Getting inside your customer's mind

Knowing what your customer wants before they ask for it is a skill all businesses want. That's why business intelligence tools...

Knowing what your customer wants before they ask for it is a skill all businesses want. That's why business intelligence tools are a growing market and Microsoft, IBM and Oracle are trying to dominate it. Murdoch MacTaggart reports

Business intelligence tools, even allowing for the specialised nature of current products, are not new. Indeed Oracle's venerable online analytical processing (Olap) product, Express, first appeared in 1970. Businesses are well aware that unless they have an accurate grasp of fundamental business data they're unlikely to survive.

The more astute firms have long understood that competitive advantage can be gained by analysing relationships between disparate sets of data, by mapping product sales to particular geographic regions or times of day, by experimenting with different store layouts and by evaluating the impact on turnover or profitability. The much-cited example of

placing six packs of beer close to packs of nappies could only have been found by a rare, inspired guess or, much more consistently for similar encouragement of impulse buying, by intelligently analysing sales data across a number of different dimensions.

This market has typically been dominated by expensive products running on powerful hardware and requiring complex and expensive consultancy to set up. As the number of dimensions to be evaluated increases, the processing power and temporary storage needed will typically increase exponentially. The impact on existing systems of running complex queries against live data can often be so considerable that the common practice has been to build data warehouses, specialised collections of data abstracted regularly from live operational data, cleaned of inconsistencies and put into a standard form. In other cases, specialised tools may be used to run queries against operational data or to interrogate localised subsets for immediate managerial information.

The growth of Internet trading has brought with it new demands for up-to-date analysis of data. Gigabytes of raw data, derived from users' interactions with Web sites are available daily and could provide valuable insights and necessary information when processed.

What is also new is Microsoft's entry into this arena. Microsoft is traditionally at the low end of the scale in enterprise terms and although this is true to some degree with its entry into the business intelligence marketplace, what is particularly interesting is the extent of its impact in a very short time and its potential for shaking up this marketplace. According to the authors of the Olap Report, the new "Olap triad" will be Oracle, IBM and Microsoft - a statement which should be considered against the current ranking of the present generation of Olap suppliers (see table).

MicroStrategy, currently number four in this table and showing a steady revenue growth, has reacted by moving towards the high end and by welcoming the impetus it sees Microsoft giving to the market overall. Chris Ward, business information marketing manager for Oracle says that "Microsoft is a threat, but not a serious enterprise threat".

Comshare, another major and long-established business information company and one whose products are highly regarded in terms of usability and sophisticated analysis capabilities, has reacted by embracing the Web interface approach wholeheartedly and by continuing its strategy of maintaining interoperability with relevant significant database managers such as Oracle8, IBM's DB2 and Microsoft SQL Server.

The SAS Institute is also moving more towards the high end by ensuring that a range of operating platforms and databases are supported.

Business Objects claims that the main effect of Microsoft's entry into this area will be felt by those suppliers which are largely reliant on revenues from Olap servers while others, such as themselves, which offer wide connectivity and have specialist front-end analytical and reporting tools will gain rather than lose.

The specialist consultancy company Druid contrasts Oracle and Microsoft as companies providing full solutions for different ends of the market with Cognos and Business Objects as the two main players supplying front-end tools which can work with servers from many different suppliers.

Microsoft's impact

According to the Butler Group, business information should not be the preserve of a favoured few power users or key decision-makers but should be available right through a company. Only by taking this grass-roots approach, it says, can a company operate as intelligently as it should. Furthermore, unless firms learn to do this, they risk being wiped out in the new, dangerous model brought about by global trading.

The problem in the past has been both cost and difficulty. Large software suppliers selling into the high-level enterprise space typically do not grasp the budget and time constraints which affect smaller businesses. Oracle, for example, says it has reduced prices significantly in many areas in order to target small companies. Yet one example given - its Fast Forward Budgeting package which bundles together (Olap) Express Server, Financial Analyser, Oracle General Ledger, Oracle Support and Oracle Training in a total package which the company guarantees to install and configure, so providing a risk-free contract - still costs close to £60,000 for a 10-user version. Contrast this with the raw price for a five-named user copy of the standard version SQL Server 7 (albeit without general ledger, training, consultancy or support) of under £1,000.

In the case of Olap Services, Microsoft bought the technology but not the company. Panorama, the Israeli-based developer, continues selling its product in specialised areas while Microsoft has incorporated it into its SQL Server 7. Although the product comes as part of the SQL Server 7 suite, it can be installed on its own and need not run in conjunction with the RDBMS. Note that there are two editions of SQL Server 7 and that it is only the enterprise edition which includes the partitioning capability within the Olap Services module.

Although the product was launched only at the beginning of 1999, there appears to have been a huge take-up. Nigel Pendse of the Olap report calculates that around half a million seats had been sold during 1999, that makes Microsoft's market share, in terms of numbers rather than value, greater than that of all the other Olap suppliers combined. Licences sold for SQL Server 7 are not the same as working installations of Microsoft's Olap Services, of course. Nevertheless, the deployment in such a short period of time is impressive.

Opinions vary as to the quality of the Microsoft's Olap Services but most objective commentators suggest the product is impressive but somewhat let down by the available clients. The main client offered by Microsoft is Excel's PivotTables facility - but the capabilities here fall a long way short of being able to take advantage of what the Olap Services can offer. However, a number of third party clients are available, most of them going well beyond Excel's basic level of functionality in this area. These include Nova View from Cognos, [email protected], ProClarity and Seagate's Worksheet and Analyzer products.

The Olap Services themselves seem extremely fast in building cubes and in aggregating data, while the size of database which can be handled is impressively large. Pendse takes the view that while Olap Services is less functional than Oracle's Express, it has better scalability.

Although there remains a question mark about the viability of Microsoft products in the enterprise computing marketplace the favourable evaluations of Windows 2000, in terms of stability, scalability and resilience, are starting to dispel these.

Olap Services in SQL Server 2000, due for release in the second quarter of this year, will be extended and include data mining capabilities. At the lower end, where the client functionality provided by Excel may be adequate, the opening up of this market to smaller firms is extremely welcome. Bigger firms may need more functionality than Microsoft's Olap Services can currently provide, while others will find it adequate for much of what they do and at a significantly reduced price compared with competitive products. If nothing else, this is likely to have a significant effect on overall pricing.

0Olap: the major players

Attempting to estimate the world market for business information is a hopeless task. In part this is because of the wide variations in definition claimed by different suppliers, in part on account of the bundling of business information tools with other products, or the multi-functional nature of many products.

The highly regarded Olap Report does, however, regularly attempt this in the specialised area of Olap products. Despite its long involvement with, and experience of, this market, it still has to estimate extensively. Nevertheless it has estimated the total worldwide Olap market, including implementation services, at around $1bn in 1996, $1.4bn in 1997, just over $2bn in 1998, $2.5bn in 1999 with an expectation of very close to $3bn for this year and rising to $4bn for 2002.

It ranks the major players during 1999, together with their market shares, as follows:

1 Hyperion Solutions (including all resellers) 27.8%
2 Oracle 11.2%
3 Cognos 10.9%
4 MicroStrategy 10.2%
5 Microsoft 7.5%
6 Business Objects 5.2%
7 Comshare 3.1%
8 Applix 3.0%
9 IBM 2.9%
10 Sterling Software 2.7%

Read more on Business intelligence software