It's almost impossible to open a newspaper without reading news of some compliance or governance issue arising inside service based enterprises. These are no longer restricted to the headline grabbing stories of Enron or Parmalat but include a broad range of reported failings. In the year to 31st March, 2004, the FSA, Britain’s chief financial watchdog imposed a record £12.4 million in fines, an increase of 35 percent over the previous year. The figure is almost certain to be higher in the current year. More worrying, the FSA managed to bring its first criminal prosecution for market misconduct. Even worse, it slapped Abbey with a fine for ‘not knowing its customers.’ Bank of Scotland hit the regulatory banana skin for losing records. Elsewhere, Standard Life blamed poor publicity arising out of its spat with the FSA over the way it reports its numbers as a prime reason for a dramatic fall in business. So what’s going on?

Misconduct among financial institutions is nothing new but the weight of regulation is now embracing what might otherwise seem to be trivial offences. Yet in an age where information is everywhere, there should be no reason why companies should fall foul of regulation. Or is there? It’s not as though companies haven’t spent money on improving their systems. In the late 1990’s companies were falling over themselves to invest in ERP, CRM and other transaction related solutions. While much of the frenetic spending has gone away, AMR Research believes that executives will spend an additional $14 billion on enterprise applications over the next five years in a global market they estimate to be worth an eye watering $60 billion. The question today is whether companies are focusing their attention on the right kinds of application. AMR affirms that spending on software that tackles compliance related issues will experience higher than average growth. But other research indicates that companies don’t necessarily spend in the right ways.

Specifically, involving as many people as possible into decision making processes. According to Forrester Research: "One significant disappointment has been the failure to involve more than 5 percent of the enterprise in making data into intelligent and actionable information. The business intelligence ghetto — as some industry pundits have described it — represents the exclusive analytic realm where only the most highly trained data moguls ("power users" inside and outside IT) spend their most productive time." Most recently, Forrester adds that as many as 30 percent of companies are concerned about information quality lapses to the point where these may require boardroom intervention.

Many of the problems stem from the fact that information is everywhere but nowhere. In today’s high speed, data rich world, the best performing companies are those that have a firm grasp on their information assets. Sadly, that is a very small minority. The reasons for this parlous state of affairs are not difficult to understand. Applications have proliferated throughout organisations as they latched onto the idea that improving process speed would translate into competitive edge. That turned out to be a half truth. While many companies have seen significant transaction processing improvements, implementing systems designed to automate processes does not of itself improve overall operational efficiency.

If anything, proliferation has led to what Gartner refers to as ‘spaghetti IT’ where discrete applications play crucial roles but which are not integrated in any meaningful manner. Certainly the information they generate is not integrated or managed. Elsewhere, global enterprises have allowed their international subsidiaries and divisions to paddle their own information canoes, adding to the spaghetti IT effect. Systems are duplicated, processes are not standardised in a manageable way and it takes forever to get even the basic monthly reporting pack up to the boardroom.

Finally, there is the impact of the spreadsheet. Everyone likes a free lunch and this is the impression that is created when PC’s or laptops are installed. They usually come with a full suite of office productivity software, including the ubiquitous spreadsheet. From a financial management perspective, this has proved to be a disaster. According to longitudinal research undertaken by the University of Hawaii 20% to 40% of all spreadsheets contain errors and error rates as high as 95 per cent have been recorded.

The problem is that spreadsheets, which were originally designed as general, ad hoc analysis tools have become woven into the decision making fabric. They are an integral part of the reporting process, seen by some as sacrosanct. This is a fallacy. In many instances, spreadsheets represent the fragile last mile of business intelligence because they are born out of the premise that humans can operate as report servers. Even a cursory view demonstrates this is an untenable view.

There’s an old saying that goes: "If you give a set of numbers to six accountants then you’ll likely get six different sets of results." This is because numbers are capable of interpretation. The same goes for spreadsheet construction. Since the tools are generic, they have little if any functionality that conforms to generally accepted accounting principles. This allows an infinite variety of construction, much of which is not documented and which allows finance people to be as creative as they wish. As spreadsheets grow, become more sophisticated, get added to and consolidated, they become unwieldy until finally, one day, they break.

The alternative is to take a structured approach to business intelligence and reporting in particular. Companies need repeatable, reliable, robust reporting processes that can be rolled out to the entire enterprise. As a first step, that means parking the concept of the human report server. It does not mean abandoning spreadsheets where they add genuine value. Neither does it prevent executives from lifting information and analysing it for specific day to day tasks. What it does mean is that reporting must start from a base that conforms to well understood, documented and auditable principles and processes. That may include the inclusion of spreadsheets, but only where they have passed muster.

The next step is to understand where the most important information resides and integrate those data sources. This is not easy because disparate applications understand the same data in different ways. What’s more, it is becoming increasingly apparent that financial information generated from a variety of back office systems alone does not provide the context decision makers require. Web reports for instance add context to results. Companies therefore need ‘meta data,’ a kind of dictionary about data that ensures that the right pieces of data are brought together. Creating a meta data repository takes time and effort because it forces people to think about what they’re reviewing from a different perspective. Rather than seeing data as it relates to their job function, they start to see it from an enterprise perspective. A beneficial side effect of creating meta data is that it exposes data redundancy, allowing companies to improve their efficiency by isolating and phasing out those processes that don’t add value to business information.

Once companies have reliable meta data tied to repeatable processes, then decision makers start to see information that has real business relevance and which is prepared in a timely manner because everyone knows how long it takes to complete processes like the period close. But there is one final twist.

It is said we live in a world where decisions have to be taken in real time. No where is this more true than in financial services where decisions about equity pricing for example have to be taken on a minute by minute basis. Make the wrong decision and you lose millions. Worse still, you may even step over the regulatory line. Make the right decision and you’re a hero. So exposing relevant, aggregated business information in real time becomes crucial. This includes the same information that is used elsewhere in the enterprise. This does not mean there is a single information truth but multiple views dependant upon need and enriched by non-financial but highly relevant data. For example, if you’re trying to predict gas futures then it helps to know historic values compared with given weather conditions. When this level of data integration and dissemination is achieved, then information is everywhere – period.