Reuters technical development: A story of enterprise and innovation
Monday 15 June 2015
As well as being arguably the best general and financial news service, Reuters is and was a wonderful and very valuable business with innovative and revolutionary products that changed the way markets worldwide operate.
In the last four decades of the 20th century especially, this was due to the company's in-house technical development. Standards for professional products were high in terms of availability, speed, accuracy and, latterly, openness. Technical boundaries were pushed to their limits to build unique global products. In the early days, we had to use in-house developed technology.
But later, often at the behest of the powerful technology management in client organisations, Reuters used off-the-shelf industry standard software so that clients could weave together our products with their technical environment. Too often, the trade-off for this customer-friendly strategy was product roll-out delays, unfriendly user interfaces and our inability to deliver product change to all clients quickly. Duplication between geographic Reuter areas and our many acquisitions, solving much the same problem many times, did not help.
Martin Davids, a pillar for more than 20 years of the technical development function at Reuters, has gathered as many contributors as possible to a chronology of this work before individual memories fade.
Here is part of the story:
What was the purpose of the Reuter product line?
Reuters created a globally available decision support product family for traders in financial markets. In terms of scale and geography and market range Reuters was mostly first to market and the leader. Starting with Stockmaster, which gave European traders real time price visibility of US markets which in turn gave them the confidence to deal in increasingly liquid markets, we moved to the more innovative challenge of creating a competitive market on screen (the Reuter Monitor) in foreign exchange - the biggest market in the world. Traders wanted to see "dealable" prices on the same screen, so we created the Reuter Dealing service which logically culminated in Reuters becoming the electronic broker showing banks the best price in the market for them at any instant. A long journey from Stockmaster. But our clients wanted to use this interconnected product line as part of their own systems and networks and alongside competitive offerings. Hence the slow creation of our open systems product set including the new idea of "logical data" which clients could use for their proprietary software applications - the datafeed, still the pinnacle of the Thomson Reuters financial product line.
We created a new world-wide business (now with many players) which specifically made financial markets more transparent around the world
What factors determined our technical development approach?
Geography. Reuters was divided into profit centres and the ambitious leaders of those profit centres wanted to deploy technical product development suitable for their part of the world. They also created long lasting duplication with consequent lack of the fundamental economy of scale we should have exploited.
Economics. Was broadcast over satellite and cable a feasible way of exploiting economies of scale and avoiding the high cost of terrestrial lines whose tariffs were still set by monopolistic communications providers? Or did the inherently two-way nature of much of the product line defeat this worthy aim? The argument raged across the Atlantic for 20 years.
Subsidiaries. We bought a lot of them mostly to attempt to improve rapidly the international product line. But a plethora of subsidiaries tended to solve the same technical problem or task over and over again and in turn frustrated attempts to roll out enhancements simultaneously to the whole world client population. The Bloomberg company had one centralised product line and therefore could achieve greater manoeuvrability as time went on.
Openness and industry standards. By making our products and data "open" we were on the whole much more useful to our clients. On the other hand introducing change in the field became more complex. Like software vendors we had to offer backward compatibility. By using industry standard software, notably Microsoft Windows which we pioneered in financial markets, we saddled ourselves with persistent user unfriendliness. Managers could not demonstrate our arcane product line. It was very different from today's browser world or indeed the "antique" simplicity of a green screen Bloomberg.
What went wrong?
Complexity and duplication coupled with the extraordinary difficulty of ever switching off any part of the product line for good meant that change delivered to all clients took forever and the cost of maintenance and upgrade was not helped by the tsunami of market data kept year in, year out.
User unfriendly interfaces courtesy of our worthy adherence to industry standard software (why build when you can buy?) had a profound negative effect vis a vis competition and our internal understanding of our own product line and how it was used.
Speed of data delivery, time to market of minor new features and screen appearance usually trumped deep understanding of data, especially historical data. Bloomberg's progressive lead was based on fixed income where data knowledge is compulsory.
What was outstanding and innovative?
We got mini-computers and then personal computers into dealing rooms.
We solved problems of escalating data volumes, 24-hour non-stop operation, geographical roll-out in the most testing circumstances mostly as trailblazers.
We obtained permission, notably in Europe and Asia, for third party communication (by our clients) over circuits from monopoly telecoms providers. This was truly a revolution enabling not only our own transaction services but also the entire value-added private network business.
Before the Internet, we deployed a packet switched network to permit point-to-point communication at a feasible price point.
We created a new world-wide business (now with many players) which specifically made financial markets more transparent around the world. This transparency made the markets more liquid and therefore more useful.
The chronology captures what happened and shows where we took some right as well as some wrong turnings. It includes the Reuter Monitor, Dealing, Rowgrabber, RICs, trading room systems, data feeds, Armstrong and Globex among many others.
The first part will be published here next week. I hope you enjoy reading it and if possible contributing to its evolution as a record of a great business achievement.
David Ure joined Reuters Economic Services in 1968 as a trainee. After time reporting the London commodities markets and working on the London and Brussels RES desks, he started to ascend the management ladder, spending the 1970s mostly in product development (mainly the Dealing system development), the 1980s in charge of Reuters in Europe, and the 1990s as an executive director back with product development and strategy. He left the company in 2004. ■