15 January, 1999
Internet Portals
The growth in the concept of the portal
site was one of the internet phenomena of the past year. In
addition to the major search engines, a growing number of
smaller special interest sites have been expanding their
content and providing peripheral services in order to become
portals. During 1999 we believe you will see increased
customisation and personalisation of portal sites, and beyond,
we see the portal extending outside the browser to provide
information to a wider variety of client applications and
devices. We believe that, in future, portals will not just be
accessed through the web browser, but will also provide
information to other desktop applications and to mobile phones,
set-top boxes, and handheld devices.
The Importance of retaining customers
The earliest portal sites were search
engines such as Yahoo, Excite, Infoseek etc. These sites
quickly realised that, while they were being used as an
integral part of the web experience for the growing mass of web
users, this usage was only transitory. Once users found the
destination they were looking for, they would move off the
search site to their destination. It was around this time that
a possible answer to the question of how to make revenues and
profits on the web began to emerge. Initially, it was felt
that you could charge for information. However, the web is a
free-wheeling information culture, and the moment you start
charging for information someone else will provide similar
information for free. In fact, the only companies that can
successfully charge for information are the few that have a
lock on exclusive and desirable content, such as the Wall
Street Journal. The second thought was to provide information
for free, thus driving traffic and allowing sites to charge for
advertising. Yet, it is becoming apparent that advertising
revenue alone may be insufficient to fuel the sort of growth
that investors expect.
Today's thinking pins future web revenues and profits on
e-commerce. The web is becoming more than an information
resource - it is becoming a platform, a medium by which people
will manage their lives. Increasingly, the mass market
consumer will buy clothes and holidays, shop for cars,
mortgages and houses, and bank and pay bills - via the web. It
will be the websites over which people transact that make
money. This is the web revenue model - provide valuable
independent information for free and profit from e-commerce
transactions. In addition to on-site transaction revenue, it
is likely that the advertising of transactional services (which
aims, in most cases, to take users off-site) will increasingly
incorporate revenue sharing on resulting purchases in order to
offset the loss of potential buyers.
This raises the important question "on
which websites will people transact?". People will transact on
sites that they know, trust and visit regularly. Those that
provide a wealth of useful information and make comparison
shopping easy. Those that the user feels a part of, and that
understand the individual user's needs and preferences. The
search engines now see this opportunity. People already know,
trust and visit them. Now, they must provide a wealth of
useful information, get to know their customers, and make
transacting easy. It isn't only the search engines that
recognise this. Increasingly, sites containing
special-interest content are getting to know their users
better, providing additional services, and building a sense of
community - thus becoming vertical (or special-interest)
portals. Most portals and vertical portals are trying to
become more 'sticky' - by providing content that encourages the
user to stay on the site for longer instead of moving on. The
longer a user stays on a site, and the more targeted the site's
content, the more likely they are to make an e-commerce
transaction.
This vision of an e-commerce based
revenue model is confirmed if you look at the valuations of
internet companies in the US. Typically, sites that gets a lot
of visitors are valued at around $10-$20 per page impression
per month. Those sites that get to know their users (by
turning visitors into registered members) are valued at around
$100-$200 per registered member. And those sites where
registered members are known to make a high volume of
e-commerce transactions (such as Amazon.com or E*Trade) are
valued at several thousand dollars per member.
Web usage patterns
There are four principal types of web
usage. A given user may switch between different modes of
usage, but the following generally holds true:
Multiple speciality users use the internet to satisfy two or more 'passion centres'. There are very few sites on the internet that are designed specifically to serve two areas of interest (although certain combinations such as technology and finance, food and art, sport and parenting, could probably attract a fair number of users).
General interest users do not have a strong primary area of interest, and will use the web as a form of leisure and entertainment, for access to the online community, and for specific information as and when required.
Launch pad users are looking for
guidance as to where to go on the web (this includes a high
proportion of novice users). Many of these will use the
portal as a search engine that is a pointer to their
eventual destination.
'General interest users' and 'Launch pad' users are clearly prime targets for portal sites. As they become more customisable and personalisable, general portals will also be able to capture 'Multiple speciality users', who will prefer to have their needs served on a single site rather than switching between several. 'Single speciality users' will continue to prefer vertical portals due to the depth of content and functionality that they offer in specific subject areas.
From website to internet information centre
A portal site essentially aims to capture the attention of users seeking information on the internet. Now consider the channel through which users find this information:
Traditionally, internet companies
(including portal sites) have concentrated on the internet side
of this picture. The thinking has been that while the client
side provides access, the internet side is where the services
are delivered. However, in reality, the internet is not as
platform independent as envisioned, and the client hardware and
software used can have a deciding influence on how internet
services are used. At the browser level, both Netscape and
Internet Explorer guide users to preferred portal sites - for
example Netscape Communicator leads you to Netcenter if you
press the 'Search' or 'My Netscape' buttons, or if you enter
keywords in the URL field.
At the desktop level, a variety of
desktop applications and features are internet enabled. These
include applications that feature scrolling tickers or desktop
information such as Pointcast, Backweb, and Yahoo Ticker, as
well as word processors, spreadsheets, and design programs.
Increasingly, these applications will drive traffic towards
internet portals. For example, AOL's instant messaging client
now allows you to conduct searches - taking you directly from
the desktop client to AOL's portal. Productivity applications
such as calendars, project planners and other organisation
tools particularly lend themselves to the internet environment.
As every application becomes internet enabled, more and more
of these will use information stored by portal sites and will
drive traffic towards portals.
Increasingly, internet capabilities are
being provided by the operating system itself - for example,
Apple's Sherlock component in Mac OS 8.5 allows you to conduct
searches directly from the operating system, using a variety of
portal partners to deliver aggregated results (thus bypassing
the browser). Microsoft's Active Desktop bypasses the browser
too.
The hardware companies themselves are new
entrants in the battle for portal traffic. Recognising that
the internet is a key driver for PC sales, manufacturers are
increasingly working with partners along the access chain to
make the internet a seamless experience. Many manufacturers
are placing 'internet' and 'search' keys into their keyboards.
Dell is developing its own portal in collaboration with Excite,
and in December announced it is providing a co-branded version
of AOL to its US customers. Compaq's CEO announced at Comdex
'98 that his was no longer just a PC firm, but that it would be
developing a series of home 'gadgets' closely integrated with
the web. Beyond the PC, manufacturers of handheld devices,
mobile phones, and set-top boxes will increasingly influence
user choice of portal.
Some UK portal players
The UK portal market can be divided into two categories - UK divisions of US portal companies, and local UK portals.
MSN UK should be regarded as the one of the leading UK portal because of the depth of its UK content and functionality, its presentation, and its personalisation capabilities. It is the only major UK portal to feature a customisable home page (as opposed to a separate customisable site such as 'My Yahoo').
Excite UK. Excite is focussing on the development of its e-commerce revenue through features such as one-click ordering, and through the intelligent use of customer profile information. We expect many of these features to be migrated to the UK during 1999. Excite UK has already succeeded in gaining sponsorship for its Money & Investing channel, and will likely look for more sponsorship deals during this year.
AOL UK does not consider its UK website AOL.co.uk to be a portal (unlike its parent AOL.com). The UK website acts simply to draw people in to the AOL brand and the full AOL service. However, if the proprietary AOL service can be considered a portal, then it is currently the most visited portal in the UK - achieving 111.3 million page impressions per month. It also has the key advantage of having established loyal user communities. Although features can be copied by others, established communities are very hard to move.
Dixons' Freeserve, among UK portal providers, is the most formidable competitor at present - having built its user base from zero to over 500,000 (overtaking AOL UK) in just nine weeks, through the provision of free internet access. Freeserve has executed extremely well, developing a simple installation process, a good portal site and a reliable network. It remains to be seen whether Freeserve's portal site will become a major competitor as e-commerce revenues begin to grow.
Virgin Net represents a big UK brand
that has been fairly low profile on the internet to date.
Nonetheless, the power of the Virgin brand, and the
possibilities of combining services from the 238 Virgin
Group companies onto the internet should not be
underestimated.
What happens next ?
In the United States, we have seen the following transitions in the internet market:
We see two acquisition trends in the UK.
Firstly, we expect the final stage of development shown in the
diagram above to begin taking place in the UK very shortly.
There are a number of well established consumer and media brand
names in the United Kingdom, including the likes of Sainsbury,
Marks and Spencer, Barclays, Virgin, BT, Top Shop, Burtons, BBC
etc. These established brands will increasingly buy internet
companies in order to establish their positions in the internet
space. BT's investment of $10 million for 50% of Excite UK
represents the first major example of such a development.
Secondly, we believe that smaller UK internet companies that
provide a technological infrastructure or have built an
established customer base (particularly supporting e-commerce),
will increasingly be bought up by larger internet players.
These will include providers of services such as classified
advertising, internet auctions, comparison shopping, travel
guides etc.
Christmas 1998 was a major year for the
e-commerce market in the United States. Year on year sales for
online merchants grew around 230%, with revenues reaching $13
billion. AOL reported average e-commerce spend per US member
of around $90 ! We believe that the UK will experience similar
gains by Christmas 2000. Portals will gear up for e-commerce
by transforming their focus from search engine to shopping
engine, offering vendor-independent comparison shopping, single
click ordering, product reviews, and online auctions.
We believe that 1999 will also see a
continuing proliferation of portals - with many more special
interest sites expanding their content and functionality to
become vertical portals. We will see more customisation and
personalisation of UK portal sites - eventually leading to
componentisation, which will allow individuals and corporations
to create their own internet and intranet portals featuring
components from established portal players. For example, an
intranet portal at Shell might consist of industry news from
Newspages, search from Lycos, weather from the Met office,
commodity prices from Reuters, customisable personal content
from Netcenter, and links to a variety of internal corporate
databases and workflow systems.
Finally, we believe that the portal
providers will increasingly deliver personalised information
based on the user's preferences to a variety of applications
and devices. Imagine if an internet portal could store your
bookmarks, contact addresses, custom dictionary, e-mail
settings, one-click order settings etc - and could make these
centralised preferences available to your desktop, your word
processor, your mobile phone, your television set-top box and
your PDA. By providing such services, the portal will move
beyond a website into a personal internet information centre
accessible from anywhere.
Jay Marathe
Head of Consulting
Durlacher
Please contact us for
further information on Diurlacher's research, consulting, or
corporate finance services
http://www.durlacher.com/
jmarathe@durlacher.com
020 7459 3641 © Durlacher Research Ltd,
1999
