Skip to content

A Marketer’s Guide to the Next Web

2010 October 13

Most people think that the Web is a pretty big deal.  It created a universal platform for communication and content that unleashed an astounding amount of creativity.

However, there is a second Web brewing that could be even more powerful.  We won’t see much of it, because it won’t be a platform for us, but for machines and it will unleash the power of information by linking data that we have long sequestered in disparate places.

The consequences for marketers will be profound.  New terms such as metatags,  API’s and the web of things are  already entering the media lexicon.  The next Web will fundamentally change not only the way we interact with consumers, but how we work with each other. Here’s a summary of what’s going on and what to expect.

The First Web

Tim Berners-Lee created the basic architecture for the Web in 1989 as a way to organize documents for CERN, the European nuclear research lab.  Physicsts were always coming and going and needed a system that would allow them to access and share documents no matter what software they were written on.

His invention consisted of three components: URL, HTTP, and HTML.

The first two are mere formalities.  To get an idea of what they do, think about a standard business contract which, in the preamble, designates the parties and the body of law under which the contract will function. URL and HTTP basically form standards for computers to identify each other and signal how they will interact.

The third part, HTML or the HyperText Markup Language, however, was a stroke of genius.  It’s what makes the Web so special.  There are two basic elements:

Markup Tags: These are what give the web its universality.  They work very much a the script in a play.  Just as scripts have little instructions in the text so that the actors will know to present the dialogue (wryly) or (angrily), HTML tags instruct computers to <make this bold> or <make this blue>.

HyperText links: These give the web its connectivity.  Berners-Lee originally created the web so that people could find important documents.  He realized that any hierarchical system would fail because importance is all about context.  Therefore, he made it easy for anyone with a web page to tell people to <go to> to find information that might interest them.

So the first Web was about being able to find and share documents.  Over time, those documents have become more sophisticated; they have become dynamic (i.e. they change regularly) and can include video, etc., but the basic premise hasn’t changed.

The Next Web

The Next Web, or the Semantic Web, goes a level deeper.  It’s not about documents themselves, but about the data that underlies them.

Imagine you want to sell car.  You can upload the specifications to different web sites and the users of those web sites can see what kind of car you have, what price you are offering, and so on. However, what if the person who wants to buy your car doesn’t go to the sites that you’ve posted on?

You could sell your car much more easily if you could just upload it once and all car web sites could access it.  The key is to get computers who don’t speak the same language to understand that they are talking about the same thing.

The Semantic Web incorporates two technologies to do this: Metatags and Ontologies:

Metatags: Metatags are “information about information.” A new markup language, RDF (Resource Description Framework) , does for data much of what HTML does for documents.  It marks up databases with universal tags that can be interpreted elsewhere.

Ontologies: These work like cross language dictionaries.  They interpret the RDF tags and their relation to other RDF tags.  Of course, ontologies can be either global or local, so they can apply to the web as a whole, a specific industry or even a certain type of file.

The two technologies, when used in conjunction, result in universal databases. Much like the first Web can be accessed from any browser, the Semantic Web will allow us to combine data from a variety of systems and platforms.

For instance, in our earlier car selling example, the company who made the car can be referred to as a “manufacturer,” a “producer” or a “make.”  A person would recognize these terms to be equivalent but a machine wouldn’t, which makes it difficult to combine databases.

However, with RDF tags and Ontolgies, a database using one set of terms can be translated into a universal database. So if car sites had their databases tagged and used a common ontology you would be able to upload your information just once and it will be in a form that anybody who wants to sell your car for you will be able to use.

What’s Going On Now

Mostly due to the enormous amount of semantic infrastructure that needs to be built, the full impact of the Next Web is probably still a decade away.  However, real technologies are in development and some have already been deployed.

In his book, Pull, author David Siegel gives a range of concrete examples of the changes underway.  Here are some of them:

Retail: At the Metro Group’s Future Store Initiative, they are developing a number of semantic initiatives combining technologies such as RFID, mobile shopping assistants and deploying metatags for quality assurance of perishable items.

Finance: There is already a semantic financial reporting language called XBRL (eXtensible Business Reporting Language) which not only helps create financial reports, but allows for automatic filing with regulatory agencies without having to reformat information.  In effect, the database is the filing.

Not only will XBRL reduce costs, it will help cut down on errors and increase transparency.  It has already been adopted by institutions such as  U.S. Federal Deposit Insurance Corporation (FDIC) and the Committee of European Banking Supervisors (CEBS)

Construction and Architecture: Much like semantic technologies can weave together all the complex and diverse data elements that go into building financial reports, the building information management (BIM) standard can do the same for designing and constructing buildings.

These are just a small slice of thousands of efforts underway.  To get an idea of what’s ahead, take a look at this short TED talk about open data in action by Tim Berners-Lee himself (post continues below)

What Semantic Technologies Will Mean for Marketers

While there are many exciting things going on with semantic technologies, the impact on marketing and advertising has been minimal so far.  RDF tagging is a time consuming, labor intensive process and the amount of data to be tagged is massive.  A truly semantic marketing environment is a long way off.

Nevertheless, the possibilities are exciting.  Here are some of them:

Research and Analysis: A enormous amount of time an effort go into combining data from different databases.  We have one set of systems for expenditures, another for consumer data, another for media consumption and still more for our own proprietary research.

Being able to combine all of these things at the database level would not only save a lot of time, our marketing product would improve immensely.  As more media becomes digitized, we will be able to generate insights across the marketing landscape, from broadcast, to social to POS.

Planning and Buying: Probably the least efficient area of advertising is media planning and buying.

Planning is usually done in one software platform (although often different ones for different media), buying in another and billing in a third financial platform.  In some cases, an agency or media supplier have integrated everything internally, but they still are using separate systems from other market players reducing efficiency and increasing errors

Obviously, being able to stitch together all the various platforms using one industry ontology will vastly increase productivity.

Point of Sale: To effectively utilize point of sale advertising, you need to understand which points of sale are the most valuable.  That means you need to combine data from dozens or even hundreds of retail partners all with separate systems.  Again, this is a classical semantic problem that the Next Web will be able to streamline.

ROI: One of the most difficult things about optimizing marketing ROI is that there is so much data in so many places, just getting it all in one place is a challenge.  So by making databases universal, we’ll not only be able to make marketing more efficient, but more effective too.

What Won’t Change

While many digital advocates believe that the Semantic Web will bring about an exclusively “pull marketplace,” that’s unlikely to happen.   As emerging technologies make transactions easier to facilitate, brand preference will only become more important.

That means creating and broadcasting your message will still be the most important thing marketers do.  Although the tools will change, and we need to learn them, the basic marketing job will still have to be done.

– Greg

10 Responses leave one →
  1. October 13, 2010

    Great article on semantic web. One added point. Even without “standards” noted, automation and other IT tools allow for marketers to access huge data mines and multiple sources of customer info on the fly today in ways that were unimaginable just a few years ago. So let’s say you are a CVS Extra Care loyalty member. Until now, it could track that you have made a purchase, and perhaps know your purchase volume, and determine that you are now eligible to for the special “seasonal suntan lotion offer” with a coupon printed on the receipt. Static. Large retailers are now building IT infrastructure that can merge your contact info including mobile, along with past purchase behavior and even credit card purchase data to deliver a personalized and much more relevant offer in real time. Say you have allergies and purchase Claritin every spring. Members can literally walk into the store, have data reviewed and receive for example a “buy 2, get free” offer for Claritin delivered to your phone right on the spot. The capabilities exist and are being built by some leading retailers here, now.

  2. October 14, 2010


    Great example! Unfortunately, most of the systems now are proprietary. The semantic infrastructure underway will help facilitate data connectivity and make the whole process more efficient. However, the privacy issues are obvious and will have to be dealt with.

    Thanks for a fantastic comment!

    – Greg

  3. October 16, 2010

    Good post, Greg. It provides an exciting vision how semantic web can change marketing and makes me think. Standardized data format will obviously create a new digital world.

  4. October 16, 2010

    Thanks Stan!

  5. fenderbirds permalink
    October 18, 2010

    nice article, keep the posts coming

  6. October 20, 2010


  7. November 12, 2010

    Fantastic article! I am reminded, though, that there are areas (such as my town of Chicago, IL) where thousands of residents don’t have Internet connectivity unless they visit a library. They don’t have smart phones or computers. They live in impoverished neighborhoods but they have some money to spend and can be smart consumers. I have met several business leaders who assume “everyone” is connected digitally and paper (of any sort) is dead. I hope marketers don’t make the same mistake and overlook these digital deserts.

  8. November 12, 2010


    You make a very good point. Thanks for the reality check.

    – Greg

  9. walter daniels permalink
    June 15, 2011

    I love the final comment in your article. In truth, the more things change, the more they stay the same. I “stumbled” across the ‘Net in about 1994, joining an e-mail list for people interested in advertising. I look at what it’s like today, and it blows my mind.
    But, I also see how much the old basics not only apply, but are even more important.

  10. June 15, 2011

    Thanks. Walter. Glad you liked it!

    – Greg

Leave a Reply

Note: You can use basic XHTML in your comments. Your email address will never be published.

Subscribe to this comment feed via RSS