Now "google" is a new word in the dictionary—an indication of how widespread and well known Google searches have become for the public at large.
Just as the Internet itself, and then Google, went from novelties to necessities almost overnight, Enterprise Search is coming on strong.
Within the corporate world, information on demand isn’t quite so easy. Many organizations have implemented basic search capabilities, with unsatisfactory results. Frustrated, business leaders ask why the search for information should be so much harder at the office.
Accenture believes that being able to easily access all available information, then use it to meet business challenges, explore opportunities and exceed the competition, is essential for achieving a higher level of performance. And that goal is increasingly within reach because Enterprise Search technologies are advancing rapidly. Business leaders who are disappointed in their search capabilities thus far will find that more sophisticated technologies truly can meet their business needs.
Enterprise Search Technology at Work
Early search capabilities were limited: a user could only perform basic keyword searches and sort results using parameters such as creation date. Much more is possible today. True Enterprise Search can function as a technology platform, providing access to any and all data across the enterprise. Both structured and unstructured data can be searched. And natural language processing enables the search engine to understand the intent behind a user’s query and provide a meaningful response.(For more on the technology itself, see "How does it work?")
How can today’s best Enterprise Search technology help a busy executive? Imagine Ms. Elizabeth Jones, head of the Breakfast Products business unit of FoodWise, a convenience foods company. Her business unit has introduced a whole-grain breakfast flake, known internally during development as Product 217 and now launched worldwide as WheatMax.
Sales are slow, and Ms. Jones is concerned. She types "Product 217 Sales" into the Search Bar on her portal. The search engine understands that Product 217 is also WheatMax and will return information for both names—an important capability for searches that involve both internal and external sources.
It’s a broad query, and the search engine can return information from a variety of internal sources, such as sales records from products from FoodWise’s enterprise data system; research reports on consumer habits and trends for breakfast products; and status reports from Ms. Jones’ staff. It can also access external data such as news stories from pre-defined sources such as CNN.com, bbc.com, and various industry publications.
The results of the query are categorized using the company’s predefined taxonomy, so Ms. Jones gets information categorized as product financials (with the most recent sales figures shown first), product research, team status reports, external news, and so on.
Ms. Jones, though, is looking for insights and answers:
- How do WheatMax sales compare with those of other breakfast cereals, inside the company and by competitors? A search for all cereal sales yields an industry analyst’s report, which shows no clear trends.
- Did recent events, such as the record cold winter in Europe, affect the sale of cold breakfast cereals? By broadening the search to encompass all her business unit’s breakfast cereals, then specifying sales for Europe over the past year, she can see a significant seasonal drop across brands.
- What about the whole wheat factor? A search for sales data on all whole-wheat products within FoodWise shows that sales are steady in the United States and Europe, and on the rise in Latin America—a new marketing opportunity to explore? And, a TV news story mentions that many Americans now eat breakfast while commuting, which raises possibilities for product reformulation or packaging. In fact, she learns, some of her team exchanged e-mails on the possibility of WheatMax breakfast bars early in the product development process.
Ms. Jones now has new insight into her product sales numbers, and additional avenues to explore for boosting them.
Early Adopters Gain Benefits
According to IDC¹, companies in a number of industries are investing in new Enterprise Search technology because they are convinced they can quickly recoup their investment and then use their enhanced capabilities to outpace competitors.
Accenture’s own experience confirms the IDC findings. Early adopters are moving ahead—and realizing significant benefits:
- A single view of all enterprise information. A good example is Statoil, Norway’s largest company, with nearly 24,000 employees in 29 countries. Statoil’s new intranet search capability allows collaboration and information sharing round the clock worldwide, with easy, correct and secure access to huge volumes of data throughout the information life cycle.
- Tangible productivity enhancement, better regulatory compliance and improved decision making. Companies report that they have reduced the average employee’s time spent looking for information by 15 to 30 percent, and also lessened the time spent re-creating lost information2.The cost and time of adhering to regulatory compliance demands and responding to litigation requests can also be reduced.
Proven ROI. In call center cost management, Enterprise Search enables more questions to be answered automatically online, which can save from $4 to $40 versus a call handled in the traditional manner by a call center representative. In e-commerce, enhanced Enterprise Search has improved click-through rates, improved sales force access to product and customer information, and thus increased revenue3.
For example, a large European retailer offers customers a 150 percent rebate on same products sold at lower prices elsewhere. New Enterprise Search capabilities allow business managers to monitor the Web for competitive prices on those products that can accurately be considered comparable to the retailer’s own products. By ensuring that its own prices are appropriately aligned and that its generous rebate offer is only extended when necessary, this retailer is using Enterprise Search capabilities to protect and enhance its ROI.
What lies ahead?
While Enterprise Search offers real benefits today, next-generation technology offers even more intriguing possibilities. These include:
Advanced Analytics and monitoring, which makes it possible to tap information essentially in real time by monitoring content as it arrives in databases or on the Internet, then responding to predefined queries with up-to-the-minute responses.
Factiva, a Dow Jones and Reuters company, is already beginning to use this concept. Factiva provides business news and information from more than 10,000 authoritative sources that enables professionals to make effective and informed decisions. In the past, data was indexed at intervals, and users initiated their searches by querying a database that was static and so always somewhat out of date. Now, Factiva users can pre-define search queries called Triggers, which are automatically checked against incoming data, triggering an e-mail or other personal alert to the user in real time.
Sentiment analysis, an emerging technology that can extract the sentiment and tone of a document (negative or positive around a specific subject) by analyzing the text itself.
For example, a senior executive might search for analyst reports on the Internet after the annual earnings announcement for his company is made. By applying sentiment analysis to this search, the text for all of the analyst coverage on the company can be analyzed to determine whether there was an overall positive or negative response in the marketplace to the recent earnings announcement.
Multimedia search capabilities, which enable users to query video and audio sources. For example, a user might access the particular portion of a newscast that addressed the impact of a possible railroad strike, or get a list of all recorded call center phone calls in which users complained about the company’s latest product release. Government agencies could utilize audio search capabilities to identify potential threats such as the use of the word "bomb."
Guided information discovery, an advanced capability made possible by several emerging technologies such as concept clustering and entity extraction. These technologies make it possible for a user to explore information without having a specific or defined query in mind, because the search engine will be able to extract relevant ideas or topics from material in dynamic response to the user’s input. Inputting a few key words will produce a list of perhaps 20 possible topic areas, each of which can be explored through drilling down until the user ultimately finds the right topic and the relevant documents or data.
For example, a pharmaceutical company employee might search the intranet for more information on a heart medication that the company is developing. The search engine could generate not only a list of documents that mention this drug, but also the names of people affiliated with the drug’s development who are mentioned in these documents. Clicking on their names would bring up their contact information so the employee could reach them via e-mail or phone.
Just as the Internet itself, and then Google, went from novelties to necessities almost overnight, Enterprise Search is coming on strong. New generations of technology promise to give business leaders what they want: the ability to find information they can use—even when they’re not sure exactly what they need. Savvy leaders are learning about this new Enterprise Search technology and putting it to work so that they can start achieving higher performance now.
How does it work?
Enterprise Search platforms sit on top of, and integrate with, various information sources across the enterprise—everything from ERP systems to Web servers—as well as with external sources such as analysts’ reports, blogs or extranets. A search platform provides the ability to intelligently and dynamically access, retrieve and analyze information in real time, regardless of data format, structure or location.
Think of Enterprise Search as having three components:
Indexing
Just as the index of a book indicates where key pieces of information are found in the text, the Enterprise Search platform does not store data but rather points to its location.
Data from all sources, of all types, and from all locations across the enterprise, are read and brought into a single search index. Note that unstructured data, which now accounts for over 80 percent of enterprise data, can now be included in the search. A PowerPoint document is unstructured, but the text within it can be read and indexed. A phone call can be converted to text via speech recognition technology. Even photographs can be indexed by reading the metadata (data about the data) associated with the image.
Of special importance to business, data security can be managed and maintained through the indexing process. For example, each document can be assigned an Access Control List, which can then be indexed. In that way only authorized readers can tap into any particular document.
- The query-response process
Users typically express a query as a few key words or a phrase, which the Enterprise Search engine must understand, and possibly expand upon or improve. This ability to comprehend the thought behind a query, called "natural language processing," is a new technological advance.
Enterprise Search matches the query with information in the index, then ranks the documents in order of relevance. Vertical search applications can be built to filter and prioritize for particular industries or specific needs. Furthermore, responses can now be personalized so that results are ranked and displayed in a way likely to be most helpful for that particular user. For example, a California-based accountant searching for legal help could have relevant state laws appear at the top of the response, with the most recent legal changes given highest priority.
- Search refinement
A comprehensive response will by definition be broad, sweeping in more information than the user wants. A user should be able to "drill down" by providing additional qualifiers, honing the response for greater accuracy.
1 IDC Technology Assessment: "Is Disruption Inevitable? Modeling the Future of the Search and Retrieval Software Market," Susan Feldman and Anthony C. Picard, August 2004, IDC#32643, Volume 1, Tab: Markets, pp.6–7.
2 IDC: "The High Cost of Not Finding Information," Susan Feldman and Chris Sherman, April 2003, IDC #29127, Vol. 1.
3 IDC: "The High Cost of Not Finding Information," Susan Feldman and Chris Sherman, April 2003, IDC #29127, Vol. 1.