Ovum: Tie and Mighty
That bow-tie might be eye-catching, but don?t let it detract from the important message in Ovum?s report on corporate financial systems, says Guy Clapperton
Ian Wesley, managing consultant at Ovum, has a worried air about him. The higher-profile, co-writer of the research company?s forthcoming report on corporate financial systems has been popping into the office frequently to actually write the thing, but the staff have not been greeting him with the warm hello?s that might be expected. The reason is simple. They have not recognised him.
?He wasn?t wearing his bow-tie,? says a perturbed Wesley. In fact Dennis Keeling, one of the best-known dandies in the business, has been seen sloping in wearing distinctly casual apparel. ?It?s like he?s in mufti or something,? adds Wesley.
Keeling?s bow-tie is indeed a legend in the accounting software industry. He may have brought the Bank of England around to his point of view
on the importance of European monetary union to the accounting industry, and seen things move in Threadneedle Street as a result. He might indeed have founded the much-needed Business and Software Developers? Association (BASDA) and through its offices forced numerous overseas software suppliers to get their act into gear as far as VAT implementation goes; and he might well have persuaded Customs & Excise to co-operate with, rather than dictate to, those same software providers when it comes to VAT and Intrastat implementation.
In fact a great many businesses that have never heard of him probably find their activities a great deal easier to manage because of his efforts. But he remains best-known for the cloth draped about his windpipe, and it is with some relief that the small throng gathered for the launch of his latest oeuvre notes that it is duly in place when he enters the room.
The document itself is arguably just another IT report. Resellers get to see squillions of them every day, with each one angling for a look-in during a snatched coffee break. But there are strong messages for the IT community this time around. Client/server has failed to deliver on its promise. Open systems are not as open as had been previously thought. The vast majority of implementation is being done through third parties such as systems integrators and resellers, but this has been bad news for the consultant. Windows NT is growing, but so are the Windows support issues, and the fastest-growing areas of all are ? wait for it ? mainframe and AS/400. That?s right, those Jurassic blokes.
Then there are the issues driving the businesses forward. First there is the Year 2000; people are getting jaded with this at exactly the wrong time. Then there is European monetary union. Think you can treat the euro as just another currency when it happens? Dream on, it?ll work entirely differently. And there are the wishes of the international clients. They want to have their information easily accessible through the same system worldwide and it isn?t happening. Not only is it not happening, there are few ways of making it happen given the current shape of the systems involved.
What stands out above all else is the resurgence of the market for resellers in a market that looked as though it was about to go exclusively direct or through consultants at best. Keeling, with his consultant?s rather than his analyst?s hat on, is not altogether in favour of this. ?We?ve tried using parties and I believe it?s going completely the wrong way. I like to deal direct with the developer so I?ve only got one neck to wring.? It?s idiosyncratic, but it?s a point of view.
More obviously contentious is the idea that client/server is basically a dead duck for the moment, with the largest reference site made available to Keeling consisting of only 1,000 users. He believes that despite the noise made about client/server a couple of years ago, the boom has not happened and the network PC and the intranet are likely to supersede its growth before it actually happens.
However, the client/server market has its strong adherents in spite of what Keeling says. David Turner, marketing manager of Coda, which has seven per cent of the UK user base market for accounting systems (Tate Bramald Consultancy figures, 1996), has seen strong growth. ?It grew at around 100 per cent the year before last and about 33 per cent last year, and it?s looking like around 50 per cent this year,? he says.
The mainframe competition is seeing new business not because of any intrinsic growth in the field but because of the year 2000, he suggests. ?We haven?t even done our port to DB2. IBM is always on at us to do it but we made a commercial decision because there was a lot of growth in client/server.? Having said which, a great many companies (including Oracle) which only a while ago were committing heavily to the platform are now hedging.
Open systems is another example of technology that Keeling believes has yet to make the grade. JBA?s experience illustrates this usefully. A representative confirmed that although the company had released Unix versions of its suite recently, its main market remains and will continue to remain in the AS/400 environment. Jyoti Banerjee, managing director of Tate Bramald Consultancy, suggests there is a very profound reason for this ? the stuff works.
IBM even issued a press release about how many users had sent the company a Valentine?s day card because they were so enamoured of their AS/400 this year. The words ?gits? and ?sad? spring all too easily to mind, but there must be a germ of reason in it somewhere. Equally sad, at least for its advocates, is Keeling?s conclusion that open systems have a ceiling of 1,000 users, beyond which they are pretty useless in financial IT. Banerjee believes there are reasons outside the technology to explain this. ?We did a study of IS managers and 55 per cent of the 800 companies involved said they didn?t have the right people to manage open systems. That?s pretty duff.?
In spite of all Keeling?s other findings, though, there were a few distinct strands of wish lists that entrepreneurs and business owners had for their software when I did a round robin call. The replies received fell into two broad categories: the ?we want to be certain of surviving the century changeover? brigade and the ?look, EMU is coming very soon and we don?t know what to do about it? group. This is the point at which a journalist sits back and says ?those old chestnuts again, let?s write about something else?, whereas for the poor business owner who has to run a company these are probably the most important issues around at the moment.
First, the good news. Every package Keeling looked at for the report was fully year 2000 compatible. Corporates will have to look at their heritage systems, and resellers with skilled Cobol programmers on their books will be able to hike their rates before the end of the century, but as far as the current financial systems go you?re OK. EMU is more of a problem because it is misunderstood.
Last time PC Dealer wrote about the EMU issue, one of the major stumbling blocks was that the rounding issue had yet to be sorted out. Put simply, if one country is rounding its conversions down to three significant figures while another is rounding them to five significant figures, the conversions won?t match and the figures will be skewed. The Dublin summit conference decided to round to six significant figures throughout Europe, killing the problem stone dead ? yes, we have a ?politicians getting something right? situation, which made the bow-tie wobble visibly.
Less easy to resolve is the mechanism by which European currencies other than the euro will be converted to currencies outside EMU during the interim period in which the new coinage is introduced. As of 1999, dollars (for example) will be converted into euros, not francs, deutschmarks or anything else. The euros will then be converted into single-country currencies. There is not a single corporate accounts package that can handle this as yet, and there is not a single international company that won?t have to deal with it.
The 1999 date is in dispute, with feelings in Germany, for instance, running high against integrating the currencies. But to hell with democracy ? as Keeling points out: ?That?s public opinion, but when you talk to the banks they regard it as a foregone conclusion.?
All the vendors are working on the problem. None of them feel they can wait for the politicians to make their minds up. Unfortunately, none of them have any choice.
This is compounded by the fact that heritage systems don?t really talk to each other. Ovum?s international respondents had what should have been a simple request: they wanted to be able to extract data on their financials from anywhere in the world and present it on a monitor easily. This is what computing is all about ? straight number-crunching and bringing the data together so that it makes something more productive than there was before. Regrettably it isn?t happening; platforms and environments have grown organically and through acquisition in almost every case and made life most difficult for anyone wanting to build anything coherent in a great hurry.
Keeling?s position as an analyst is to note the trend rather than suggest a solution; mind you, any resellers who want a piece of the profit available for sorting all that information out and who haven?t read PC Dealer?s recent masterpiece on data warehousing have only themselves to blame. Banerjee concurs with Keeling again: ?When you were in the proprietary world you could have had any front end, any hardware, any middleware and any email system. Under open systems it?s not possible to have tested your system against absolutely any combination because there are too many.?
The report?s findings are surprising in many ways. Open systems looking more dated than mainframe, client/server losing share to AS/400 ? it would all have looked so unlikely only a couple of years ago. But the market intelligence from many people, including his competitors, bears him out, with the only obvious dissenters having vested interests in him being wrong. The market can turn around, but for the moment the document ? and the Tate Bramald reports ? form an invaluable guide to how the lansdcape does not have to look as you might have expected.
In fact he looked so satisfied with his work that I?d have sworn I saw him visibly relax to the extent of loosening his tie on his way out ? but then again, maybe I imagined it. Never mind client/server, mainframe et al; that would have been extraordinary.