Taxonomy – The “Name of the Game”

Organizational Background

SPX is a multi-industry manufacturing leader with operations in more than 35 countries and approximately 15,000 associates worldwide. Headquartered in Charlotte, North Carolina, SPX generates more than $5 billion in annual revenue.

Through acquisitions and focused organic growth, SPX has evolved from a U.S.-centric automotive component manufacturer with less than $1 billion in sales, to a globally diversified provider of highly engineered, innovative products and technologies that support two critical needs of modern societies: power and energy; and foods and beverages.

The Challenge

As a global company operating in many countries, across multiple business units, SPX deals with a huge number of customers, suppliers, and partners around the globe. Many of these contacts are involved in multiple projects, in different business units, across different countries.

SPX realized that by astutely managing their documents and email correspondence, they could create ‘business intelligence’ out of them. For example, by capturing knowledge stored in documents and email messages, SPX could locate suppliers who were submitting inordinate numbers of claims across business units and geographies, thusly saving the company millions of dollars.

The problem is that SPX always has thousands of concurrent business projects ‘in the works.’ The ability to analyze and monitor the business through documents tracking was a great idea, but it wasn’t possible because of the complexity in finding and identifying the relevant documents.

Although SPX was using SharePoint as a repository for documents and email messages; many important documents were not getting into SharePoint and those that did were not classified so they couldn’t be located.

The SPX Strategy

Because of legal implications associated with claims, SPX chartered the Legal Department with finding a solution for mining documents and email messages to find potentially-strategic patterns. The success of this project was seen as a powerful way to mitigate costs and streamline operations.

Taxonomy – The “Name of the Game”

Before considering changes in technology, the project lead formed a global committee that included representatives from each department and business unit in the company that participated in the project, including all the project heads, which included 150 participants working in 11 different teams. Each corporate segment and geography was represented in these teams.

The committee was tasked with developing a taxonomy that would accurately classify documents without being too complicated to be practical. Since SharePoint is centrally-managed at SPX, it was relatively straightforward organizationally to pull the team together.

The definition started at the business unit, or ‘segment level’, which is comprised of Industrial Products and Services, Thermal Equipment and Services, and Flow Technology. Each segment spans many projects, but projects within a segment are related to the domain expertise of the unit.

The committee worked for 12 months to define the document storage structure as well as the metadata that would be employed for each library. At the end of the project, a master list of libraries was published. This list dictates the library where each document should be stored by end users. It also specifies the metadata used for each library.

To make the system simple, the committee decided that no library require more than four document properties as metadata. It was also decided that a maximum of one (in some cases two) properties could be supplied without a default value. The use of default values for the bulk of the SharePoint properties makes upload easy, since users only need to manually specify one property value. An example of default values is “project correspondence type” for email messages, which can be ‘internal’ for intra-organizational email or ‘external’ for inter-organizational email. Since most of corporate email stays within the organization, a default value of ‘internal’ is set for an email message uploaded to SharePoint. Users can change default values if needed.

Once the taxonomy was clearly defined, it was incorporated into the company’s SharePoint system. Now, when uploading documents, employees would be required to specify document metadata, so that the taxonomy strategy would be fully implemented.

At this stage, SPX faced several more challenges, namely, how would they get business users to upload documents to SharePoint, since they were used to sharing documents as email attachments. Next, even if they did upload a document, how would they correctly classify a document or message? And who would want to spend all the time and effort to go through this exercise for every document and email message they had? Lastly, when all the documents were inserted into SharePoint and correctly classified, how would project leaders be able to find important information, for example, which suppliers were filing an inordinate amount of claims across the organization.

The Solution

SPX sought a solution to help their business users easily capture and classify important documents in SharePoint directly from Outlook; where users spend the bulk of their work time. SPX built an evaluation scorecard and then evaluated five vendors.

After a rigorous selection process, SPX selected harmon.ie. In addition to its technical superiority, SPX singled out harmon.ie as the only company to present a business solution that would help SPX achieve the goals of enabling knowledge capture and sharing (as opposed to just enhancing document upload). harmon.ie was also deemed the easiest product to use.

How Does It Work?

Like most workers, SPX employees spend a large portion of their day in their Microsoft Outlook email application. One harmon.ie app presents SharePoint within Outlook.

With harmon.ie Outlook Edition, when an employee wants to save a document or email, they merely drag it from their email window directly into a folder into the harmon.ie sidebar, whereupon it gets automatically uploaded into SharePoint.

During the upload, default metadata values for the document or email message are automatically added to the document’s properties in SharePoint, in a process completely transparent to the end user. For the remaining metadata fields, harmon.ie displays a simple to use “Properties Editor” (shown below) that prompts the user to specify data, mostly from a set of pre-populated choices. To ensure accurate classification, users are not allowed to complete the upload process without completing the required set of metadata values.

The end result? Large numbers of documents and email messages are uploaded to SharePoint and classified according to the corporate policy, without requiring employees to change the way they already work today.

Employees can also locate SharePoint documents quickly from harmon.ie using views that have been defined in SharePoint. Default views include views that display documents by content type. Custom views are generated on an ‘as-needed’ basis by employees who put in requests for new views.

Custom views requested by employees represent a large part of the knowledge stored in SharePoint documents, since it defines how employees want to ‘slice and dice’ the information locked in those documents. harmon.ie makes it easy to access these views from a pull-down menu within the harmon.ie window, right within Outlook.

The Rollout

harmon.ie was initially exposed to knowledge experts in each of the department participating in the project. Some training was provided to users by the project lead. Documentation was created that provided instructions about which documents should be stored in particular folders. SPX held weekly review meetings with project managers to get feedback. Two items that were discussed weekly included metadata changes, as well as network performance, particularly to the countries with poor network infrastructure.

The project started as a pilot of 10 projects; from departments like supply chain, proposal management, and engineering. These project managers are the same people on the taxonomy committee, so their participation validated the system’s completeness and usability.

The global rollout includes a six-month migration period.

Looking Forward – Knowledge Management, Mobile, and Social

The project goal is to reach an additional 1000 projects, including all the corporate users who will eventually need to access the systems.