Feb

17

Index to Analytics

Filed Under B2B, Big Data, Blog, Financial services, Industry Analysis, internet, Reed Elsevier, Search, semantic web, STM, Thomson, Uncategorized, Workflow | Leave a Comment

A long time ago the Financial Times formed a joint company with the London Stock exchange to exploit the FTSE share price index. I seem to recall that this was not a success, but a colleague at that time joined the board and I recall asking him what an index was for. He replied that it was a sort of branding statement, and it also said that you had the underlying data from which to create the index, should anyone want to look at it. And was that a good business? Well, not really, since few people were able to make sense of the underlying data. So it was mostly a brand thing, then? Well, yes. And a brand thing where, since most people refer to the “footsie”, the brand reference is lost in speech.

I do not believe that they have the same view at FTSE now, and in a world currently rampant with indices it is interesting to check on the progress of players like Argus Media (www.argusmedia.com) who have used indexation powerfully to elevate a small player in energy and commodity data markets into a very powerful one. I wrote about this in November 2009 (https://www.davidworlock.com/2009/11/battle-of-the-indices/) and envisaged the war between Argus and McGraw Hill’s Platts in oil markets as a classic  David-Goliath story – but one which would need to be followed up by the victor to consolidate the gain with a wider service base. To quote: “index publishing is becoming an interesting value phenomenon.  It creates lock-in around which workflow activities and value-add analytics can be built. It gives brand focus and recognition.  It provides contract opportunities to supply and maintain service points on client intranets. In truth, it is sexier than it sounds.”

In light of this I was delighted to find that Argus Media had made an important purchase in analytics software this month.  Fundalytics “compiles, cleans and publishes fundamental data on European natural gas markets” and is a first service acquisition of this type that the company has made. Starting with natural gas, however, it should be possible to create a wider range of analytics activities, across energy markets, which are currently so very active, and other commodity areas like fertilizers where the company is building a stronghold. Competition is obviously fierce, with direct pressure from Platts, about double the size, and RBI’s smaller ICIS. And then there are the market information players who have always used the data and its primary analysis to form notation services for both players and investors – the Wood Mackenzies and IHS operations, and, at a further remove, Michael Liebrich’s New (now Bloomberg) Energy Finance and Thomson Reuters’ Point Carbon. It is understandable that there would be heavy competitive pressure in such an important field, and rewards will align with the industrial, financial and political clout the whole field invokes. But of the companies mentioned here, some are primary data producers, some secondary, and some create market commentary without owning a data farm at all. Can they all survive, and, if not, what sort of equipment do you need to succeed?

This is why the Argus Media purchase is more important than its size or value. If we have learnt anything from the consolidation of service markets in the network in the past decade then it is, surely, that relatively few players are needed to provide the whole range of internet services, and that users do not lust for more – indeed, they seem to want one sure place to go, and an alternative  in case their preferred supplier tries to abuse his pricing control. You could point to the history of Lexis and Westlaw in law markets for part of the history of this. Then they want from those two lead suppliers the ability to secure access to all the core data that they need: to both use that data and its analysis on the supplier service, and suck data into their own intranets to use in conjunction with their own content: to access APIs which allow them to create custom service environments and maintain them as fresh value add features are developed by the supplier: to use the supplier as the architect/engineer for workflow service environments, where news and update is cycled to the right place at the right time and where compliance with knowledge requirements can be monitored and audited: and, finally, they want the supplier to run the Big Data coverage for them, using his analytical framework as a way of searching wide tracts of publicly available data on the internet to secure connections and provide analysis which could never have existed before.

This is a formidable agenda, and I am not suggesting that anyone is close to realising it. Those who want to enter the race are probably now securing their content on an XML-based platform and beginning to buy into analytics software systems. And it was the latter point which so interested me in regard to Argus. If the human race descended from a tree shrew, then there is no reason at all why a smart data company close to London’s fashionable Shoreditch tech-zone, should not be a lead player in the future structure of service solutions for the energy and commodities markets!

Feb

12

The Point of Utility

Filed Under B2B, Big Data, Blog, Education, Financial services, healthcare, Industry Analysis, internet, Publishing, Search, semantic web, social media, STM, Uncategorized, Workflow | 3 Comments

We have had points of inflection and points of sustainability. Time then to propose a new “point”, one which applies universally throughout the world of information services and solutions, but which I found last week dramatically illustrated in the world of STM. Just as in the early years of the networked world we observed a point of disintermediation, at which the network effect removed real world service entities and made process cheaper and quicker, so we can now see places where re-intermediation is required, introducing a new service layer to streamline the network’s own inefficientcies, moving cost to a different place, but still often reducing it while increasing network efficientcies. And once this re-layering has taken place, the point of utility satisfied, and opportunity is created for major increases in the collection of neutral data about how the network processes in question work, and from this derives still greater efficiencies.

The ideal illustration of this is Mendeley (www.mendeley.com) and I listened with great appreciation last week when its co-founder, Jan Reichelt, described the problem which he and his colleagues had set out to solve. The inability to describe PDFs containing scholarly articles arose from the nature of network access and storage. Academic buyers knew what they had bought when they acquired it, but since the PDF envelope had no external description of what it contained, it was hard to cross search one’s collected downloads or manage a research project’s collection. And publishers, being publishers, adopted different practises regarding use of metadata, or even numbering systems. And as sources of grey literature, of files of evidence, or collections of abstracts became involved, local content access and control became a major overhead. Jan and his colleagues knew – only a few years ago they were researchers themselves.

So Mendeley was launched as an environment to regularize this, and to ensure that academics are able to better exploit the acquisitions that they have made. As a result those primary drivers of network efficientcy can be accomplished – saving money, making better decisions, and ensuring cost-effective compliance. This Point of Utility exploitation then has certain network knock-on effects. The service, like Mendeley, becomes an important part of the navigation of end users, and indeed may become part or the base for the user access dashboard. Once the Point of Utility has become an interface, then it is able to pick up all sorts of feedback data from the way end users act through the interface. This data about workflow will indicate usage and popularity, the common processes that users employ in discovery, the way in which resources in the system relate to each other, and the subjects that researchers really search (as distinct from the disciplines that journal editors think they subscribe to). Once this activity gets underway then the new interface owner can begin to suggest workflow improvements, and resell across the market the high value data which derives from the actual patterns of usage. There is a Point of Utlity in every network environment and Mendeley, through their knowledge of researcher proclivities, have camped on one of these exciting fault-lines in STM.

Some of these opportunities arise from publisher activity – lack of collaboration, lack of standardization, lack of knowledge about what happens to articles post-sales – and some from the same features in the user community. This is not a blame game. Mendeley has taken the initiative and must be welcomed as one of the foremost workflow players in the sector, especially since the launch of the Mendeley Institutional Edition last month, which takes the intermediary role into the the academic library, in conjunction with and powered by Swets, who have quickly grasped the point. This, as well as exporting the Mendeley API (http://www.mendeley.com/blog/design-research-tools/winners-of-the-first-binary-battle-apps-for-science-contest/), will turn fast growth into a torrent: Mendeley already have backlogs despite having captured 160 million references for 1.5 million users. Some publishers (Springers’s short article previews) clearly get it – others, as ever in this sector, plainly adopt a policy of “this too will pass”).

But of course it will not. Far from returning to normal , the science knowledge market is in a ferment. Visit Dryad (http://datadryad.org/) and go to Giga Science  (http://www.gigasciencejournal.com/) and observe the impact of Big Data on this sector. My friend Mark Ware, in an excellent note for Outsell (https://clients.outsellinc.com/insights/?p=11693), has given chapter, verse and analysis. Content control in the scientist’s workflow is becoming a multiple media nightmare. Does the metadata accurately describe the contents of video diaries and observations, or their audio equivalent? Can we source the data behind that report, and do our own analysis? How many unpublished, unreported studies have validated these results? What has been said and at which conferences about how far down the track this research team has gone? Where do we find the right mix of experience to staff this next enquiry? Regardless of its peer reviewed status, who actually used this work – and, if they did not use it, what did they rely upon instead? Mendeley is a promising beginning, but there is a long road ahead. Stefan Glanzer (Last.fm) and Alejandro Zubillaga (lately head of Warner Digital Music – where he must have seen parallel problems) put in the seedcorn and should be congratulated. They have a real start-up (my spirits rise when I visit an office that has its bike racks and its table football in the foyer!) with the wind behind it.

One last check on where that wind is blowing. Visit ResearchGate (www.researchgate.net) and look at the ways in which scientists are beginning to indulge in meaningful social networking. I have been told for 25 years that scientists and academics are too competitive to use social networking. Like much of the received wisdom of the pre-networked society, this is at best a half truth. The whole truth is that there are no longer simple generalizations that hold true about researcher behaviour. That is why they flock so quickly to the Point of Utility.

« go back — keep looking »