W3C home > Mailing lists > Public > public-lod@w3.org > July 2012

Re: Linked Data Business Models?

From: Kingsley Idehen <kidehen@openlinksw.com>
Date: Thu, 26 Jul 2012 15:52:07 -0400
Message-ID: <50119FE7.7090007@openlinksw.com>
To: public-lod@w3.org
On 7/26/12 2:45 PM, Mike Liebhold wrote:
> On 7/26/12 11:23 AM, Kingsley Idehen wrote:
>> Q: What does Linked Data offer in this context?
>> A: Powerful mechanism for data virtualization that turns a collection 
>> of disparate and heterogeneous data sources into a coherent, 
>> persistent, and change sensitive mesh of entities and their 
>> relationships. For a number of years, this has been achievable via:
>> 1. Linked Data Views over RDBMS data sources
>> 2. Linked Data Views over other data sources.
>> What do enterprises buy?
>> 1. Data Virtualization Middleware
>> 2. Database Management Systems
>> 3. Identity Management Systems
>> 4. Collaboration Systems
>> 5. Professional Services -- training, consulting, and custom 
>> development.
>> My $0.02 .
> I agree with all of these, Kingsley, but in the real world there are 
> serious organizational obstacles that need to be transcended for an 
> enterprise to appreciate the benefits you outline.
> I was once tasked with the design of a shared, interoperable content 
> repository for over 20 discrete publishing companies in one media 
> corporation (Times Mirror), and many of these individual publishers 
> had smaller business units with separately managed repositories.  It 
> turned out that the individual IT executives saw little local benefit 
> for the extra work labeling their content with consistent and extra 
> metadata for benefits in other organizations, and completely resisted, 
> so the project floundered and ultimately failed.

Yes, because they had to invest time they didn't have in a proposal that 
(in their eyes) was at best of mercurial value.

A Linked Data product, packaged appropriately, would basically alleviate 
their real concerns by getting to the value, pronto. In a nutshell, if 
Virtuoso (our product) was presented to the same managers, the value 
discovery sequence would go something like this:

1. install product
2. identify data sources and their associated access mechanisms 
(typically: ODBC or JDBC data source names, SOA service endpoints, 
document URLs etc..)
3. attach the data source names to Virtuoso using its in-built 
functionality and UI for this task
4. generate Linked Data views over these data sources -- basically 
push-button on go
5. open up the faceted browser interface to demonstrate all the new 
insights unveiled by the new entity oriented Linked Data view
6. show how ODBC, JDBC, any Web Browers etc.. has access to this new 
conceptual view
7. show how a single hyperlink is the basic unit of insight (report) 
8. show how everyone in the value chain can alter their view of the data 
exposed by these hyperlinks that have become the new canonical data 
source names (as opposed to the line-of-business dbms specific data 
source names from ODBC, JDBC etc..).

> In order to achieve the benefits you list, the most senior executives 
> have to be deeply sold, and willing to assert their authority over 
> local data fiefdoms.

Yes, and the value proposition as to be presented in a manner palatable 
to said audience. In this case the key message has to be about unified 
(virtualized) views over disparate data sources en route to improved 
enterprise agility etc..

We (vendors) must make value palpable via a minimal number of message 
exchanges with our customers and prospects. This is how its always 
worked. This is how the RDBMS market came to be many years ago. I know 
that because I experienced it first hand.

> fwiw.
> Mike Liebhold
> Senior Researcher
> Institute for the Future



Kingsley Idehen	
Founder & CEO
OpenLink Software
Company Web: http://www.openlinksw.com
Personal Weblog: http://www.openlinksw.com/blog/~kidehen
Twitter/Identi.ca handle: @kidehen
Google+ Profile: https://plus.google.com/112399767740508618350/about
LinkedIn Profile: http://www.linkedin.com/in/kidehen

Received on Thursday, 26 July 2012 19:51:28 UTC

This archive was generated by hypermail 2.3.1 : Wednesday, 7 January 2015 15:16:24 UTC