Why inter-operability standards are essential for an open and competitive market

| 1 Comment
| More

Bryan Glick's summary of the "Big IT v. SME's" debate and the need to change supplier behaviour raises many questions, not least "how we bring about genuine competition?".

A little while ago, when blogging on the need for robust policies to preserve competition in the on-line world as a whole  I promised to reprise my script to a recent BASDA (the Business Software Developers Association) conference. I was asked to address the importance of inter-operability standards. These are boring but essential to genunine competition. Without  effective action on standards, the lobbyists of the oligopolists can still make a credible case to "Sir Humphrey" that the Minister will be happier with hiring consultants to plan a high cost/risk "delayed big bang" project (i.e. promises today, problems tomorrow: for his successor) than a low cost/risk incremental change programme. The former is safer for the minister - he will have moved on before the chickens come home to roast. The latter opens up the potential for criticism while the minister is still in office, whether the trials work (post code lottery because only a few have benefited) or not (waste of time/money, however small).

Chi Onwurah is correct in saying that we still need the big suppliers. If government is seeking to contract 25% of its business to SMEs that mean the big suppliers will still account for up to 75%, But securing value for money, with systems and contracts that can evolve over time, as needs and technologies changes, require fundamental changes in behaviour on the part of both government and its suppliers, small as well as large. The era of painfully negotiated, comprehensive and inflexible outsourcing contracts and PFI deals is coming to an end. Its demise will, however, be neither easy nor painless, unless and until major suppliers can find democratically accountable, (for public money and quality of service), ways of working flexibly and profitably with "families" of nimble, low overhead, innovators. Some suppliers are well down this route. But even they face problems because they risk cannibalising bread and butter revenue streams in favour of a reduced share of lower margin new business. The true winners are those (as yet only a handful) who have worked out how to use the opportunity to jobs back to the UK while dramatically cutting their off-shore, outsourced costs: the "win, win, win" strategy. 

I was, however, accused of "scaremongering" when I blogged on the possible implications of a recent criticism of a ministerial refusal to plough good money after bad until the end-user trials of the people processes the technology was to support had demonstrated success. I should therefore make a couple of disclaimers before I reprise the script I used when speaking to BASDA last month.

First as an occasional journalist and regular blogger:

I first wrote for Computer Weekly in 1973, when part of my London Business School Master's project appeared as a ten part series on "Why Computer Systems Fail" (£15 per thousand words was good drinking money in those days). I have been an occasional contributor ever since. I have also written for others. I even had a column in Computing for some years. I started this blog in September 2007 when I was "convenor" (alias programme advisor) for the CW500. The aim was (and still is) to put political matters into IT context and IT matters into political context, mainly for the benefit of Heads of IT (whatever the current titIe of the poor sod who carries the can for delivering systems that work), for users (alias victims) and for investors (from finance directors to fund-managers).

Apart from occasional speaking engagements and advice on thought leadership opportunities I have not worked for a supplier since I left ICL in 1977. I then had five years outside IT as a corporaate planner for a UK-based multi-national before joining the NCC, originally to set up a technology assessment operation. When I left the NCC in 1986, I took with me the operation that I had joined the NCC to create: helping banks, fund managers and major users to appraise new technologies and the associated investment opportunities. That has entailed both avoiding vested interests and taking a cool look at innovations and market enthusiasms. 

Second as a volunteer, unpaid, politician:

In 1978 I was co-founder of the Conservative Computer Forum. About the same time, I volunteered as an ASTMS representative (I paid the political levy and remain a member of Unite) on the TUC studies for the Labour Party on the impact of new technology. Some of the studies on which I worked appeared in the policy papers of both sides in 1979: e.g. telecoms liberalisation and the micros in schools programme. Others did not: e.g. telecoms privatisation and IT Year. 

 In 1981 I was one of the co-founders of the all-party Parliamentary IT Committee (PITCOM). Shortly afterwards I stood down as chairman of the Conservative Computer Forum (after joining the National Computing Centre I was barred from appearing on party political platforms). In 1993 I agreed to organise the re-launch of EURIM. Until 2010 (when I stood down as Secretary General of EURIM) I devoted my energies to working on an all-party basis.

In 2010 I agreed to do a three year term as chairman of the Conservative Technology Forum with a remit to try get the younger generation to do as we had done in 1978-9: collating industry inputs into peer-reviewed recommendations for submission to those responsible for drafting party policy. I made clear that I still regarded most matters IT and tele-communications as cross-party rather than partisan and would continue to seek support for all-party, pan industry studies where practical.

I stood down as chairman of the CTF on March 31st but remain Vice-chairman (Policy Studies). The list of topics for which CTF is currently seeking volunteers and submission is on their website . The nearest Labour equivalents are probably Labour Digital   and the Digital Government programme although Labour has also announced studies into Digital Skills and on the Digital Creative Industries. When speaking to industry audiences, including via this blog, I strongly encourage listeners and readers to be active via the party of their choice - because the silent majority gets what is deserves - ignored.
Now to the meat of this blog: the script I used when speaking to BASDA on why action on inter-operability standards is as important as action on public sector skills if we really do wish to change the behaviour of government and its suppliers towards using IT to serve the community. This is relevant to the issues raised by Bryan because without such action, we risk perpetuating practices condemned as unfit for purpose by the National Audit Office, the Public Accounts Committee and the Public Administration Select Committee     

Why inter-operability standards are essential for an open and competitive markets
•    The on-line world is dominated by  an interlocking hierarchy of cartels
•    The Western economies are in recession 
•    Therefore monopolists can only  grow by vertical integration
•    Politicians and regulators are in ignorance, captivity or both
•    Competition cannot survive without enforced open standards
•    The layers of standards: technical, operational, service and business 
•    Who is the opposition: customers and competitors?
•     Who are your potential allies: customers and competitors?
•    You have to work to make standards less boring for politicians

Good afternoon.  Standards are boring. I have been involved in the politics of IT for over forty years and have met only three politicians who can stay awake, let alone understand what is at stake. The first was the 3rd Viscount Chelmsford, who helped expedite the chip and pin card standards which now save us £billions by causing criminals to focus on American issued cards instead. The second was Malcolm Harbour MEP who saw e-Procurement and e-Invoicing thorough the European parliament. The third is Stephen McPartland who is trying to get the UK to implement what has already been agreed [insert link to report] because the slow take-up by the UK public sector is still costing us £billions year in fraud and inefficiency.

[In retrospect I should have said half a dozen and included Andrew Miller, Chi Onwurah and Stephen Timms] 

Standards may be boring but they are important and those who do not participate in their formation get stitched up because the on-line world is dominated by monopolists running cartels which use standards games to lock in their customers and rig markets to their own advantage.

The on-line world is dominated by an interlocking hierarchy of cartels
•     Chipsets: ARM, IBM, Intel and who?
•    Global access: Google and Microsoft
•    UK access:  BT, Arqiva, Virgin, Vodafone/O2, EE  and resellers
•    Global switching : Alcatel-Lucent,  CISCO and Huawei
•    On-line Advertising:  Google and its acolytes
•    On-line  Retail:  Amazon, eBay and who?
•    Identity: Experian, Lexis Nexis,  Equifax , Call Credit   
•    Content: News International, TimeWarner, Netflix
•    Storage: EMC, Hadoop. Oracle, SAP   
•    Cloud: Amazon, Google, HP and IBM

The Internet has been described as a cartel masquerading as anarchy.  The basic concepts may be simple and imply freedom and lack of control but the reality is that it is now run by American Patent and Copyright lawyers and lobbyists for their clients. To take a simple example. 10% of the world population now transact over mobiles. That nearly always means an Arm or Intel chipset set in a phone that uses Android or Windows operating systems to find sites via Google. In the UK your mobile signal will reach the peering centre via one of half a dozen national mobile and wifi services sharing  Arqiva Masts and BT routers and exchanges.

•    The Western economies are in recession 
•    Therefore monopolists can only  grow by vertical integration

But the Western World has not recovered from a banking crisis caused by a government created lending bubble in parallel with the dotcom boom. The monopolists are therefore trying to eat each others' lunch. BT is moving into content to attack Sky and squeeze Talk Talk. They are retaliating by moving into infrastructure in co-operation with City Fibre.  Amazon's resale agency operations are attacking e-Bay. e-Bay is therefore helping traditional retailers fight back against Amazon. The movie studios are backing Netflix to take out the pirates on one side and the US cable operators on the other. Google is taking on almost everyone.

How long before "G**gle-pay" makes an update on your financial status available to its customers before the payment reaches your bank?

•    Politicians and regulators are in ignorance, captivity or both
•    Competition cannot survive without enforced open standards

Central governments and regulators are bewitched and enslaved by the lobbyists and lawyers employed by major players. Google is now the second highest spender in Washington, behind GE but ahead of Apple, AT&T and everyone else. In London it is said to spend more than the rest of the ICT industry added together, save for BT which may get through even more.  More-over many others are in their thrall because almost all on-line promotion depends on positioning in Google searches.

The layers of standards: technical, Structural, service and business 
Component    Data definitions, Chipsets, software modules
Architecture   "Rules" for interconnectivity of devices, modules etc
Performance  Capability and performance of products and services
Process        Business models, service level agreements, liabilities

There are many layer models but this slide may help make sense of some of the games being played.  Almost all modern systems involve hierarchies of components from multiple sources.  X805 is the ITU framework standard for complex networked systems. It tracks dependencies, which versions of which components and processes are used, to which standards they were audited, by whom and when. Its use depends access to libraries of the relevant audit results. Meanwhile major players talk open standards, in order to invade new markets and then seek to lock in customers via products and services that are vertically integrated using patented or customised sub-features.  Last week I was given the example of a telecoms supplier who had to take on extra staff because BT uses a non-standard on-line invoicing and payment system, and he has no wish to be locked out from serving other customers.

The critical importance of Data Definitions as a foundation of effective inter-operability has not been over-taken by the wonders of fuzzy big data. Forty years ago, as a very young senior programmer, I was loaned to the auditors to help them check the value of our work in progress. One of my tasks was to help them merge part number master files with different prices for supposedly identical electrical components. It was only later than I learned that some of the biggest differences were to do with testing for mean time between failure. A well known German carmaker lost its reputation for reliability when it used components which began to fail after barely one year instead of ten to fifteen.

We have similar confusions over standards for certificating identities and contracts.  Some services accept no liability for errors, like government departments claiming Crown Immunity. Many US based services limit their liability for $500 under the laws of the Commonwealth of Delaware. Notaries take liability for up to a £million and Scriveners, the hidden monopoly at the heart of world trade, accept unlimited liability - but exercise a rigor that is alien to most of the on-line world .

Why is progress so difficult given that HMG is supposedly in support?
•    Who is the opposition: customers and competitors?
•    Who are your potential allies: customers and competitors?
•    Those who have nothing to gain from procrastination but pain
•    You have to work to make standards less boring for politicians

Until recently UK central government commonly spent 30% more than Germany or Holland and double local government for equivalent IT services. It then tried to get the oligopoly suppliers to offer cuts of at least 10%. It could not get them to do so, except in return for extended contracts. It has therefore blocked new spend other than via frameworks, such as G-Cloud and PSN, which supposedly embed open standards. It is also trying to enforce spending cuts on Local Government of around over 30% and on some central government departments of 50%. These can only be met by cutting and merging services.    

HMG also aims to spend 25% of its spend via SMEs, but has yet to develop procurement routines that will enable these to lead innovative bids with support from large players. But one of the advantages of inter-operability should be to reduce the risk of failure because the product or process can be supported or replaced by another supplier, albeit not necessarily at the same cost. In the private sector this can be seamless with processes, as well as components, dual sourced. 

So what is not to like?

The first problem is that those responsible for policy and implementation are used to big consultancies earning big fees for organising big procurements. They do not have the mindsets, let alone skills, to plan incremental change, with small procurements staged over time. These are also alien to most major consultancies.

One consequence is that little business is yet flowing through the G-Cloud and PSN frameworks and the silos of state and their strategic partners have been spinning out the contracts of yester-year, hoping to ride out the drive for reform. 

They will fail because the Treasury has run out of money. In consequence some large suppliers beginning to look at how to work with SMEs to bid for Local Government business, where the markets are more open. They are not finding it easy because those in key roles find it much easier to justify prolonging existing contracts or to use subtle variations, such as non-standard e-procurement and e-invoicing to lock SMEs into their supply chains as opposed to bidding in their own right or, worse, via their competitors. 

Also we should not forget that one of the reasons for opposing change to common processes, particularly e-invoicing linked  robust, well-controlled  procurement and payment processes is to protect or prolong opportunities for fraud.    

I therefore recommend you submit evidence to the Labour Party Digital Government policy studies in parallel with inputs to Cabinet Office consultations, Select Committee Enquires and relevant Conservative Party policy studies. I also recommend very public co-operation via the all-party Digital Policy Alliance. We need to ensure that the opponents of reform have nothing gain to gain but pain by attempting to procrastinate until after the 2015 election.

I have a morass of material from reviews of DWP, HMRC, Cabinet Office and EU proposals for Identities, Real Time Information and so on. Most indicates the value of twin track strategies for tax and benefits: separating the routines for the 80 - 90%, whose fortunes are reasonably predictable, from the processes for helping the 10 - 20% who are in most need as they move drift in and out of work (usually with the small firms who are making such a predictable hash of RTI) or otherwise lurch from crisis to crisis. I also have much material on procurement and invoicing which suggests we should facilitate choice between competing frameworks - so that customers can split procurements according to which provide them with attractive choices/prices and suppliers can support those which reach attractive customers at affordable cost.

Once again effective inter-operability standards are central to success - but first we have to put the benefits into political context - so that they cease to be boring.

1 Comment

Real information standards are about how the business interprets data, not to be confused with data standards, which are about how computers represent data. Given this distinction, there are two important reasons for adopting information standards.

The first is risk, in particular the risk that when you send information from one system to another, it will be misinterpretted by the users at the other end - an example from twenty years ago was the differce in meaning between "confidential" in UK and US government - while both used it in the sense of "be careful who you tell, but it's not as secret as secret", the governments had different business processes when it came to periodic downgrading of classification. The cost of aligning business processes across all aspects of govenment would be horrendous, but one needs at least to align the usage where the probability and consequence of misinterpretation poses a substantial risk. This is why, for example, payment mechanisms need to follow standards.

The second is most interoperability standards are about sending information from one place to another, but the same problems apply when sending data across time - that is, when reading data that is more than a few years old. The lack of a data standard results in being locked into a proprietary product - and worse if the vendor ceases trading. The lack of an information standard interpreting the data standard then makes it impossible to be sure that any replacement software, although it reads the data correctly, is then making the correct interpretation of the data. This is particularly troublesome where complex models are used, such as in Computer Aided Design (See Alex Ball's paper from the Digital Preservation Coalition).

These may not be major problems yet in an organization that still thinks in paper-based systems, however, with the launch of the Queen Elizabeth being vaunted as a capability for the next fifty years, such information standards will be the only way to ensure that the design and therefore the safety information can be kept up-to-date thoughout its life. Fortunately, the companies involved have been investing in developing the necessary information standards for the last thirty years, and lobbying the software suppliers to adopt them. (Several years ago the investement was estimated at over $500 million.)

Product-based industries have been adopting model-based information standards for the last twenty or so years, to the extent that some groups are describing the model-based enterprise. The reasons for adopting this approach are compelling, including direct reduction of cost though eliminating the Machine-Man-Man-Machine interfaces (print it out and type it in), enormous improvements in quality and in time to market. But this approach is completely dependent on information standards to build the supply chains needed and to sustain the products over their operating life. These standards will become more important as "The Internet of Things" takes hold - not necessarily for the individual consumer, but for a large organization which, say, needs to maintain a diverse fleet of vehicles and doesn't want a separate IT system for every manufacturer.

However, the important lesson for government was that these product standards were developed by the end-users, because they are the ones that own and understand the business processes. Such information standards are counter-intuative for data processing companies, whose objective is to develop a single software system that can be sold to many customers.

The political problem is, as noted above, investiment in standards takes time to pay off, and it's any additional spending in this parlaiment's budget will not show up as a saving until the next parliament.

Leave a comment

About this Entry

This page contains a single entry by Philip Virgo published on June 30, 2014 2:29 PM.

I apologise for sloppy blogging: I should have said oligopoly not cartel yesterday was the previous entry in this blog.

Are you suffering from Crapband in the City? Tell the Corporation. is the next entry in this blog.

Find recent content on the main index or look in the archives to find all content.


Recent Comments


-- Advertisement --