• sharithurow

    Hi Matt-

    I support schema.org. The problems? (1) It’s far too technocentric. (2) There are no professional information architects, usability professionals, GUI professionals, etc. in the schema.org group to keep the focus on users.

    GUI = graphical user interface
    HCI = human/computer interface(s)

    At least information architects conduct usability tests on real users and put data into proper context. Has anyone reached out to professional information architects? Many of them have degrees in information sciences.

    I took the “library” out of that degree part due to the overwhelming stereotype of “librarian.” There are many information scientists who understand users.

    Maybe you’ll get more widespread use of schema.org if it weren’t so technocentric and the group reached out to people who can help put that information into a better frame of reference.

    I find the technocentric focus (and arrogance) quite troubling. Guess what? So do users. I observe it all the time.

    My 2 cents.

  • Durant Imboden

    It’s probably just a matter of time until schema markup is incorporated into user-friendly tools. Even in DIY form, it’s more drudgery than technical challenge. Whether it’s worth doing probably depends on your objectives. (Making it easier for a search engine to publish your business’s phone number, address, hours of operation, etc. on a local SERP page is likely to be more valuable than coaching a search engine on the differences between headlines, images, and body text.)

  • Ben Heligman

    Then what is the advantage of markup?

  • http://www.problogger.gr/ Pano Kondoyiannis

    better visibility, better brand ;-)

  • gabsgabsgabs

    I really don’t feel it’s that complicated unless you want to try and use every single markup a page can have…

  • Jason Sykes

    Completely off topic.

  • http://www.manapool.co.uk/ Evil Tactician

    It’s really not that complicated. Any individual which created content which could benefit from schema.org mark-up, should be able to implement the mark-up as well.

    Whether or not it will be useful, depends entirely on your website and the type of content you feature.

  • Durant Imboden

    It depends on the site and the business or organization that owns it.

    Here’s an example: Let’s say you own a dental clinic. With schema markup, you can make it easy for the search engines to display your name, business hours, address, and phone number directly on a local SERP. That’s likely to attract more business than a simple link to your Web site would do.

    Not every site owner is primarily interested in Web traffic. Brick-and-mortar businesses such as shops, clinics, restaurants–or local government services, for that matter–may be more interested in listings directly on the SERPs that function as free ads.

  • http://www.discountonlinefitness.com steve

    We use Magento for our ecommerce site. Unfortunately unless I hire an experienced programmer to edit the template files on our website it’s not an easy task to add markup coding to our product webpages. There are some really over priced plugins we can purchase that markup the pages but not sure if they actually work correctly and are up to par with googles standards. I just need more $$$ to get it rolled out! :)

  • http://www.powrsurg.com/ POWRSURG

    My company runs a custom-built CMS. We’ve built in a ton of options to let users add their own schema markup, and we automatically include some markup to mark them as pages of type WebPage (which can easily be changed), unless it is designated a product page. Off the top of my head I only know of one client that has actually used it themselves, and they are a client that has paid someone to do SEO/SEM/SMM.

    While some of this markup is great, there are A LOT of definitions that have no examples on schema.org. For example, why should I implement SiteNavigationElement?

    Google has a section which explains how some microdata will be used in their search engine. That’s great and worth adding support to cover those. But what about the rest?

    Also, am I just out of the loop, but I feel like they have been silently adding new definitions without telling anyone. I keep on discovering new definitions every so often without seeing them announced.

  • http://andymci.com/ Andy McIlwain

    I think that the user-facing implementation of schema.org within authoring/publishing tools (e.g. CMS platforms) is best kept to the developers/contributors of those platforms.

    In WordPress, for example, we have plugins like All In One Schema.org (https://wordpress.org/plugins/all-in-one-schemaorg-rich-snippets/) that make the experience far less daunting for users.

    There’s definitely a need for more education around proper schema.org usage, however. The “what and why”, rather than the how.

  • Christopher Shaffer

    If you have a basic understanding of html, it really isn’t that hard.

  • Christopher Shaffer

    Steve, you should be able to implement it with some basic HTML knowledge. Feel free to send over any questions.

  • http://www.manapool.co.uk/ Evil Tactician

    Exactly. It’s actually incredibly simple – so the term technocentric is rather far from the truth.

  • http://www.seoskeptic.com/ Aaron Bradley

    I’m sympathetic to your complaint that schema.org is “too technocentric.” That is, insofar as there could be better, more user-friendly documentation and implementation guides.

    Aside from that I’m curious, though, as to what sort of “usability tests” you think should be run – or, more precisely, on what web-accessible artifact or artifacts?

    schema.org itself – and I’m speaking here of the thing, not the website – is a formal vocabulary, and as such (in order to work with syntaxes like RDFa, microdata and JSON-LD) is necessarily technical – although I don’t know if it’s the technical nature of the vocabulary itself that leads you to call it “technocentric.”

    Perhaps not because you do speak of putting “that information into a better frame of reference.” On that point I agree, and for my part have been doing what I can to help provide that perspective.

    Perhaps its a failure of “the [schema.org] group” not to have “reached out” to non-technical stakeholders more, but anybody can make contributions to that group, which is actually made up of people that care very much about the vocabulary, and are very responsive to criticisms and suggestions for improvements.

    And actually one of the strengths of schema.org is that – for an initiative that was initially put together by the major search engines behind closed doors (as it were) and announced as a done deal to the world (to much criticism), subsequent development has largely been conducted collaboratively, in the open, with any stakeholder (including information architects and usability experts) able to make contributions (huge kudos to Dan Brickley for fostering that environment): I’d love to see you and other information architects take part in those discussions, and help make schema.org a better vocabulary, and implementing it an easier experience.

    Are they “arrogant”? Sure, some people surrounding schema.org are, but the bulk of them aren’t. And in a process in which one isn’t prohibited from participating, allowing the arrogant to ride roughshod by withholding that participation only reinforces that arrogance.

    Sure, fighting such battles (real or perceived) can be tough, but I think worth it. A highly skilled SEO developer I know, too, initially railed against what he typified as “elitism” in that community and distanced himself from it. I convinced him to give participation another shot, and he’s since become an extremely active – and respected – voice in ensuring the concerns of marketers and developers are accounted for in the vocabulary development process, and schema.org is better for it.

  • http://www.aplus.net.nz/ Tanzil

    I totally agree with you on the benefits of using Schema. But
    I think for the most site owners, it is a bit hard to use Schema. To use Schema,
    you need little bit knowledge (how to use) about it. Since most site owners are
    commonly non-technical people, they may be a bit afraid of using it.

    Another point is – I just checked my Yoast SEO Plugin. It is
    commonly used by many WordPress site owners like me. I did not find any option
    for Schema feature. I doubt if any other good plugin like this offers schema

    Last point, we do not regularly find schema as a hot topic.
    So, many may have no heard of it yet. I guess schema is still a hidden mine.

  • http://www.instantatlas.com/ David E Carey

    Yeah I’m sceptical about Schema (a little). I have 135 pages recorded in GWT as having structured data. However on recently testing one of these pages which has been established for many years may I add. The meta description is served in SERPs not the metaitemprop tag for description. So what is it doing? If I remove meta name =description will it simply elect metaitemprop tag for the page description.

    If so then that means that meta name=description is still the default choice for presenting a text overview of my page in the SERPs.

    I’ve not go into much detail on this but wonder if anyone has experienced similar analysis

  • hdc77494

    Help me out Here. Are the vast majority of webmasters really that incompetent or is there a lot more to this story?

  • http://ultimateunit.com Gunjal shrivastava

    Is there a guide to this available, most preferably an ebook.

  • http://bloggersideas.com/ Jitendra Vaswani

    yeah ri8, pakistan is full of spammers

  • http://www.amazon.com/gp/product/B00J0DRS7G RealWorld

    Google simply can’t find it. I’ve added the markup to both HTML and WordPress sites using the schema.org mark-up tool to generate the code, then go to Webmaster tools to verify it’s found and I have never had Google find it. I’ve tried everything I can think of, verified it’s in the right place, right format, and even put it in the wrong place just in case the instructions were wrong….still nothing. As it is, it’s worthless.

  • SEO Dim Sum

    In my experience, it’s easy to mess up schema mark-up. One wrong tag, or one wrong placement of a tag will completely ruin it, despite what Google’s mark-up checking tool tells you. My best advice, and what I did for the e-commerce site I work on, is to find a similar site who has mark-up working for them in the SERP (for example, I looked at what Zappos and other e-commerce sites were doing for their product/offer markup) and looked at their source code to see what they were doing to make it work. I also plugged in URL examples these sites into Google’s checker to see what it would return and used that as a way to triage my site’s issues. This approach worked for me and helped me get the product/offer mark-up to work.

  • sharithurow

    “Incredibly simple” is contextual. I’m a developer. I implement schema.org. I recognize that it is easier for me to implement than others.

    Maybe I see it more because I actually conduct usability studies and work with people who are learning how to do mark-up. I see how people struggle with it and the instructions.

    I stand by my comments.

  • http://www.manapool.co.uk/ Evil Tactician

    Anyone who struggles with as simple a mark-up as schema.org, shouldn’t work in development.

    Schema.org *is* aimed at developers and people who *implement* the code – anyone with a basic grasp of code or mark-up language such as html should have an easy time with it.

    If they’re struggling – you need to question either the people you are teaching or your teaching methods. I stand by my comments.

  • sharithurow

    Hi all-

    Wow, I did not expect sucha discussion on schema.org, either. Pretty cool.

    Gentlemen, and please correct me if I am wrong, I don’t believe you are aware of a number of things:

    (1) Search is a part of information architecture. Many information architects are search (information retrieval) experts. They have considerably more technical knowledge than the average person.

    (2) Information science(s) is also a field of study and work. If you read this excellent book called Handbook of Information Science by Wolfgang G. Stock and Mechtild Stock, you might have a different opinion as to what constitutes “technobabble.”

    (3) I have worked with and studied under usability professionals who have degrees in artificial intelligence, human/computer interfaces, usability engineering, and so forth. I do not immediately stereotype and classify usability professionals as nontechnical.

    My usability colleagues and I understand that “incredibly simple” is contextual. I think mitosis and meiosis is incredibly simple to understand. I don’t insult people because they do not believe these fundamental biology concepts are not easy to understand.

    (4) Librarians and information scientists are pretty skilled at taxonomy and ontology. In fact there is a very technical ontology forum people can participate in. I believe Peter Yim is in charge of it.

    So with all due respect, I don’t believe schema.org is only for developers. Just as SEO is optimizing a website for people who use search engines, schemas can be used to optimize listings for PEOPLE WHO USE search engines. IMHO, I think schema.org is in desperate need of some skilled taxonomists.

    In summary, I think some of you are really missing the human (factors) part of human/computer interfaces…and that includes some folks at schema.org. It’s my opinion.

    Quick note to Aaron: I am serving and have served on multiple boards of professional associations that understand human/computer interfaces. I am more than doing my part in the search, IA, UX, and usability industries.

  • sharithurow

    Hi Aaron-

    The usability test I would probably do (multiple iterations) is a performance test. My usability director has over 20+ years of experience. She has tested software and other interfaces many, may times.

    That’s my gut answer. The tests I’d select depend on context.

  • http://www.seoskeptic.com/ Aaron Bradley

    Thanks for your considered response Shari.

    I can’t answer for anyone else to whom your comments may be directed, but I certainly aware of all four of your points (particularly #4, as I worked for a decade as a technical services librarian).

    I absolutely agree that schema.org is not just for developers, and I’ve certainly done what I can to reiterate that message with schema.org sponsors.

    As to schema.org being in “desperate need of some skilled taxonomists,” the initiative is awash with taxonomists and ontologists. If there’s a failure in human factors design pertaining to schema.org it is not due to a lack of taxonomists working on the initiative.

    And I wasn’t suggesting that you personally weren’t somehow pulling your weight in not taking an active part of this initiative, Shari (I would think the “I’d love to” in “I’d love to see you and other information architects…” expresses desire rather than admonishment, but if there’s any doubt, I did not mean to admonish).

    Let me, however, expand on what I said in the interests of making my point more clearly. While much of schema.org originated from a top-down process from its sponsors, the development of the vocabulary has been highly collaborative. This makes it relatively unique in the world of search engine development, which is normally quite prescriptive.

    If you or anybody else finds the vocabulary wanting you’re of course justified in describing those shortcomings without doing anything personally to rectify them. But because it’s a collaborative process I don’t think it’s as contingent upon the sponsors themselves to take up perceived slack – say, for bringing on board taxonomists.

    To that end I was remiss in not giving a shout-out to the other schema.org Dan. :) Namely Dan Scott, Systems Librarian for Laurentian University, who has made innumerable valuable contributions to the vocabulary’s development.

  • sharithurow

    Good to know. Thank you.

  • John Biundo

    I wasn’t sure I understood the adjective technocentric in this context (and yes, I googled it), but I think I can guess from Shari’s comments, and the ensuing discussion, that she’s saying that a lot of webmasters/web developers have a hard time applying schema.org markup.

    Some of the responses to Shari suggest that “it’s really not that hard”. And I doubt that Shari’s saying *she personally* has a hard time with it. But it does seem valid to suggest that the user facing documentation for applying schema.org markup to a web site is fairly lacking. As are the tools and the support channels. Taking a tour of a decent sized sample of sites that have implemented it (and inspecting their markup) is enough to demonstrate that many people aren’t getting it quite right. It’s actually remarkable to me that the search engines are extracting as much signal as they are, given all the noise.

    It also seems true that simple cases are simple, and complex cases can be extremely hard. What are complex cases? Deeply nested structures (think localBusinesses with products, offers, reviews, and aggregateRatings). The concepts at work here are pretty simple, but the syntax and structure can be a bit daunting. Again, with still immature documentation, tooling and support, this can leave many frustrated.

    A big complicating factor that I’ve seen is with e-commerce sites (among others) that have a fair degree of automation in their page production process. This could be a full blown CMS, or could just be lots of dynamic content creation. In either case, getting the semantic markup embedded properly is a real challenge. It’s not a simple matter of reaching in and dropping an itemprop here or there! It’s a complex orchestration to get it right. This particular issue doesn’t seem to be a flaw of the schema.org vocabulary per se; any embedded markup would have this same problem.

    When you combine deeply nested structures, automated page production, and an immature support infrastructure, I’d have to agree that getting it right can be a real challenge. But I don’t believe for an instant that this suggests that there weren’t “proper architects”, or that usability tests or “reaching out to professional information architects” are the cure, or that schema.org is inherently flawed. And I haven’t personally witnessed any “arrogance” on the part of the group’s leadership. To the contrary, there is a remarkable openness and willingness to help individuals and interest groups who take the time to ask.

    One area I hold a lot of hope for is the use of JSON-LD. Google has been somewhat circumspect about their support for this syntax as a source of rich snippets. On the one hand, the markup helper tool produces JSON-LD syntax for many of the major schema.org types. Good sign! On the other hand, the structured data testing tool doesn’t read JSON-LD, so you can’t validate it! On the third hand, there IS a wonderful JSON-LD testing tool buried in the Gmail Actions in the Inbox section that DOES read, extract, and validate JSON-LD! So all signs point towards JSON-LD becoming a first class citizen in the schema.org syntax family. Hooray – it can’t come soon enough.

    The big advantage of JSON-LD is it avoids the above issue with properly weaving schema.org Microdata markup into your page production process. You simply insert an island of code in a script tag in the header, and you’re done. I’m very curious whether anyone has found any instances of JSON-LD in the wild that yield rich snippets. I’m running some tests myself and hope to come to some conclusions soon, but hoping even more that Google comes out with some clear direction on their support for JSON-LD and rich snippets.

  • sharithurow

    Thank you, John. You were accurately interpreting my thinking and communication.

    JSON-LD sounds promising.

    I don’t believe that the schema.org folks are arrogant. I wasn’t aware that there were taxonomists working on the project (guess I was having a mind fart that day – there are technocentric taxonomists in the world).

    In the tug-of-war between users vs. technology, I see users getting the “proverbial short end of the stick” on a daily basis, even when I’m not working. I constantly witness and hear complaints from “arrogant” support people.

    I get it — support people see and hear the complaints all of the time. They see the answers all of the time. So the answers seem obvious to them.

    I say that if the answer is so damn obvious, then why does support keep getting calls, emails, requests, etc. about the answers? Support people often sigh (I hear them on the speaker phone and note the number of sighs I hear), and blame the users (“well if you just did this…”).

    The problem often lies with the interface and the people who came up with the unusable interface and its support materials…comments that make me REALLY popular with support staff.


    Insulting people who are new to an interface is not a way to encourage them to learn more.

    I would support schema.org more if they had more people who are truly working for better interfaces…and that means taking constructive criticism and advice from information architects, user interface and usability professionals.

    I’ll leave it at that.

    Thank you, again, John.

  • John Biundo

    My guess is that the lack of clear direction with JSON-LD is that it presents an opportunity for spamming. With any markup, it’s possible to add data (e.g., meta elements) that’s hidden from users and opens the door to spamming.

    Google tries to limit this potential via quality guidelines, but with JSON-LD, and the ability to drop in semantic data that’s potentially completely disconnected from the display, it seems the door is cracked somewhat wider.

    And even if spammers aren’t charging through that door — the burden of synchronizing what’s visible on the page with what’s “invisible” in the JSON-LD scripts is somewhat greater. So the really interesting question is: does JSON-LD work today (does it actually produce rich snippets in the SERPs)? As I mentioned, the tools are there to generate it, it’s documented on schema.org, it’s extracted and validated by the email markup tool, so it seems like it *ought* to work, but does it?

    Here’s hoping that Google (and the other data consumers) figure out how to address this issue (it seems like they should be able to), and we all can take advantage of the easier integration of JSON-LD into the page production process.

  • Christopher Shaffer

    @sharithurow:disqus I am NOT a developer and only have a basic understanding of html and can implement schema with no problem. You “actually” conduct usability studies? Did you conduct a study on schema implementation? What was the sample size and experience of the group?

  • sharithurow

    Guys, I don’t want to debate this anymore because at the core of usability is understanding that your personal opinion and mental models are not necessarily the opinions and mental models of actual users.

    Continuing to say/type things such as, “I think it’s easy…” or “I’m not a developer…” indicates to me that you might not have this understanding of usability.

    Without a common understanding and vocabulary, I don’t see that continuing to comment on this thread will be useful to anyone. So I’ll put in some useful resources as my last comment in this thread.

    I write about usability and search a lot in Search Engine Land and Marketing Land. That provides my frame of reference. But I suggest reading books by Jakob Nielsen, Alan Cooper, Ginny Redish, Peter Morville, Louis Rosenfeld, Eric Schaffer, and Jared Spool.

    Humanfactors.com has a good newsletter. UIE.com has good webinars and podcasts. Susan Weinschenk (The Brain Lady) has an excellent blog.


    Here’s a good article on personas by Kim Goodwin:


    Oh yeah! Edward Tufte is always interesting, IMHO.

  • John Biundo

    Hi Shari,

    I know you wanted to make this the end of the discussion, but I think discussing usability of schema.org is extremely useful and important, and while this might not be the best forum, it is widely read, so I’m going to continue with another few comments.

    I spent an hour this morning with a really smart SEO guy who is struggling to understand how to best markup reviews for a major site. There are just a bunch of difficulties in getting this right. I’ll mention a few:

    First, it’s important to remember that Schema.org is a consortium, and their documentation about schema is NOT logically the same thing as a user guide for “how should I mark up my pages to get rich snippets”. Google is a (contributor to and a) consumer of schema.org information, and has their own rules (algorithms if you wish) about when to award rich snippets. Why does this matter? Because you currently cannot go to one place and understand exactly what markup Google wants for rich snippets.

    The Google rich snippet docs are supposed to address this need, but they are deficient for two reasons. First, they have only been partially updated to reflect schema.org (they appear to have been originally written based on other vocabularies (e.g., hReview/microformat), and partially updated (with some example code) to schema.org).

    For example, the Rich snippets – Reviews page (https://support.google.com/webmasters/answer/146645) refers to hReview in the main body, and only provides a code example with schema. Because the models differ, there’s an impedance mismatch in trying to follow the guidelines. Case in point: the discussion on this page refers to “Review-Aggregate”, but there is no “Review-Aggregate” type in schema.org. With some work, you end up deciding that what they are saying here (about the concept of “aggregated reviews”) maps to the schema.org type aggregateRating. Terminology soup! It requires way too much energy to map the guidelines to an implementation, causing you to have to constantly cross-reference the guidelines page, the schema.org page,and make leaps of deduction and intuition.

    Second, they currently REFER to schema.org, but since schema.org doesn’t define what’s required vs. optional, and Google doesn’t replicate the model and annotate it with these attributes, you’re left wondering.

    Some would argue that “the one place to go” is the structured data testing tool. But the tool only tests your results and gives limited feedback on structure and syntax errors. To give a very specific example: is the itemReviewed element required on a Review? There is no easy answer to this question. Schema.org doesn’t indicate required properties (nor should they; they provide a model, and the consumers (Google in this case) decide how to process it. It’s up to Google to say whether it’s required).

    Now they do indicate that itemreviewed is a required property, but they do so in the context of hReview, not schema.org/Review. So you’re left to generate some sample markup, run it through the Testing tool, and see what happens. If you omit this property, the tool complains with a generic “Error: incomplete microdata with schema.org” message. What caused the error? You don’t know until you test with and without the itemReviewed property in your markup, whether that’s the problem or if it’s something else.

    Google could help by providing more detailed documentation, and also by enhancing the testing tool to provide better feedback.

    Then there are larger questions such as this example: what to do if you have a page that has both an aggregate rating and individual reviews (as is the case on many IYP type sites on their merchant pages). Google seems to say (on the same page I listed above) you shouldn’t mark up both, but conceptually this seems like bad advice. And in fact, if you DO mark up both, it validates fine in the testing tool and you can get rich snippets for the aggregateRating in the SERPs. So marking up the reviews is not “harmful” (if your aim is getting rich snippets), and it sure seems like the right thing to do from a semantic web perspective. Why is there confusion in the Google documentation?

    This kind of ambiguity, lack of detail, and inconsistency is rampant. As I’ve said before, it’s amazing to me that the search engines actually DO extract meaning from this noise. It’s a tribute to some amazing engineering. But the whole process needs to be WAY easier. In my opinion, the good news is that there’s a solid foundation, and the need is in the areas of tooling, documentation, and best practice development in the SEO community. It’s more a sign of lack of maturity than of fundamental flaws. So I think we can get there on this ship if we can get the right oars in the water and pull in the right direction!



  • Christopher Shaffer

    I’m not trying to back you into a corner by asking about your experience with UX. I’m simply asking what studies you have done or seen related to schema that cause UX issues? From a developer standpoint, if they can’t write schema, they can’t write html. From an end user standpoint, if schema is implemented correctly it doesn’t impact UX whatsoever. I’m simply curious as to where your opinion was coming from, not trying to attack you.

  • sharithurow

    Wasn’t taking it as a personal attack. I think this is surprisingly (as in unexpected) an interesting topic.

    Other stuff is going on with me, guys, that has priority. I’ll revisit this topic some other time.

  • http://www.manapool.co.uk/ Evil Tactician

    I too would be interested in your answer to Christopher’s question. You mention you’ve done a study on the usability of Schema.org yet you refuse to even make a basic comment or example on it despite the fact that this would be incredibly relevant and interesting to this discussion.

  • sharithurow

    The studies I do are paid for by clients and, therefore, are not available for the general public. I respect and honor my clients’ privacy.

    I communicated what i wanted to communicate, for the time being.