This is the final market we look at in our Creative Destruction 7 Act Play series.

In Part 1, we looked at the overall market for enterprise software to see where semantic web technology could fit. The basic conclusion: it is part of the data integration business.

In this post we dive a bit deeper into the types of opportunity for semantic web vendors and how they can position to win a big share of the $229 billion enterprise software market.

The 7 Pillars

One of the best insights on semantic enterprise comes from MK Bergman's January 2010 post on Seven Pillars of the Open Semantic Enterprise. I will quote liberally from his post but do read the whole thing and subscribe to his RSS as he writes very well. You can get the essence from this picture:

Here are Bergman's 7 Pillars, with my comments.

Pillar #1: The RDF Data Model
Bergman points out that "there are now more than 150 off-the-shelf ‘RDFizers’ for converting various non-RDF notations (data formats and serializations) to RDF. What this practically means is that the integration layer can be based on RDF, but that all source data and schema can still reside in their native forms."

That fits with what we are seeing in the consumer web - RDF momentum is strong.

Pillar #2: Linked Data Techniques
This is a no-brainer. The beauty of the LinkedData model is that it enables any business group to be a publisher. This is like a better version of the Intranet and is close to the Publish and Subscribe model that business folks find intuitive.

Pillar #3: Adaptive Ontologies
Bergman sees "a shift in roles and responsibilities away from IT to the knowledge workers themselves". He believes that the "any subject matter expert or knowledge worker likely has the necessary skills to contribute to useful ontology development and refinement."

This sounds like the data modeler telling the business folks that "really, it is quite simple". Quick, tell me the fundamental difference between and ontology and a data model. Until creating ontologies is as easy as creating a spreadsheet, ontologies will remain the province of the specialist, just like data models. Which is fine, but not the crossing the chasm breakthrough we are looking for.

The good news is that this represents a great opportunity for an inspired, breakthrough software product. Watch this space!

Pillar #4: Ontology-driven Applications
Bergam writes that "Ontology-driven apps fulfill specific generic tasks" that "accommodate a relatively common abstraction layer that responds to the structure and conventions of the guiding ontologies." In his vision, this "limits software brittleness and maximizes software re-use."
This sounds very much like Model Driven Architecture (MDA) which in highly skilled hands has produced marvelous results. But very, very few people know how to do this well. When average skilled developers attempt MDA, the result is what Joel Spolsky calls "architecture astronauts" that are always architecting and never delivering.

Pillar #5: A Web-oriented Architecture
Bergman says that a "Web-oriented architecture (WOA) is a subset of the service-oriented architectural (SOA) style, wherein discrete functions are packaged into modular and shareable elements (”services”) that are made available in a distributed and loosely coupled manner. WOA uses the representational state transfer (REST) style."

This seems totally right and Semantic Web technology fits the WOA neatly. This will be key to adoption and ties in with his 7th Pillar.

#6: An Incremental, Layered Approach
Bergman reminds us that "Semantic technology does not change or alter the fact that most activities of the enterprise are transactional, communicative or documentary in nature." Therefore "an obvious benefit to the semantic enterprise is to federate across existing data silos."

This is where he shows the layered architecture we presented earlier. My problem us that "federate across silos" sounds worthy but does not sound like a business case. The killer app is still missing.

Pillar #7: The Open World Mindset
Bergman writes that "Enterprises have been captive to the mindset of traditional relational data management and its (most often unstated) closed world assumption (CWA)." He goes on to say that a "closed-world mindset carries with it certainty and logic implications not supportable by real circumstances."

He says "this is not an esoteric point, but a fundamental one" and I agree. Getting budget for a purely INTERNAL project is tough. The existing vendors/technologies have the inside track and internal projects tend to be about cost saving. External projects that bring in new technology/products or open up new markets are strategic and get CXO attention. This is where the enterprise semantic projects are likely to be found.

Semantic Standards Across Enterprises

This stylized representation of an enterprise comes from Collibra:

Collibra are selling something they call "data governance". They point out that despite decades of effort, data quality is still pretty bad. So business users cannot really rely on the data and spend way too much time double checking, re-verifying and reconciling data.

But Collibra's picture illustrates that it is the external data that matters. That is where the $$$ flows from. So it is no use having internal semantic standards. These have to work across any enterprise. That is why we see a lot of momentum in cross-enterprise semantic standards such as:

Bio RDF for Bio Pharma data

XBRL for financial information

The success stories around Bio RDF and XBRL illustrate another key point:

End user solutions for line of business managers get traction first.

IT may be involved with decisions around systems that use Bio RDF and XBRL, but the real decisions are taken closer to the business front lines. In both cases there is a divisional budget other than IT - R&D in the case of Bio RDF and Finance in the case of XBRL.

Knowledge Management Is In A Coma. Long Live Business Intelligence

Semantic Enterprise vendors tend to pitch their solution as "knowledge management". This sounds good, knowledge is a key resource and so on. But KM has been even longer in the oven than semantic web and has lost a lot of credibility. It is hard to find a KM budget.

But you can find Business Intelligence (BI) budgets and they are growing rapidly even in a recession.

According to Gartner:

"The worldwide appetite for BI platforms, analytic applications and performance management software in 2008 increased 21.7% on the previous year, from $7.2 billion to $8.8 billion."

One reason BI is growing fast is the explosion of data created by online systems. So this is ideally suited to Web Oriented Architecture as described by Bergman.

BI projects can be big or small and they can be close to line of business. In other words you can get a "quick win" in the BI world. All the IT controlled data technologies/processes - EAI, ETL, Messaging, MDM, Data Quality, Search and yes, Semantic Web - are enablers for BI.

If Semantic Web is a better tool for BI, it has an enduring and central place in the enterprise. If not, it will go the way of many once-promising-often-hyped-but-now-forgotten technologies in the dustbin of IT history.

Anzo: It's The Spreadsheet Stupid!

Cambridge Semantics is another Semantic Enterprise vendor. They have a broad, horizontal product strategy and have focused on the one data tool that every business user tends to use every day - the spreadsheet.

To many of us, the spreadsheet is our BI, our MIS and our data management suite of tools. Sad but true.

Anzo enables data that has been isolated in individual Excel spreadsheets to be accessed, shared and managed. As Steve Kludt, EVP of Marketing explained in our recent review of Anzo, this means that these "spreadsheets can become part of the IT data structure--you get security, workflow, and governance.”

That is huge, because it "turns Excel into an application builder". In the history of IT we have seen countless technologies that would enable the average business to create applications. They all failed, despite $ billions invested, great technology and great marketing. Meanwhile, millions of business folk create applications in Excel every day.

Excel to relational databases has been much promised but failed to become mainstream. Excel to RDF maybe easier. If so, Anzo could be a killer app for the semantic enterprise.

The Bowling Pin Markets

In Crossing The Chasm terminology, a "bowling pin" is a niche market where there is a real need for the new technology. Once the technology is proven in a few of these niches, it may be ready to cross the chasm to the mainstream.

In the case of semantic enterprise technology, the obvious bowling pins are:

1. Intelligence. As in CIA, MI6 type of "Intelligence". Information is a) totally mission critical b) horribly complex c) distributed all over the place and d) unstructured until they apply some structure to it. Intelligence agencies tend to have close contacts to the academic world that is currently funding semantic web research. The trouble is for vendors, they cannot easily publish the case studies!

2. BioPharma. We explored that here. This still looks exotic to your average enterprise, but a lot less exotic than Intelligence.

The other markets are less obvious. There is traction all over the map. Semantic web is mission critical to Publishing, but the folks running Publishing today don't always see that. Huge data sets are critical to financial services, but the semantic traction seems to be unclear at present.

-----------------------------------------------------------------------
• Don't forget to propose your startup for our Semantic Web Impact Awards. The deadline is Sept. 15.