Here is a picture:

Both these forums support the functionality of labeling an answer to the question as a “top answer.” Google is clearly picking up on that, structuring the data as such and then displaying that directly in the search results.

This is all likely part of Google’s ongoing, slow and consistent effort of building out their semantic search engine.

Google has expanded this over time with forum sitelinks and listing the number of posts in a specific thread or blog post. This is just one more step in that direction.

Hat tip to Frederik Hyldig from SEOPORT.

Related Stories:

• Google Adds Visited Pages, Past Hour & Fewer Shopping Sites Filtering
• Up Close With Google Search Options
• Google Starts To Classify Content Types In Web Search
• Google Rolls Out Sitelinks Display For Forums
• Google’s Last Visited Time Stamp Gets Down To The Minute
• Finding Search Engine Freshness & Crawl Dates

When you look at the type schema you see close to 300 metadata elements that can describe various entities (called Things here).

There is lot of specificity in a few categories such as Local Businesses, Civic Structures and Events. Two of these categories have real implications for local and mobile SEO because of the richness of the metadata.

Below you see the high-level organizational taxonomy for the MicroData entities. It has up to five levels depending upon which category you are looking at. Here you see that three levels are associated with Local Business, while the Persons and Product groups have no subcategories.

Things

The high-level property elements for Things include Name, URL, Image and Description. These properties are applied to all of the following seven subcategories.

1. Creative Work — Text and metadata include ratings, reviews and offers.
2. Events — Metadata elements includes time, who, place and related events.
3. Intangibles — This is dynamic metadata about quantities and values.
4. Organizations — Metadata to include location, founders, employees and contact info.
5. Persons — Metadata includes title, awards, affiliations, family relationships, personal information and contact information.
6. Places — Metadata about photos, maps, addresses and contact information
7. Products — Metadata about brand, model, manufacturer and reviews.

Let’s take a look at each of these seven categories in a little more detail to see what might be valuable to a B2B marketing executive. Not all will be useful; some will be pure media plays, while others will be B2C oriented.

Creative Works

The properties available in the Creative Works category are metadata elements used to describe all the elements of a piece of creative work. This includes things like identifying headlines, the editor, genre and author. Many more properties are described at the creative works link above.

The properties apply to all the subcategories in the Creative Works group, which include the following types:

Events

Eleven properties are available in the Events category for describing web pages that contain event information. Events can also be broken into super or subevents. For example, if you are putting on several presentations at trades shows, these can be called out individually. Subcategories for events include:

Intangibles

The four properties available in the Intangibles category describe dynamic or intangible data. Examples include the format of a book, the condition of an item, the price, the availability and the seller of the item. Sub-categories include:

Organizations

The 16 properties available in the Organizations category are for classifying data elements that deal with common company information. A large number of the properties focus on contact information, company facts and employees. There are also ratings and review properties.

Person

There are 27 properties available in the Persons category for classifying data elements about a person. You find properties that describe organizational affiliations, alumni status, job title, nationality and the name of the person’s spouse.

There are no subcategories for the Persons group, which is surprising. When you review business websites in any number you see a handful of very defined groups of people. Almost every website has a listing for their Board of Directors, C-level executive, advisory / technical boards and other named teams.

Often, the names of investors are provided as well. You could tag a person with the job title property. If you have a board with seven members you would have to do this seven times. Having a named subgroup that includes more than one individual makes more sense because this reflects current common practice on the Internet.

Place

The 12 properties in the Places category describe unique locations such as a building or a landform. You can specify that a place is part of a larger geographic location. Beyond this the properties are what you would expect: maps, photos, address and telephone numbers.

The Local Business subgroup below is the same one found in the Organizations category.

Products

The seven properties in the Products category are for describing the elements of a product. These properties include specifying a brand name, the manufacturer’s name and product identification numbers, to name a few.

Given the prominent role that products play in Internet commerce, I was surprised at how few properties there are, and I think this will be an issue for successfully deploying the MicroData standard.

Here are some of the elements that I think are missing:

• High-level designation: is it a B2B product or a B2C product.
• Age range for the intended product.
• Industry target: e.g., healthcare, entertainment, financial and technology.
• Product category: e.g., books, clothing, garden, grocery and sports.
• Gender: shoes for example.
• Price or price range.

To be fair, some of these properties are found in the Intangibles category, such as price and seller property. But why these properties are there instead of in Products is a real mystery.

The second problem with this category is there are no subcategories at all for the Products group. It’s almost impossible to find a shopping site that does not have a product taxonomy that is used to navigationally organize company’s wares.

These taxonomies exist for a reason: They work for both the website owner and the consumer. This section really needs to be expanded in the next version of the MicroData specification.

Local Business

I mentioned earlier that a couple of categories had lot of subcategories in its hierarchy. Local Business (part of Organizations) is one example with 27 subgroups that break down into 120 categories.

One of the Local Businesses subcategories called Stores has 30 subcategories in its group alone.

Business names could be used to complement what is missing in the Product category, but it’s far from optimal. For example there is no Internet, Software or Technology categories in the existing taxonomy. The subcategories for Local Business include:

• Animal Shelter
• Automotive Business (ten subcategories)
• Child Care
• Dry Cleaning / Laundry
• Emergency Service (three subcategories)
• Employment Agency
• Entertainment Business (seven subcategories)
• Financial Service (four subcategories)
• Food Establishment (eight subcategories)
• Government Office (one subcategories)
• Health And Beauty Business (five subcategories)
• Home And Construction Business (eight subcategories)
• Internet Cafe
• Library
• Lodging Business (four subcategories)
• Medical Organization (seven subcategories)
• Professional Service (ten subcategories)
• Radio Station
• Real Estate Agent
• Recycling Center
• Self-Storage
• Shopping Center
• Sports Activity Location
• Store (30 subcategories)
• Television Station
• Tourist Information Center
• Travel Agency

Data Type

The Data Type feels like it is unfinished. Sometimes it’s useful to describe content at a high-level as structured (e.g., numbers) or unstructured data (e.g., text). There is value in specifying a structured data element if you have complex web pages with tables, graphs and text.

For example, being able to define a table, having it survive the indexing process intact as a single entity and returning it as a single search result is certainly valuable. Being able to define an image as a chart or an Excel table would also be useful, but these data types are not supported.

• Boolean
• Date
• Number
• Text

Summary

When you look at the subcategories for most of the above groups, it’s easy to see how the MicroData specification is of value, especially the Local Business category. For example, the properties in Local Business provide more markup opportunities for your local SEO strategy.

However, the lack of definition in the Product category makes this a less-than-compelling tool for those firms that sell brick and mortar products through their websites.

Enterprise search engines are well-versed with named entity extraction tools and techniques, which is how this information has been generated in the past decade.

The major search engines are essentially telling content providers that if they name all their entities in their content, they can do a better job of extracting meaning from them to help improve search results. This is a win-win proposition, right? Maybe!

Other similar initiatives known as the Semantic Web and RDFa have not made much inroads in the mindshare of business owners yet. In fact, business owners have a hard time staying on top of current SEO markup practices associated with standard HTML tags.

Will this format gain traction among businesses? I hope so, but there are a number of hurdles that stand in the way at the moment.

Here are some challenges for moving this technology forward. Some are technical, some are not.

• This technology represents a real cost to implement, and the decision to spend those dollars are usually made by the senior marketing executive. In fact, it is very likely that most marketers at any level do not understand the relationship between the various micro-tag technologies and improved search results. There is no real educational strategy in place to promote this specification, and the PR machine is not going to get the job done.

• A second hurdle is the enormous amount of content that is currently deployed on the Web. It would be a herculean task to go back and reprocess this content. Since the value proposition is not well understood by marketing executives, this reprocessing is not going to happen in any meaningful way for some time. In fact, even if the value proposition was well understood, it’s not clear that the cost of processing hundreds of millions of documents is worth the cost.

• The presentation of information at Schema.Org was written by technologist for technologists. Nowhere is there a clear statement of benefits, or a call to action targeted at the real decision maker who happens to be the CMO – not the CTO.

• The hierarchal presentation of entities and related information is hard to review at Schema.org – it’s not apparent that there are only two parent nodes (level-1) with seven child nodes (level-2) at first glance. This information really needs to be repurposed to make it easy to access, and understood by non-technologist decision makers.

• The sale of products and services is at the heart of every B2B/B2C website doing business on Internet. I was surprised at how few product properties there are and that there are no sub-categories at all for the Products category. Maybe this is being worked on, but it is a serious flaw in the MicroData specification. There is a well developed set of business categories, but in most cases these will be too high-level.

For example, a site tagged as an Office Equipment Store implies that it sells office equipment, but this category does not have properties to describe a primary product or multiple products.

Further, the Product category does not have a defined product taxonomy such as those found in the office supplies category at Amazon.

Going forward, I think business owners can be motivated to start marking up new content if they are presented with real examples of why this format is important. But first, the product category needs to be fleshed out to meet the needs of business owners.

Both Yahoo and Google are using similar in-house technology (SearchMonkey and Rich Snippets) to enhance search results in very meaningful ways – this is an excellent use of the technology, but most marketing professionals do not make the connection between these enhanced search results and the new Microdata initiative.

The bottomline is that businesses sell products and they want to sell more products. Their ability to markup their products is limited with the current specification.

I think this is a great idea, the specification needs to be fleshed out, and the sponsors of this initiative need to do a better job of explaining this technology and its benefits to marketing executives. Perhaps these are improvements that are in the works.

Next time, in part two of this article, I will provide a detailed review of the functionality, and how the MicroData specification can be of practical use to marketing executives when thinking about their website strategies.

Microdata is a type of structured mark up that can be used to provide semantic meaning to content on web pages. The microdata types currently supported are documented at schema.og. You can also take a look at the announcements from each search engine on their blogs:

• Google: Introducing schema.org: Search engines come together for a richer web
• Microsoft Bing: Introducing Schema.org: Bing, Google and Yahoo Unite to Build the Web of Objects
• Yahoo: Introducing schema.org: A Collaboration on Structured Data

It appears as though the three search engines will be using this meta data solely to enhance the search results display for now, much like is already done with Google’s rich snippets and was done with Yahoo’s SearchMonkey.

This makes sense for Yahoo, as they control only the user experience of their search results now that the indexing and ranking of their search results now come from Bing. But Google and Microsoft could use the data in many other ways –such as metadata about what queries a page is relevant for and to obtain more accurate and detailed information about business listings for Google Places.

Google is, in fact, using the structured markup from microdata in certain instances, such as with its recently released recipe search. Google uses metadata about recipes (cook time, number of ingredients…) to provide a faceted navigational search.

You can see a complete list of currently supported microdata types and the syntax for them on the Schema.org website.

Once you’ve marked up your pages, you can use Google’s rich snippet testing tool to make sure that the markup is correct and can be read by the engines.

What About Microformats & RDFa?

While Google and Yahoo both have supported their use with their rich snippets and SearchMonkey programs, respectively, neither format is supported as part of schema.org. However, the engines say that the existing support for these formats will continue (even though they imply they’d like you to switch. From the FAQ:

“If you have already done markup and it is already being used by Google, Microsoft, or Yahoo!, the markup format will continue to be supported. Changing to the new markup format could be helpful over time because you will be switching to a standard that is accepted across all three companies, but you don’t have to do it.”

Should you go through the trouble of marking up your pages?

The answer is entirely dependent on your situation. If you’re building out a new site structure and want to have support built in, especially as the engines use microdata in other ways, then it makes sense to include it. However, if you are prioritizing work on your site and have other items to tackle, such as canonical URL issues or a need to invest in creating quality content, those items should probably come first.

As with sitemaps.org, actual implementation may take sometime. The engines will likely want to see how the markup is being used on sites and will test the data internally.

As you can see, all of the ads on the screenshot above — whether above the organic results or in the right column — are showing the small magnifying glass icon for Instant Previews. The preview image itself seems to be work exactly as the feature works in organic search; the search query, “lcd tvs” in this case, is called out and shown in bold on the preview.

When Instant Previews launched in November, Danny Sullivan reported that Google was planning this:

Ads, by the way, will also get an Instant Preview feature in the future, Google told me. There’s no set timeline for this, however.

Google has already pushed Instant Previews to mobile devices and created a video preview format, too.

(screenshot by Zeon Search via Twitter)

At last, Google has finally gained a page preview feature. Called “Instant Previews,” this is a way for you to see what’s on a page that’s listed in Google’s search results without having to leave Google.

The new feature seems promising, another way to save time in the searching process, because there’s less need for people to “pogostick” with their search results.

Pogosticking is when someone clicks from a search listing to a page, then clicks back if unsatisfied to select another result and so on. With the new Instant Previews, they may be better able to immediately spot the most promising pages among those listed.

Up Close With Instant Previews

In the new feature, which was spotted being tested last month, a small magnifying glass appears next to each listing:

Click on that magnifying glass, and a preview of the page that’s listed will appear to the right side of the search results:

Once you’ve enabled Instant Preview mode, you can also use your arrow key to move down through the other results. As you select one, it will be highlighted with a blue background. The preview will change to show the highlighted page:

More Than Just A Thumbnail

Previews that show a small “thumbnail” image of a page are nothing new. They’ve been employed by various search engines over the years. Ask.com, for example, launched its Binoculars service way back in 2004, through the feature appears to have been dropped as part of Ask’s recent makeover into a Q&A search engine. Google even began offering thumbnail previews last year, through its search options feature.

Bing tried a new twist on previews when it rolled out Quick Previews as part of its launch in 2009. By placing your mouse to the right of any listing, you get a longer textual summary of information displayed from the page itself, as with this example below:

Google’s system seeks to marry the two: an image of the page as well as extracting text. Rather than a small thumbnail image, it shows a much larger picture of what the page looks like. The larger image, and one that’s not in a standard square shape, especially helped with people finding the previews useful in testing, Google told me. In addition, the previews highlight the relevant text on the page — similar to Bing — but within the context of the image.

Callouts & Tears

Within the images, relevant text to your search is highlighted through “callouts” that make it easy to read:

Pages can have more than one callout:

Pages might also have have what Google call “tears,” where a page might appear to have been “torn” to show you the overall context of the page but also show you parts relevant to your search. Below, an example of a tear is shown between the two green bars:

Instant Previews & Ads

The images previews cover up Google’s ads, when they appear. Is this a problem for advertisers? Google says largely no.

The company told me that most people scan the search results page quickly. If they’re interested in the unpaid “editorial” results, they’ll tend to stay looking at them rather than going back and forth to the ads. In other words, previews aren’t blocking ads because if someone decides they want to focus on editorial results, they’re already ignoring the ads.

Ads, by the way, will also get an Instant Preview feature in the future, Google told me.  There’s no set timeline for this, however.

Site Owners: You Can Block Previews

Don’t like the idea of Google making previews of your pages? As with Bing’s previews, you can also opt-out of Google’s. To do so, just make use of the nosnippets meta tag. Meta Robots Tag 101: Blocking Spiders, Cached Pages & More explains more about this.

Also, be aware that the descriptions shown for your pages in the results may be different than the callouts on the pages. Sometimes they match; sometimes they don’t. Different algorithms are at work.

Nice Addition

Google suggests that Instant Previews will make it easier to spot if a page contains a table, chart or picture you might be after, and says those using it will be 5% happier with their results. I agree. It should be a useful feature for many searchers in these or other cases. There have been plenty of times when a short preview would have given me a better idea if a site might be worthwhile to visit or not. And, if you don’t like it, well, don’t click on the magnifying glass and you won’t see it.

The preview begin rolling out today worldwide and should be in place by tomorrow for everyone. If you don’t see them yet, visit this special page at Google about Instant Previews and then click on the “Try It Now” link. Or, just click here to go directly to a special version of Google that will enable it.

To learn more about Google Instant Previews, also see Google’s blog post about it. For related coverage of this story across the web, see here on Techmeme.

One of the most obnoxious in recent memory is the blog plug-in tool “snap shots.” Many blogs installed it because it looked “cool” but did very little in the way of adding value.

Bing does a version of this currently, which is relatively helpful but probably not widely used. An arrow to the right of each SERP link enables a small preview window. It would have a greater impact potentially if it showed more of the page and was more prominently called out on the SERP:

Roughly two years ago TigerLogic released a browser-based preview and search refinement tool called Chunk It, directed at consumers. Like most of these tools, few people used it.

Now the company has reinvented Chunkit as “yolink.” The new tool now does new things and abandons its consumer orientation for developers and publishers. From an adoption standpoint, that’s the right choice and gives it a fighting chance. The company just released an API that allows yolink’s enhanced search capability to be integrated into publisher site search and vertical search engines.

Yolink integration exposes more information about a document or page to users and makes that information clickable. You can see the tool in action in different contexts here.

Below is an example of Google Patent search, the current version and then the yolink-enhanced version. Text is extracted from behind the link, keywords are highlighted and users can click on the passage to go directly to the relevant discussion in the document.

While the presentation of the enhanced results and text are not particularly attractive there’s clear value in the service — at least for “power users,” academics and researchers. The question is whether niche search engines and mainstream publishers will see enough value to integrate it for “ordinary” users.

Click here to view the embedded video.

Google has now expanded this support internationally. (While the documentation is available in 16 languages other than English, it’s unclear if international support extends beyond these 16. We will update this post when we get clarification from Google.)

When adding structured markup to your pages for use in rich snippet display, remember that you can use the preview tool to test the result and can request that Google take a look at your pages for consideration (rather than wait for it to be found during regular crawling). In the latest post, Google notes that including structured markup doesn’t guarantee rich snippet display:

“As you’ve probably heard by now (several times), we’re taking a gradual approach to surface rich snippets. This means that marking up your site doesn’t guarantee that we’ll show rich snippets for your pages. We’re doing this to ensure a good experience for our users; but rest assured we’re working hard to expand coverage and include more web pages.”

Proponents of structured markup look at increased support as a signal that Google may eventually use metadata such as that found in microformats, microdata, and RDFa in indexing and ranking as well. That hasn’t happened yet, but as more sites have reason to use structured data (such as to get enhanced listings in search results), Google may find its use widespread enough to be valuable in these other areas.

Events In Rich Snippets

Adding to last year’s announcement of rich snippets for reviews and people, Google has created one for events, too.

Using the hCalendar microformat, Google will show information and links to specific events as part of the snippet below a search result. If you offer event listings, check Google’s documentation for more on how to use this. Note that Google says this will be rolled out gradually, and that using the hCalendar microformat won’t guarantee that rich snippets will show for your site.

Answer Highlighting

Google has also announced plans to add more answers and information into its search results. Google says it’s using the research behind Google Squared to power this effort to extract facts from web pages. When Google identifies a search as factual in nature, and can find the answer on a web page, it may highlight the answer in the snippet.

Google says this feature is rolling out “during the next couple days” on Google.com.

Answer Highlighting may have a negative impact on search traffic to information-based web sites. Rather than referring traffic to sites and pages like these, Google is now explicitly pulling the facts from these pages and showing them in its own search results.

Here’s an enhanced result for Danny Sullivan:

The main link points to Danny’s Facebook profile, and there are additional links to add Danny as a friend, view his friends, send him a message, and … what you’ve always wanted to do … poke him!

Facebook is the seventh SearchMonkey app to be turned on for all Yahoo search users. The others are LinkedIn, Yelp, Yahoo Local, Citysearch, Zagat, and Wikipedia.