Venturing into Semantic Technologies ?

One of the most interesting sessions at the consistently engaging Semantic Technology conference in San Jose back in May was the 'Semantic Venture Panel: Investor Opportunities and Pitfalls.'

Alex Iskold did a good job of capturing the essence of the session for ReadWriteWeb and I recommend a read of his post, which freed me from having to write up the detail of this excellent discussion. We also talked about this event (and Linked Data Planet) in June's episode of the Semantic Web Gang, and Cambridge University's Nico Adams did a sterling job of maintaining a video diary throughout the event (although unfortunately he doesn't appear to have been in the venture capital session).

As Alex notes, the panel comprised Vulcan Capital's Steve Hall, Intel Capital's Eghosa Omoigui and Palomar Ventures' Amanda Reed. Each brought a different perspective to the conversation, determined by their own experiences and the type and stage of investment most commonly made by their fund.

So why return to a conference held back in May now, at the end of July? A piece by Stacey Higginbotham, syndicated to BusinessWeek from GigaOm this week, reminded me of comments made during the panel, and raised some questions that I thought were interesting enough to share. In VCs Have Their Heads in the Clouds, Higginbotham asserts that;

"the game-changing potential of [Cloud Computing] services has venture firms sitting up and taking notice. Indeed, after spending the past few years pouring money into Facebook applications and me-too social networks, venture firms are starting to invest in infrastructure again, with both hardware and software plays tied to the cloud."

Whilst recognising (as Talis CEO, Dave Errington, was quick to remind me when discussing this internally) that there is not one received wisdom in the venture market, there remains a clear disconnect between the thrust of Higginbotham's piece and some of the panel's messages.

Responding to my question about finding the interesting market opportunities for deploying semantic technologies, for example, Palomar's Amanda Reed suggested (and I'm paraphrasing, as I try to decipher my handwriting) that;

"lots of companies are filling a space just now that's essentially about routing around the hole left by the Semantic Web not being here yet."

And when Alex Iskold (doubtless wearing his AdaptiveBlue hat) asked about the amount of capital required by consumer-facing applications, Amanda was quick to identify the role played by Cloud services such as Amazon's in lowering those initial costs. Eghosa followed this, illustrating the dramatic cost savings to be found by outsourcing from North America to the talent pools of Eastern Europe (where costs can apparently be 10-15% of those in the Valley).

Each of the panellists, in different ways, appeared to suggest that they were not keen to invest in underlying infrastructure; the 'context' to most organisations' 'core,' to paraphrase Geoffrey Moore. Yet they recognised the increased cost of financing companies that had to reinvent wheels and 'route around' infrastructural gaps, and welcomed the cost savings of utilising existing Cloud infrastructure built at others' expense.

There is clearly a disconnect here, despite the points raised in the BusinessWeek piece. I am not, for a moment, suggesting that these investors are 'wrong'. Rather, I'm wondering if the system itself makes it difficult - or even impossible - to finance underlying infrastructure of this sort by traditional venture-backed means?

So many of the previous utility infrastructures were begun with public funds and Government mandate. Even when telecoms, railways and the like were privatised, favourable contracts, first mover advantage, and inertia gave them ready access to steady funds and existing customers to fund programmes of modernisation and reinvention.

As we move beyond the previous infrastructure of fibre, routers, switches and servers to a new infrastructure of facilitating compute utilities, how do we fund these... and how do we secure sufficiently large markets to make the initial investment defensible whilst ensuring the survival of competition and choice?