UBS probably has the best website of any broking firm. Can you talk about how the website might have changed analysts' approach?

The website hasn't changed the analysts' approach, but it has made it easier for our clients. It allows them to monitor analyst recommendations and target prices, as well as to see what analysts have said over time. For many clients it is an important tool with which to evaluate the research product.

So the internet has made the research world more transparent?

Yes. But it's also valuable because it allows us to see which clients use the website. The larger and more sophisticated the client, the more they use the website so we believe that many things that differentiate us as a research house - such as access to working/valuation models into which clients can insert their own assumptions - may have been underplayed before the widespread use of the web.

The website was a considerable investment. Do you feel it has paid dividends by positioning you at the top end of the business?

A lot of clients complain about the excess of sell-side research. Offices were filling with paper. Faxes were tried. Then emails - although McKinsey research shows that around 70% of all sell-side emails are not read. Increasingly, clients want to search for information directly. A powerful website that can deliver that information is an attractive proposition for the client.

So it is a case of pull rather than push?

Yes. Portfolio managers will tell you they get their stock-picking ideas from a variety of sources. It could be meeting a company, it could be talking to a UBS salesperson, or even reading FinanceAsia - putting ideas together, using lateral thinking. Then they want to go and prove the investment case. That's when they will go and look for meaningful equity research on the company. They want to do it as easily as possible and that will be through the web or some of the other electronic channels.

Broking houses used to print a lot of research. Do you print a lot less?

Compared to three or four years ago, we print substantially less research. A lot of what I would term 'maintenance research', such as summaries of results or presentations, is now purely electronic. We only print what we think will be of particular interest - value-added research - and physically distribute it to clients.

There's a barrage of research hitting clients, particularly by email. How do you get noticed?

We think that branding our products is important - so that clients know what they're receiving. We have two or three key globally consistent brands for different types of reports. For example, key industry questions are addressed by the branded "Q Series". As with any other industry, it's essential to build and develop global brands

Has email led to analysts speaking to fund managers less, and sending them email instead?

Our data suggests analysts are talking to their clients more. It's clear clients want more access to analysts, but our analysts' face-to-face time is the most expensive resource we have to offer. Nevertheless, we try to provide access to industry and country experts to those clients that want to pay for it.

Are you able to use the website to figure out who your most popular analysts are - ie in terms of who is read most?

Yes. We monitor the most read documents and what types of research are most useful for clients. The information allows us to decide which direction we should be going in with our product. That's where it is powerful. Historically, we relied on qualitative research to make those decisions - that is, going to see clients face-to-face, or eliciting feedback from salespeople.

Will there ever be a direct link between analyst bonuses and whether they are in the top quartile of most read analysts?

A significant component of analyst compensation is client-related - we rely on client feedback on who they believe are our best analysts. At present, we do not track readerships part of the pay equation, but if the data was sufficiently representative of client value, we might.

How much has the internet impacted analysts and how they obtain information?

It's made a huge change to the way analysts do their jobs. Most information gathering is now conducted via the web. We've also begun to introduce more advanced search tools, that give automatic alerts when the required information appears.

It has made analysts more productive.

What's the next phase? Will it be video?

We are looking at multimedia. It has produced promising results in pilots with certain clients. Some presentations, for example, feature a "talking head" which means clients do not physically need to attend a marketing meeting. This can be attractive to a client to save time, especially if they know the analyst well. We also exploring more sophisticated client datamining so that when a client visits the web they are presented with products that is of direct interest to them, such as research on stock holdings or related to their portfolios.

Will point-to-point videoconferencing over the internet kill the analyst marketing trip?

I don't think so. It's tough to eliminate the fact that a lot of this business is about forming personal relationships with your clients. You still need to see them, say, a couple of times a year, to build on the daily dialogue. So it may supplement the marketing trip, but it won't eliminate it.

Given all the money spent on this technology, has it been a good investment?

It's a business essential. The best research can involve a sophisticated database to capture the financials of the companies we analyse, portfolio risk analytic tools, on line valuation tools, a powerful web distribution channel and a good client relationship system to segment your offering to clients effectively. If you're not willing to invest in the technology, you may not be able to compete It's a big investment, but it has to be made to excel.