stokkete -


What to consider when developing a corporate taxonomy

Finding information efficiently within a company is key for user productivity. Businesses should develop an enterprise search strategy to improve content navigation and discovery.

Searching for information within a business application and coming up empty-handed can be both frustrating and a waste of time.

Creating a corporate taxonomy can help users sift through a significant volume of business information to find what they need when they need it, enabling instant access and improving the user experience.

A corporate taxonomy is a way of tagging and classifying digital content in different information systems -- such as content management systems and knowledge management systems -- to enable discovery in Enterprise search results. But creating that taxonomy is only one part of an enterprise search strategy, which helps improve business processes and workflows.

Content management and knowledge management systems aren't the only places in which businesses create and store content. Other systems include ERP and CRM applications, document management systems, digital asset management systems, email and collaboration tools, homegrown file systems, intranets and corporate websites. Enterprise search helps users in any business department locate content despite the presence of silos. Without enterprise search, users must look for content across many systems and use the vendor's own search mechanism in each. Even when users find content, it may not be the content they were looking for.

Incorporate enterprise search software into strategy

Before creating that corporate taxonomy, businesses should enlist the aid of enterprise search software. Many companies may already have access to platforms such as IBM Watson Discovery and Salesforce Einstein Search. But there are a number of other software options available from which businesses can choose. Organizations can analyze search platform options through a specification and selection approach that itemizes the capabilities of the search platform by weighing how important each feature is to the business.

Even when users find content, it may not be the content they were looking for.

Businesses that have already deployed some level of search capability, such as SharePoint, can include an existing search system in the specification and selection process, or it may not be appropriate. Instead, the business can turn on features and functions of the existing system to fill the gaps between what is in place and what it wants the system to do.

Enterprise search software can help an organization in the following ways:

  • Improve user experience and aid productivity by reducing time to find information. For example, employees can use enterprise search to locate a document without needing to log in to multiple systems individually.
  • Reduce risk by highlighting sources of truth for the same content, such as several versions of the same contract that multiple people edited and stored in different systems. Another example includes highlighting missing content, such as an unfiled customer nondisclosure agreement.
  • Enhance business insights by showing structured and unstructured business content in the same dashboard. This could include displaying recent customer complaints logged in the CRM system alongside sales orders from the ERP system and sales contracts from the document management system.

Enterprise search systems enable businesses to mine unstructured data for information which they can display alongside structured data. Enterprise search is the business intelligence for unstructured data.

Businesses should consider the following steps when developing an enterprise search strategy.

Know what the end user expects from the enterprise system

First, organizations should create user stories. User stories describe interactions with enterprise search and provide businesses with insight on how to test the system. When creating user stories, businesses should survey employees to understand their roles and what they expect from the enterprise search system. Companies should also be familiar with their organization's information governance policies and confidentiality rules when developing the enterprise search strategy.

Next, businesses and end users must agree on critical success factors to measure success. Any project or initiative should have ways to measure success, which include time saved, number of content sources included in search and anomalies reduced. The most important measures are known as critical success factors.

After companies determine critical success factors, they should assemble consistent subject matter experts and sponsors from across the business -- including competencies from performance learning and development -- to drive adoption of a new tool or process for success. If no one uses a tool or process and instead relies on old ways of working, businesses cannot measure the value of the process.

Once organizations determine which new tools to adopt, they must build a roadmap of prioritized activities. Businesses should tackle high-priority content sources, first applying minimal but necessary metadata classifications. Once a framework and process are in place, companies can add more content sources and metadata.

Understand and organize content with a corporate taxonomy

Next, businesses should create a corporate taxonomy -- a collection of all the words that companies can use to describe content across all business functions -- and metadata characteristics to support user stories. Tagging content helps improve the accuracy of search results. Tagging, or adding metadata, means making unstructured content structured. Then, businesses can process it alongside structured content.

For example, unstructured content within a business's CMS might include a document with instructions on how to set up employee voicemail. To enable this document to turn up in search results -- and turn it into structured content -- organizations might describe the content with tags such as "voicemail," "how to set up voicemail" and "phone system."

Businesses can speed up the corporate taxonomy structure and tagging processes by using enterprise taxonomy management software that provides automatic classification services using natural language processing and semantic analysis. Vendors that provide this type of software include Synaptica and Wordmap.

After companies add structure to data, they must identify content sources, data structures and system integration capabilities. Structured content sources include CRM and ERP systems. Content within these systems contains data fields, such as name, address, product code and unit price. Unstructured content sources include file systems, document management systems and digital asset management systems. A file system has some structured content, including file name and directory path; however, the files themselves are not structured. A document management system contains structured data -- metadata tags, file names and version numbers -- but also unstructured data, which makes up the content of those files.

An enterprise search system must integrate with other applications to gather content for analysis. A mature enterprise search system can connect to many different business systems, such as enterprise content management, payroll, ERP and websites.

Next, organizations should map content lifecycles, specifically content ownership. Users create, change, store, archive and delete content. Understanding the flow and who is responsible for the integrity, accuracy and retention of content tells businesses who can help define metadata.

Next Steps

How a content tagging taxonomy improves enterprise search

Dig Deeper on Enterprise search platforms and strategy

Business Analytics
Data Management