choice, but instead it will generate a site hosted on the Telnic service. Justin Hayward, Communications Director for Telnic, is quoted as saying:

“We consider .tel to be the first global live contact site directory. Once contact details are populated in a .tel, anyone can type a known .tel address into any browser or use keywords that describe the person or business they want to find. Keywords are free so the more keywords that are used and the more descriptive they are, the easier it is to be discoverable.”

On the surface, this all sounds good, but the first problem I see with it is one of adoption: if people know to look for you, they will not be likely to type in a .TEL domain name — everyone looks for .COM. Even if you specifically tell someone about your .TEL URL, you’ll expend extra time explaining that, yes, .TEL *is* a type of web URL, and even then they’re just as likely to type it in as “something.tel.com”, which is operated by Tokyo Electron company, and not the proper Telnic page URL.

This is the exact same issue with .MOBI, which was intended to be used as an authoritative URL for the mobile-friendly versions of websites. Most people don’t know/understand the protocol, so they won’t be naturally typing it in when out and about with their mobile devices (and, .Mobi has the additional downside of creating one-letter-longer URLs, which make it that much more tiresome for someone on a wireless device to type in).

From a marketing perspective, .TEL domains have additional downsides. You don’t appear to be able to control the UI or look-and-feel of the generated contact pages, and slapping on yet another domain can split the effectiveness of your natural search engine optimization work. Links pointing at that additional URL will dribble away portions of the PageRank you could be sending or keeping for your primary domain.

And, how will search engines treat it? As with many of the lesser top-level-domains, they’re likely to be more mistrustful. I see zero toolbar PageRank values for the top-ranking .TEL pages, though this may be due to the domains only getting launched recently. But, their public statements touting keywords (“…Keywords are free so the more keywords that are used…the easier it is to be discoverable…”) and the fact that domainers seem to be excited by having another channel to potentially exploit makes the TLD concerning in terms of potential search engine performance.

Bucking standards in favor of creating your own proprietary one, and flying in the face of established adoption rates in terms of internet consumer behavior are not a good formula for success.

In a very self-serving blog post by Telnic CTO, Henri Asseily, titled, “Why .tel and not a free hcard microformat?“, they seem to also be taking aim against the increasingly popular hCard Microformat standard in favor of .TEL. What’s funny about this is that this is comparing apples and oranges, and Telnic has deployed example domains (see emma.tel, henri.tel) which provide a vCard at the bottom of the pages.

It’s absolutely stunning to me that they took the trouble to provide vCard off of their contact info pages when they could easily also embed all the vCard information into the page itself, using semantic markup! Meanwhile, his blog is suggesting that using .TEL in some way should be done instead of using hCard! And, I totally fail to see the significance of protecting info he’s referring to with privacy settings — are the .TEL pages a publicly-findable directory of contact info, or not?!? It’s disappointing that they wouldn’t simply incorporate the hCard and thereby gain additional advantage from the special display treatments that Google has begun applying to microformat-enriched pages.

Telnic is partly promoting their service as a way of providing individuals’ and businesses’ contact info on the internet, “even if you don’t have a website”. Ummm… don’t the online white pages and yellow pages already do this?

For companies considering adopting the .TEL for online marketing advantage, you should seriously reconsider. This is not going to become the defacto online standard anytime soon, and expending time playing with this domain is going to take resources away from efforts which are likely to be far more beneficial. At worst, linking to new .TEL domains could also subtract some of your existing PageRank value to little advantage.

The only case in which a .TEL domain could potentially provide advantage is in the case of a project to improve online reputation, if you’re looking for additional webpages to come up in SERPs, helping you to push down some sort of negative content which may be ranking for your brandname. However, there are a lot more social media sites, business profile pages, and additional strategies which you should be employing in that case, and the unproven nature of .TEL sites in organic search rankings relegate use of the new TLD to the bottom of your list of possible online reputation weapons.

Optimize Your Search Engine Listing for Improved CTR

Earlier this month when I spoke at SMX Advanced on the topic of “Beyond the Usual Link Building”, one of the suggestions I made in the presentation was about how to improve how your listings appear within the search engine results.

There are a lot of people I’ve met who tend to be hyperfocused on whether their pages rank, and don’t spend as much attention on how those pages’ entries appear within the search results pages.

It seems like common sense that if the entry looks like what a user is seeking, they’d be more likely to click upon it. Therefor, if you were to improve your search engine results page entries, you’d also likely improve your click-through rate increasing your traffic.

Compare these listings on Google for a search for “Seattle indie records shop“:

Seattle Indie Music Shops Listings in SERP

You can see that the star ratings and review on the listing for “Easy Street Records” is slightly more eye-catching if you were a records shop afficianado the stars and the dollar-sign price range and the easy-to-read sample review text give it an advantage over the listing for the record shop below it. A consumer who is rapidly scanning and clicking to find what they want is going to be more likely to click here.

How much more likely is such a listing to gain clicks? According to Vanessa Fox, Yahoo! has reported a 15% click-through-rate (CTR) increase on similar types of listing treatments! Their results were based upon comparing the CTR of typical search result listings with CTR of listings sporting their special treatments developed through SearchMonkey. The customized listings really stand out from the other listings, drawing the eye and clicks, too.

Yet, before these research results were released, I’d already seen how merely fine-tuning the listing text alone could improve both CTR and rankings. Using savvy methods for forming TITLEs and Meta Descriptions on pages, one can improve keyword relevance, ranking, and click-through-rates.

Now that Google has launched their own type of enhanced listing treatment, dubbed “Rich Snippets“, there’s starting to be even more options for optimizing listings in search results. The first special treatment they’ve enabled are the ones for reviews and ratings, and it seems clear that they intend to launch more, particularly ones related to the use of Microformats, such as hCalendar, hCard, and hProduct.

One person at SMX who liked this concept of “optimizing listings” for improved CTR was Matt Cutts, who Tweeted out a mention of it:

Matt CuttsTweet re Rich Snippets

While these tactics likely have no direct effect on search engine keyword rankings, I’ve theorized for some time now that they could have an indirect effect upon rank. Google’s frequently-discussed patent for “Information Retrieval Based On Historical Data” includes within its descriptions of ranking methods (“scoring”) the possibility that pages might be ranked according to how often they’re clicked upon when they appear within particular searches. The patent states:

“…scoring the document includes assigning a higher score to the document when the document is selected more often than other documents in the set of search results over a time period…”

Very loosely interpreted, this means that if your page’s listing is clicked upon at a better rate than other pages appearing for the same keyword search, that click-frequency or CTR could actually affect that page’s future rankings for that keyword.

It’s long been controversial as to whether Google implemented many of the methods outlined in various patents like this one, but you already have a good excuse to fine-tune your listings: regardless of theoretical impact on rankings, it could easily improve your click-through rate, improving your site’s qualified traffic!

Quick Tips on Optimizing Listings:

  • Title should be brief and state what the page is about, and who you are.
  • Meta description should be brief and expand upon what the page is about or how it may be better than others listed for the same keyword search.
  • Currently, mentioning deals/discounts/rebates may improve CTR since the economy has pushed people to be more price-conscious.
  • Implementing Microformats now on your site for appropriate types of content will likely position you to take advantage of future rollouts of “Rich Snippets” treatements in Google results.
  • Building a search application with Yahoo!’s SearchMonkey platform will help you to understand how Google’s developing similar types of listing enhancements.

Good listing engineering is a complex task involving semantic tagging, taxonomic research and development, good copywriting, and SEO knowledge. Don’t make guesses when doing this use a good expert if you don’t have experience with it.

Optimize your snippets and SERP listings, and improve your CTR and Performance!

Why Use Microformats?

Microformats LogoI’ve written numerous times about how and why to code Microformats into the webpages of local businesses (see here, here and here), yet the question keeps coming up “Should I spend the time and effort on integrating Microformats into my site’s pages?”

Just during the past couple of weeks, the question has arisen yet again, and along with it there was an additional development which further emphasizes why it’s a good thing and why webmasters should be incorporating the protocol sooner than later. More on this in a minute.

I believe I was likely the first to ever propose using hCard Microformats as a component of local search engine optimization, back in 2006 (see: Tips for Local Search Engine Optimization). Back then, I had seen how Microformats were begining to take off, and I saw indications of converging trends: the sharp interest from the major search engines in local search and yellow-pages-like functionality; the increasing uses for types of open formats and extensible semantic tagging; and, most telling of all, the involvement of a number of key technologists from within Yahoo! in the Microformats movement.

I knew that as search engines attempt to match up websites which they crawl with more formal, local business listing data, they would encounter some difficulties in using algorithms to interpret the data properly. Questions such as: What is the street address of this business webpage? What is the Business Name vs. other text on the page? What is the Street Name vs. the City Name? Other questions arise as well, since website designers mostly design towards their human audience rather than algorithms attempting to interpret meaning from raw data. For instance, what Business Category should this local business website be associated with?

Like other forms of semantic markup, Microformatting labels webpage content behind the scenes, specifically telling what each piece of data is while still displaying the webpage normally for human users. If webpages of local businesses were to incorporate hCard Microformatting, I reasoned, then search engines would have an easier time of associating the sites with map locations and business directory listings. Further, if a site contained such markup, the search engine could have a higher degree of confidence in accurately normalizing their data and matching up with user queries, so such pages could potentially rank better in the future.

However, when I introduced the idea, I was not aware of any search engine that was specifically seeking out this type of semantic data. While some Yahoo! personnel were throwing support behind the movement, there was no clear indication that their search engine would seek out specially labeled data fields nor treat them any differently.

Still, there were additional reasons for using the Microformats: they provided additional functionality for some devices and for users who installed special applications to read such content out of pages in order to easily make use of it. A great example would be the Operator Toolbar for Firefox which could allow a user to easily copy out the contact details from a webpage and save it into an address book, quite similar to how vCard electronic business card info can be transfered and harvested easily from email notes (vCard is supported by such mainstream applications as Microsoft Outlook).

The Yahoo! Local search team obviously believed that people could find Microformatting potentially useful, because they incorporated it into their Local Search results earlier in 2006.

Further supporting my prediction that this was an important and growing protocol, Google subsequently immitated Yahoo by incorporating hCard Microformats in Google Map search results in 2007.

Meanwhile, at conferences and via email, many individuals asked me whether Google Maps was “reading” Microformats from webpages. I’d spoken with a few Google engineers during this period, and they answered pretty uniformly: if sufficient numbers of sites made use of this, they’d almost certainly incorporate it as yet another signal in local search data. I knew that there really wasn’t “sufficient numbers” of sites incorporating it yet, but I continued to see indications that the protocol was growing as a trend, and a number of other optimization experts also threw weight behind supporting it as a component of good, local site design. So, I’d still have to truthfully answer, “no, it’s not any sort of factor that will directly make your pages rank any higher, BUT, you should make use of it anyway!” In most of the cases of local info pages I analyze on the web, it seemed like integrating the Microformats should be relatively low-impact in terms of development effort required.

SearchMonkey LogoIn the Spring of last year, Microformats may have finally achieved a tipping point when Yahoo! announced the release of their innovative Search Monkey development platform. SearchMonkey allowed developers to somewhat customize the display of their site’s listings when they appeared in Yahoo’s keyword search results. More to the point, SearchMonkey showed us that Yahoo’s bot and content processing systems could and did read in Microformats from webpages along with other structured data protocols including RDF and DataRSS. While this did not prove that Microformatting influences rankings in Yahoo! Local, it showed that an important step had been reached in a major search engine which could enable the protocol to be a ranking and normalizing factor in local search.

Now fast-forward to the present in 2009, and the question of whether to use Microformats is still getting posed to search marketing experts. On May 4th, someone asked well-known SEO, Michael Gray whether hCard and other Microformats matter for SEO. I think Michael gave a pretty well-reasoned answer overall, although I believe Microformat protocols are just about excruciatingly simpler than he represents, and I think there’s some good reasons to not be quite as conservative about using them as he suggests.

First of all, I believe the main advantages to using Microformats are:

  • They can help search engines identify Business Name, Address, Phone, and Categories on webpages. Variations in formatting on various sites can contribute to misassociation of data elements. Imagine “Houston’s Restaurant on Dallas Street in Paris, Texas”. If an algorithm is attempting to interpret this in order to index the business/site, how does it know for certain what element is name vs. street address vs. city?
  • They can help search engines in associating the website with their listings within the engine’s directory listings content a vital step in “canonicalizing” business information. Google gets business listings from data aggregators and business directory partners, and they have to associate all the various sources of data for a particular business location with a single business listing. This is not a simple activity! Differences in ways a business name is spelled, different ways addresses are written, and different phone numbers all can result in businesses’ listings getting duplicated and diluted in ranking ability within Google Maps. So, having Microformats on your business webpage could help it get properly associated with directory listings already within Google.
  • Microformats facilitate the ease by which users can copy off a business’s contact information to store in their address books and elsewhere.
  • Microformats could also help open up content for use by other developers in unforseen and advantageous ways. For instance, by including the longitude and latitude of your business address in your pages, others can easily port the precise location over to the mapping app of their choice if left up to just using the street address, mapping systems can frequently make significant errors.
  • It’s just not all that hard to add them to sites which display addresses of local places. Some very simple development and coding which could be done within just an hour or two are all that’s required for most sites.

Google actually does a pretty good job of “canonicalizing” classic business listing data from local biz websites, so if my theories on why it could be beneficial for SEO in the future are correct, there are a lot of sites where it likely wouldn’t have all that much impact upon performance even if/after Google begins recognizing it as a local site search signal. It could help them collapse dupe listings down to a single one, which could boost that listing’s ranking weight. But, for businesses with already easy-to-interpret addresses or where Google hasn’t had difficulty in grouping related listings together, it likely wouldn’t have any ranking effect whatsoever.

As of just last week, there’s an even more compelling reason to incorporate Microformats, though: Google is following close upon the heals of Yahoo again and has announced that they’re introducing “Rich Snippets” in Google search results pages essentially the Rich Snippets are more enhanced search result listings, allowing the display of star ratings and the numbers of reviews for content on the pages. Similar to Yahoo’s SearchMonkey which allowed some customization of search listings, Google is allowing this special content display initially for pages which incorporate hReviews Microformat.

Google SERP listing for Yelp with Rich Snippets

Many of us theorized that Yahoo’s SearchMonkey could be potentially advantageous to sites, since search result listings which look different can stand out from the crowd, attract more users’ notice, and therefor have a greater chance of being clicked upon. Indeed, subsequent research showed that SearchMonkey’s special listing treatment could increase CTR by 15%!

There’s every reason to believe that display enhancements likely could improve CTR within Google search results as well, so there are great incentives to adopt the hReview protocol for those sites which have reviews and ratings content. This is only the first stage of Google’s work in Rich Snippets, however, and it’s pretty certain that Google will introduce more types of structured data into special display within search result listings. hCard and hCalendar content are some top candidates poised for imminent introduction when Google expands this.

We’re now seeing adoption of hCard in even some high-popularity sites such as Twitter now, so it may be time to actually declare Microformats to be “mainstream”!

So you see, there are compelling reasons to use Microformatting in the here-and-now, rather than putting it off. It’s generally not difficult to implement, it enhances site functionality for good user-experience, it generally won’t interfere with existing graphic layout, it could eventually help in rankings, and it might soon help in terms of click-through rates or overall conversions.

Technorati Profile

Twitter Uses Microformats

While using Twitter this week, I realized their programmers had incorporated Microformats in the design! I noticed that my Operator Toolbar was responding to the Microformat content in the page, and making it available for me to export.

As you can see from my Twitter profile page, Operator has found Contacts and an Address available in the page. Note the “Contacts” and “Addresses” buttons in the browser toolbar are not grayed out, but are showing as clickable.

The Contacts is returning hCard Microformat info not just for me, but also for all of the 36 twitterers that I follow and whose icons appear on my profile page.

The Address is apparently supposed to be my personal profile’s address data, but it’s not interpreting quite right for me. I think this is because it places the entire Twitterer’s location content in the “adr” value, without breaking the content out into the street address, locality, region and country. Also, the hCard profile attribute isn’t included in the page’s tag.

Still, Twitter’s incorporation of the Microformats in the page code is exciting to me! Why? Well, I’ve written before about how incorporating Microformats can potentially be advantageous for the purposes of Local Search Optimization here and here. Essentially, this can help search engines to more easily interpret the address info on webpages and associate business information with webpages.

Yahoo! has been the fastest at adoption of Microformat content, with Google following close behind. Yahoo’s Search Monkey platform (which allows both Yahoo engineers and all other web developers to create applications which deliver up special webpage listing representations in Yahoo search results) has shown very clearly that Yahoo’s bot has been tooled to particularly harvest Microformat data from webpages in order to make special use of that amongst the various signals they get from sites.

Does Google use Microformats? Yes and no. Google Maps has incorporated Microformats in the display of their search results so that users can access, export and use business and address data easily. However, it’s not yet entirely clear if they spider that same data from local web pages as part of the info they collect in categorizing and ranking pages. Google Maps engineers have told me off the record that they watch all types of data like this, and if there’s a significant number of sites using it, then they will also make use of it in their ranking “secret sauce”. With a high-profile site like Twitter incorporating Microformats, there’s yet more incentive for Google to adjust their data collection algos to incorporate hCard data if they have not already.

In the past week, I wrote an article on how small businesses can and are using Twitter for local marketing. Twitter’s incorporation of Microformats further underscores the value of the service as a component of Local SEO.