The name Google has become synonymous with the concept
of searching the internet. Yet its success in raising the profile
of search tools - along with others such as Microsoft MSN, Ask
Jeeves and Yahoo - has, to some extent, undermined the seriousness
with which search technology is treated within the
enterprise.
In the past few years we have seen the emergence of tools
designed specifically to search a company's intranet, rather than
the internet. It is worth noting the difference between an internet
search engine and an enterprise search tool.
Mike Lynch, chief executive of Autonomy, one of the biggest
providers of enterprise search tools, said, "When you talk about
search, people think of internet search engines, but they are
irrelevant in the enterprise, because there is a need for access
control and security. Enterprise search tools have to be able to
understand who can see which document - and 80% of the data in an
enterprise is unstructured. It is also stored in many formats in
many repositories. So the technologies should be different."
But the most important issue to consider when choosing
enterprise search technology, according to Lynch, is that it
correctly meets the search requirements of the organisation.
What is becoming apparent is that the misconception of
enterprise search as a "nice to have" is being seriously challenged
as the volume of data stored continues to grow at a staggering
rate, and the need to identify and retrieve information across the
enterprise - from the desktop to the intranet to corporate data
repositories - becomes increasingly important in boosting
productivity and efficiency, and ultimately competitive
advantage.
Effective enterprise search technology is about far more than a
white bar and a list of results. It should play a key role in
compliance, business intelligence, market intelligence, risk
mitigation, and a host of other applications.
For example, increased regulatory pressures on organisations
require them to be able to search through and make sense of the
information they store.
At the same time, information stored across various data
containers, including e-mail servers, desktops, enterprise
application databases, content management systems, file systems,
intranet sites and external websites, could hold vital information
or competitive intelligence that can only be located or even
identified using sophisticated search tools.
Enterprise search technology should also help to minimise
duplication of effort. It is estimated that knowledge workers spend
more than twice as much time recreating content than they spend
creating new content. If they are able to focus on creating new
content, the benefits and productivity improvements could be
substantial.
Mike Davis, senior research analyst at Butler Group, said, "IT
suppliers often say that between 20% and 80% of a knowledge
worker's time is spent looking for information.
"So by identifying important information faster and more
accurately, an effective enterprise search tool should allow
employees to become more efficient and productive, increase job
satisfaction and identify anomalies in the quality of data
stored.
"In theory, enterprise search technology should therefore
provide competitive advantage."
A Butler Group report on document and records management goes
even further. It said, "The greatest challenge for organisations,
and the one that Butler Group poses as the highest risk of failure
to achieve, is the ability to locate information."
According to Davis, desktop search is the starting point for any
enterprise search strategy, as it is estimated that 80% of
corporate information exists on employee PCs.
"The reality for all of us is that a great deal of time is spent
searching for content that is created and stored locally, and
unless we have implemented a decent taxonomy on the C drive, and
adhered to it, each of our PCs is just an unstructured data
dustbin," he said.
Indeed, Microsoft recently reiterated its focus on enterprise
search when it announced a desktop search addition to MSN Search,
which allows enterprise users to scan e-mail, desktops or shared
network drives on machines running Windows 2000 or XP and
Office.
In doing so it joined the ranks of Google, Verity, Fast and
Yahoo, which already offer desktop search software to
enterprises.
Angela Ashenden, senior analyst at research firm Ovum, said,
"Desktop search is relatively new, with few people using it in the
enterprise yet, but that will change.
"Enterprise search, meanwhile, provides the ability to search
information across an entire organisation, primarily internal
content such as the intranet, but also external content, such as
subscriptions on the internet."
According to Ovum, sales of enterprise search and discovery
technology grew by a 37% in 2004 to £290m, and it is expected to
expand with a compound annual growth rate of 16.4% over the next
five years to reach £640m in 2009.
Most enterprise search suppliers offer some form of desktop
search tool as a matter of course, considering it merely the
starting point of a true enterprise search tool.
In an attempt to distance themselves from the "Google internet
search" perception, enterprise search providers are focusing their
products along the lines of information access, or knowledge
discovery and retrieval.
The "search" label, in fact, underplays some of the more
interesting and sophisticated capabilities offered by enterprise
search tools. The search engine, as used on the internet, uses
mathematical algorithms to rank results and is the most basic form
of technology.
But enterprise search and discovery products offer increasingly
sophisticated functions, such as access control and security,
auto-categorisation, taxonomy navigation, concept-based retrieval,
search analytics, and visualisation and personalisation. In future,
natural language search and semantics will replace the keyword
search.
Suppliers' claims that they can measure the return on investment
from search tools are difficult to verify. However, Whit Andrews,
research vice-president at Gartner, has some suggestions.
"Enterprise search tools decrease the time spent finding and
recreating information and offer a substantial payback in terms of
improvement in the way the enterprise does business. They should
provide insight, collaboration and efficiency," he said.
"The most important thing a CIO needs to understand is that we
are now dealing with the 'Google generation' - everybody working
for you now used Yahoo throughout school and Google in college, and
cannot imagine a world that is not searchable. Where it was a
curiosity 10 years ago, it is now critical."
So how does one select a suitable search tool? Ashenden believes
users should consider both high- and low-end products.
"Some parts of the organisation have complex requirements, and
need a high-end product, while others need a Google-type solution,"
she said.
She suggested that a Google-like product is a good starting
point, allowing an enterprise to understand what it really needs
from a search tool.
"It is then in a much better position to know what it wants, in
terms of complex search requirements, from suppliers such as
Verity, Autonomy and Fast," said Ashenden.
Enterprise search suppliers and their
products
The five leading enterprise search suppliers - Verity, Autonomy,
Fast Search & Transfer (Fast), Google and Microsoft - accounted
for almost half of all sales in 2004.
Since then Autonomy has acquired its main rival Verity for
£290m. Its Intelligent Data Operating Layer (Idol) is a high-end,
neural-network-based platform that connects more than 300 formats
and data and information repositories across the enterprise.
Google's products include Search Appliance (for the enterprise),
Google Mini (for SMEs), Desktop Search for Enterprise (desktop)
and, of course, its internet search engine.
As well as MSN Search and MSN Desktop Search for the enterprise,
Microsoft offers search through Sharepoint Portal Server.
Fast's Enterprise Search Platform (ESP) is designed to help
firms build bespoke applications with customised user interfaces
for specific functions or user groups.
Omnifind is IBM's search tool. It offers enterprise search
capabilities across corporate intranets, public websites,
relational databases and content management systems, but it is not
tuned to specific industries.
Endeca focuses on guided navigation, which exposes different
facets of the same information.
Convera's Excalibur provides a secure index of the web, and the
company continues to sell Retrievalware, its information access
suite of technologies.
Inxight recently acquired the enterprise search assets from
Intelliseek, which will complement its own natural language,
visualisation and analysis product.
Open Text Livelink Discovery Server offers a scalable platform
for conventional enterprise search.
Other enterprise search suppliers include Dieselpoint, EasyAsk,
Entopia, Hummingbird, InQuira, Isys Search Software, Kaidara
Software, Knova Software, Mercado, Mondosoft, Recommind,
Thunderstone, Vivisimo and Zylab.
Case study: BAE Systems reduces retrieval
time
Defence and aeronautics manufacturer BAE Systems deployed
Autonomy's Intelligent Data Operating Layer (Idol) platform, after
discovering that more than 80% of networked employees were wasting
an average of 30 minutes a day retrieving information and 60% were
spending an hour or more duplicating the work of others.
The first two departments to implement Idol were corporate
communications and the virtual university (BAE Systems' learning,
research and best practice division). They use the platform to
aggregate structured and unstructured information from the company
intranet and 10,000 news feeds per day.
This content is then automatically categorised without the need
for manual intervention, allowing employees to navigate very easily
through the site to access pertinent content. It also automatically
alerts employees to documents in the system that relate to their
areas of expertise and interest.
As a result, BAE Systems has been able to personalise the
delivery of business-critical information to staff across the
company, and has reduced the time spent retrieving information by
more than 90%.
The company said it achieved return on investment seven months
after the initial implementation.
Richard West, organisation and e-learning manager at BAE
Systems, said, "We discovered engineers working, in different parts
of the country, on precisely the same problem - a wing construction
issue - but in very different areas, a military aircraft and an
Airbus.
"One group took the step to establish best practice, which was
transferred to a plant in another geographical location with
multimillion-pound savings. Tools like Idol ensure this sharing of
best practice is more likely, particularly in a global
organisation."