First, there is the centrist camp - "We will implement a single system across all our operations in all countries. This will be efficient and economic."

Then there are the die-hard nationalists. "Every office should decide for itself in the light of its own requirements. The people on the ground know best."

Somewhere in the middle are what one might call the federalists. "We don't all need the same system; we just need to have the same reporting. Then everyone will have access to the same information without imposing too much control."

A single system supported centrally, should lead to cost savings in implementation and on-going maintenance;

If the same data is recorded across all countries, benchmarking for cost, profit and performance should be easier (however, bear in mind that not everyone likes transparency!);

Agents with a multi-national client can add value by using aggregated information to identify opportunities;

Multi-nationals will find it easier to manage the portfolio and identify opportunities for improvement with consistent data;

The streamlining of business processes - necessary for any good system implementation - should drive efficiency, cut cost and identify best practice to be shared across the organisation;

When you are put on the spot by the Board you will be able to answer the questions "What is the level of voids in the French portfolio?" or "What proportion of tenants failed to renew in UK compared with Germany last year?" promptly without having to promise to come back with the information a week next Tuesday (assuming the comparable information actually exists);

If the systems is a module of an enterprise system, then integration with the financial and accounting parts of the business is also possible.

If the business being conducted in each country is significantly different, is there anything to be gained by using a system which will suit only one - or none if it is a compromise choice? For example, within a managing agent, some countries may concentrate on transactions, others on management, depending on the local market

It is highly likely - in fact, I would almost bet the farm on it - that even if the type of business is the same across all countries, the business processes in each country will differ. Whose processes are going to be adopted to enable implementation of a single system? The creation of uniform business processes in one country can be time consuming and expensive; to do it for many countries requires yet more skill, experience and resource.

Different legal and customary requirements may make it unlikely that you will be able to select a system that is going to suit everyone (despite what the salesmen say!); I have seen systems designed for one country but with a few tweaks;

There may be internal resistance to having one country impose its local system on the others, even if it is reworked; bear in mind that bespoke changes to systems which enable them to service several countries can be expensive;

How do you agree the "footprint" of the system's functionality? In particular, the split between Estate and Facilities management which in the UK and US are often treated separately but in continental Europe are usually integrated;

The cost and effort of getting agreement to a specification, a selection and an implementation is huge in one country - can it be justified for several?

If one country needs a new systems and another does not because its own works well and the organisation uses the first country as a catalyst to start the project, how should the costs be allocated?

What about language - should the system be in one language or have translated user interfaces?

If you implement a joint system as a big bang, it can be costly and risky, but if you implement piecemeal, there will be no comprehensive data structure at inception so each country which has been completed will need to be re-implemented to allow for additional data in the new implementation.

Each business can select a best of breed, country specific solution which matches its specific business needs;

There are no (local) problems with language (reporting is another issue altogether);

An individual business will not have to work with a non-optimal system - maintaining staff morale;

If the businesses are individual cost and profit centres there will be no questions over subsidising costs in another location;

It will still be difficult to get agreement on streamlining the business processes but much easier than doing so internationally;

A system supported locally is likely to give more personal service to users;

Fewer internal political issues;

Selection and implementation should be quicker than for a multi-national system.

There is unlikely to be consistency of data across the systems in each country. This will make benchmarking difficult for corporates and reduce the opportunities for managing agents to deliver cross border reports to multi-national clients;

If multi national reporting is required (which an increasing number of corporates demand), then amalgamating and analysing the data from different systems will be difficult, time consuming and labour intensive;

The opportunity to share best practice and streamline business processes across locations will be lost;

The opportunity to minimise IT support and maintenance costs will be lost;

The opportunity to achieve cost savings on the initial implementation through economies of scale is lost.

It should be easier to agree a streamlined set of business processes where the countries already have some similarities.

The absolute number of systems, and hence the levels of support and maintenance required can be reduced;

It is more likely that agreement can be reached on a common language;

If the cluster countries s are able to decide on the system to be implemented there will be fewer political issues to contend with.

The opportunity to review business processes across the whole organisation will not be taken up;

If a business requires reporting across different clusters the difficulty and expense of amalgamating and analysing the data still remain;

IT support and maintenance is not completely harmonised, leading to higher on going costs;

The key high level objective of integration is to achieve uniformity of output i.e. reporting, rather than uniformity of input i.e. the underlying data. A data warehouse can achieve this objective.

It will be somewhat easier for an organisation to agree on the reporting output rather than the standardisation of underlying business processes.

The very fact of discussing and agreeing the core set of reports will also go some way down the road of integrating the business. It will be apparent how similar, or how different, parts of the business are and enable senior management to begin to look at how to gain efficiencies across the business;

Additionally, if new businesses or locations are brought in, it will be quicker and easier to integrate their reporting if a clear standard has already been laid down.

Whilst it is fair to say that specifying and implementing data warehouses should not be as huge a task as implementing a cross border property management system, do not underestimate how difficult and time consuming it will be. Many of the aspects of defining a data warehouse specification, such as defined metrics and reporting, will be common to both approaches. Interfaces will need to be built and maintained and in house specialist support and maintenance teams required; this is not a cheap option;

By retaining a variety of underlying systems the organisation will have higher levels of maintenance and support;

The opportunities for streamlining business processes that a single implementation would afford will not be taken up;

If reporting requirements change the data warehouse will require modification.