David Siegel, author of the excellent Pull, The Power of the Semantic Web to Transform Your Business, was telling me about his first book, published in 1995. This was just as Web 1.0 was getting momentum. His book was about web design. People were skeptical that it would sell. It was a great success. Very large numbers of people were looking to become web designers and there was a great skill shortage for a while which does wonders for fee rates and salary levels.
So, I asked David "where will the job creation come from when the Semantic Web goes mainstream?. What is the equivalent of learning web design in 1995?"
Has To Start With Management
David's view is that this has to start with management. Specifically he said:
"I think we'll see technologists migrate from all angles. At the moment, there are plenty of people involved in machine learning, and standards can be implemented by all kinds of technical types. To me, the interesting questions are - where will the marketing people come from? Where will the managers learn about scale? Where will we get the CIOs who can stop the bleeding of IT department spending on projects that don't scale up?"
So I decided to approach the same question from a different direction by talking to an entrepreneur who is currently running a start-up that is in hyper-growth mode (actually in that interesting transition from startup to large company).
The Entrepreneur's View
The entrepreneur is running a business that relies on generating a ton of traffic at the right price and then converting those clicks into orders. They operate a super-efficient marketing funnel.
This is what lots of companies do today.
We were catching up after a long time. I told him about my work related to the Semantic Web and showed him David Siegel's book.
He had heard about the Semantic Web, but he knew the name and that was a about it. His initial take was skeptical. This sample size of one echoes what the Pew research is telling us.
His first crack was "let's just wait for 4.0". The image he had of semantic web was that it was hype. He specifically asked "is this like Cloud. Is it just a marketing term, a concept? Or is it something I can implement and if so, why and how?".
That is basically the same question that any manager/entrepreneur wants to know. Those of us in the semantic web community need to be able to answer that succinctly and simply.
Of course, the answer will depend upon the business. But what this entrepreneur was looking for - tons of good quality traffic that can be converted easily to revenue - is what a lot of other entrepreneurs/managers are looking for across all kinds of markets and industries.
As we dug into this, it became clear that, for this business, the semantic web proposition related to SEO.
When this entrepreneur had understood the basic idea that semantic tagging (using RDFa or microformats) improves findability by search engines, the interest level went up.
The company had done a good job putting their knowledgebase online and using this to get traffic. But he wanted to take this to the next level. He was not just looking for traffic. He wanted good traffic, highly convertible traffic.
So as we dug further, the benefit of adding semantic tagging to their content was that the searcher would arrive more quickly at a point in the knowledgebase that was directly related to the searcher's problem. That would definitely enhance chances of converting that searcher to revenue.
Now that the objective was clear, the question was "how do we do this?" This is where it gets back to the mainstream jobs question. To this business, the person who would drive this would be the person in charge of SEO. Which is really the person in charge of content.
Power Tagging AND Ontologies
Explaining "power tagging to add some structure so search engines can find your content better" is the easy bit. But then we had to get into ontologies. You cannot be semantic unless you have an ontology, right?
After my rather weak quick description of ontologies (note to self, work on that), he asked "OK, that is like a taxonomy, right". "Err, yes, sort of". I realized that I was fuzzy on the difference and that was embarrassing. But from a bit of Googling later, it turns out that other people get confused on this (as this article relates).
But we got past that. I showed him some ontology examples from David's book. "Oh, we have that kind of thing. Probably not in that format, but my team knows this in our domain. How do we convert that domain knowledge into the right format?".
How Do We Answer These Questions?
Too many discussions between Semantic Web folks and business folk seem to suffer from the problem of the person asking for directions who is told "I would not start from here".
But business folk have to start from here. To get action, to get a budget and management attention, there has to be clear path from here to there. We have done a grand job explaining how wonderful the "there" is. It looks a lot better than "here". But we need to do a lot better on the directions.
What training courses would that person have to go on? What technologies, products and services would that person use? This is where the mainstream semantic web jobs will be found.