Why actually use micro-marking. Overview from Yandex

    We already wrote about how the world of semantic markup works and why there are so many syntax standards. Now we want to tell you why all this is necessary. Why are so many standards developed and where are they used. Why is it useful for everyone to know?

    There are many large tasks in web services based on data transfer and on creating a connection between objects. In this article, we will analyze examples of specific projects implemented using semantic markup. Search engines strive to improve search results: sorcerers develop, launch affiliate programs, collect “knowledge bases”. Most of these tasks are solved using markup. Sites can get beautiful snippets, transfer data to Yandex.Ostrov



    , and the “knowledge bases” that almost all search engines have (Google Knowledge Graph, Bing Satori, Yandex Entity Search) get the latest data.

    Facebook was interested in developing the application industry, so it gave developers the opportunity to promote their projects on social networks and connect with other applications. So - also with the help of markup - Open Graph publishes application entries on social networks, and Applinks allows applications to easily link to each other. Later, publishing your posts became possible on Google+.

    Read on and find out more about how the tasks listed below are solved with the help of semantic markup in different products and services:
    • Search Engine Improvement: Snippets in Yandex, Google, Bing and Yahoo!
    • Knowledge Base Replenishment: Yandex Entity Search, Google Knowledge Graph, Bing Satori.
    • Other use cases: applications, deep links, and mail.

    Search Engine Improvement: Snippets in Yandex, Google, Bing and Yahoo!


    In order to better respond to user requests and do an extended search, search engines need to solve the problem of obtaining structured data from sites. There are two solutions:
    • Getting data using algorithms.
    • Receive data from webmasters.

    Both methods have pros and cons:
    • Developing algorithms is interesting, but quite expensive.
    • It’s easier to request data from webmasters, but they can provide both real data and false information. In addition, the completeness of the provision of such data is unlikely to be one hundred percent.

    Search engines have experience in developing algorithms - they have been doing this from the very beginning. And to request data from webmasters and develop an affiliate program for them is interesting, but risky - will they add markup to improve the delivery?

    In 2008, when the markup had already begun to spread and a number of sites were ready to transmit structured data, the Yahoo! platform was launched SearchMonkey , and a year later, the first Yandex affiliate program started. It became clear that many webmasters are interested in honestly promoting their site, making it look better on the issuing page.

    Yahoo! Searchmonkey

    Since May 2008, when Yahoo! SearchMonkey started, developers could transmit semantic data and improve the search results of their site. This was the first affiliate program in the world for sites with semantic markup.


    Different data could be transmitted using markup in the syntax of RDFa or Microformats . As a result, snippets saturated with semantic data appeared instead of the usual ones. Many of them had not only an improved appearance, but also various buttons and the ability to interact with the site without leaving the search page.

    For example, by adding markup to user profiles, Facebook shared its datafor Searchmonkey. The profile information was displayed in the search results, and it was possible to add the user to friends or send a message. It was also possible to add videos, documents, flash games to snippets .

    SearchMonkey was recognized as one of the 10 best products of the Semantic Web in 2008, and in October 2010, as a result of a transaction with Microsoft, it was closed along with some other services Yahoo !. On the company's blog, you can read the posts about the development of SearchMonkey.

    Examples of advanced snippets of Yandex, Google and Bing


    Despite the fact that SearchMonkey has ceased to exist, the idea implemented in 2008 continues to evolve. Other search engines regularly launch affiliate programs for different types of content.

    In Yandex, such programs work as follows. User queries are analyzed and suitable answers are thought up for them. Then, different data sources are considered for these options, and in some cases semantic markup is the most suitable. Then an affiliate program for webmasters is developed and launched, which we will announce. Website owners add markup to pages with relevant content and, thanks to it, get structured snippets that can attract more users. Snippets can contain pictures, additional information, or simply structured text. At the same time, if ordinary snippets respond better to a specific user request, they continue to be formed for pages with markup.

    In Yandex, Bing and Google, beautiful and structured snippets are built on the basis of Microformats, Open Graph, Schema.org and their extensions.
    Each search engine has its own range of affiliate programs. Almost everywhere there are programs for reviews, recipes, addresses and organizations.

    Chords in Yandex:


    Music in Google:


    Reviews in Bing:


    Detailed information is in the documentation: Yandex , Google , Bing .

    Yandex Islands

    We decided to go further than improving snippets and enable users to interact with sites immediately on search results, where the latter can be presented as a useful interactive response with real-time information updated.

    July 3, we announced the launch of a new design.


    They also talked about several new types of islands, formed on the basis of Schema.org. Islands with an article, object card and catalog are available at beta.yandex.ru.


    Knowledge bases


    Search engines have always searched by analyzing the texts of pages on the Internet. But not in all cases, users need sites. In terms of queries, people are looking for information about the essence of the real world. And the best answer to such requests is not a link to a page, but data about a specific object. Recently, search engines have begun to solve such a problem and provide, in addition to the list of sites, additional information upon request.

    In Yandex, this is how Entity Search works, in Google - Knowledge Graph, in the search engine from Microsoft - Bing Satori. Everywhere the work is based on a similar principle:
    • the response to the request is replenished with an object card and recommendations;
    • data is taken from open sources such as Freebase and Wikipedia, as well as from accumulated knowledge bases.

    For accurate information - dates, geographic data, etc. this method is the most suitable, but in some cases you need information relevant to a particular point in time. For example, about the upcoming concerts of performers. To solve such problems, markup is used.

    At the 2013 I / O conference , examples were shown where the information marked out by Schema.org falls into the Knowledge Graph area.


    Bign Satori, unfortunately, does not work in the Russian-language search. But there are examples from the English language with interesting features: links to lectures in TED or the ability to listen to audio recordings in the column about famous personalities can be added to the object’s card .


    Yandex Entity Search already works in Turkey, and in Russia it can be seen by requests related to football: The


    technological part of the project has already been worked out, and in the future we will present our visual solution for a wide range of requests.

    Other use cases: apps and deep links


    The creators of applications for social networks seek to ensure that users spend more time in them, tell friends about it and in every possible way PR. Naturally, this is not very interesting for users: if they want to talk about what they did in these applications and what they achieved. Facebook came to this conclusion and developed its own Open Graph layout standard, with which applications can actively participate in the user's social activities. A little later, similar things began to develop on Google+.

    On Facebook and Google+, using markup, applications can publish user actions on their social networking pages. Facebook calls it “user stories”, in which you can indicate the actor (user), the object, the action with this object and the name of your application with which all this action was performed. The standards Open Graph and Schema.org are used for this.

    Open Graph can also use applications with games .



    For more information and usage examples, see the Facebook and Google+ documentation . Also, using Open Graph, you can control the preview of links to your site - we wrote about this in the first article .

    Deep links


    The problem in mobile applications is integration with other services and applications. For example, when you want to immediately download the application from the search results or buy a ticket by looking at the nearest cinema on the map. Various projects related to deep linking or “deep links” solve this problem for users and developers.

    In principle, the expression " deep linking " means a transition to any non-main page of the site. In the context of applications, we call links that allow you to not just get to the page, but immediately open the download page or the right part of the application.

    Now this topic is very actively moving forward and is especially supported by companies developing the application field. Already you can go to a specific part of the application or perform some action (for example, listen to music) on the search results page.

    This is in Google and Bing .


    Using Schema.org markup on your site, in Bing, for example, you can link individual pages to specific Windows applications . This helps to add links directly to applications in the search results. You can also add markup so that following links from the site the user immediately goes to the necessary sections in applications on Windows Phone.

    There is a way for cross-platform application communication solutions using several tags from Facebook - Applinks . Using Applinks, you can link applications to each other, and more recently, link to specific parts of the application .

    Markup in the mail


    As you can see from the story above, most of the markup projects are related to search engines and applications. There are implementation examples in other services, for example, in the mail.

    Gmail Actions allows you to add buttons for user actions based on Schema.org. Action markup is used - a standard in the development of which Yandex also took part .

    For example, you want to invite a person to an event. To give him the opportunity to save time and immediately respond to the invitation, you can use the markup to add a special button with which the user can give an answer without even opening the letter. Similarly, the user can leave a review :


    In addition to the buttons for writing reviews and responding to invitations, you can add buttons for other types of actions - save the coupon, add a song to the playlist, register for the flight and others.

    The same markup is used by Google Now if these are letters confirming various transactions. By analyzing content and tying it to time, Google Now reminds users of flights, table reservations in restaurants, and other operations.

    Microsoft has a similar opportunity: airlines can transfer data for the Cortana service (similar to Siri). To do this, you need to add the Schema.org markup to outgoing emails with flight information.

    By the way, Yandex.Mail can alsoAllocate letters from airlines and prompt about meetings that were assigned to the addressee in the letter, but does this using linguistic technologies .

    Top-up vertical service databases


    With the help of micro-markup, not only the “knowledge base” of search engines is replenished, but also other services. For example, Yandex has an affiliate program for video . You can transfer detailed information about materials to the Yandex.Video service using the Schema.org markup or Open Graph. Thus, by analyzing the data, Yandex.Video not only replenishes its database of video materials, but can also more accurately respond to requests for various parameters - by actors and heroes, genre, release year, or original language.


    In this article, we tried to highlight the most diverse and interesting products that find solutions and help services, users and developers using semantic markup. By adding several tags to the HTML code of the page, you can not only tell search engines what is on it, but also get bonuses - in the form of beautiful snippets or a link in the additional response field from the knowledge base. And using the markup in the application, increase popularity through social networks or connect with another application.

    If you have more examples of micro-markup-based projects, share them in the comments! And in the next article we will analyze an example of implementing semantic markup on a site.

    Also popular now: