Identifying Computers for Data Collection Package Deployment

Applies To: Windows 7, Windows Vista

You deploy your data-collection packages to only a subset of the computers in your organization. To help you determine which computers to use, consider the following:

  • Ensure that you return a complete application inventory, including all of your business-critical and day-to-day required line-of-business (LOB) and third-party applications. We highly recommend that this list include any application that is used by at least one person to perform his or her daily tasks. To ensure that this list is complete, we also recommend that you use a cross-section of computers from your organization.

  • Capture the runtime properties and the dependencies of installed applications, including accessed DLLs and accessed or modified registry settings. These dependencies can help locate potential compatibility issues.

  • Ensure that all device drivers are captured so that the proper impact can be assessed during an operating system or a patch upgrade, in addition to being able to locate potential issue and solution data provided by original equipment manufacturers (OEMs) and independent software vendors (ISVs).

Factors to Consider for Data-Collection Package Deployment

For greater control over your collected data, IT administrators should focus their data-collection packages for deployment to a small subset of computers based on specific categories. For example, a data-collection package targeted to users in the United States Human Resources department. This enables better categorization and analysis of an application throughout the organization.


If your organization already has a hardware-asset inventory list, we recommend that you sample each unique hardware configuration so you can synchronize with the Microsoft® Compatibility Exchange and obtain the relevant driver-compatibility issues. If you do not have a comprehensive inventory, we recommend that you distribute the data-collection packages as widely as possible to obtain this list.

In addition, you must consider the following questions during your data-collection package deployment:

  • Do you have a managed, unmanaged, or mixed environment?

  • How do you use specific applications in your organization?

  • Do you use role-based applications?

  • How do you distribute your applications in your organization?

  • What is the geographic breakdown of your organization?

  • What types of computers do you have in your organization? How are they used?

Managed and Unmanaged Environments

An organization can be categorized into a managed environment, an unmanaged environment, or a mixed management environment.

  • Managed environment. IT administrators strictly control and manage the application installation and the usage based on need and the various divisions in the organization. In this situation, an IT administrator can deploy a data-collection package on a limited subset of computers for each department, based on known needs and requirements.

  • Unmanaged environment. Users typically have administrator privileges on their computers and can install applications at their own discretion. Because users in an unmanaged environment can install any software they choose, you will need to deploy your data-collection packages to more computers than you would if you were in a managed environment.

  • Mixed environment. Your organization uses both managed and unmanaged environments, depending on an individual group's needs and administrative privileges.

Application Usage

It is very important that you provide coverage for all applications required by users in your organization. However, it is even more important that you provide coverage for your LOB applications. For the most complete coverage of application usage, you must:

  • Consult with your local administrators, support engineers, and department leads to ensure that all applications are being used during the data-collection process.

  • Ensure that "seasonal" applications are covered. For example, fiscal year–accounting applications may be used only once a year.

  • Attempt to perform the data collection when there are few employee vacations scheduled or at the beginning of the week to avoid weekends. Otherwise, you may get limited or incomplete results due to the decreased application usage.

Role-based Application Usage

Your organization may use role-based applications. These are applications that relate to job function and the role a user performs within your organization. A common example is accountants (a finance role) and their finance-related applications. Reviewing application usage in conjunction with job function and roles enables better application coverage in your organization.

Software Distribution Practices

You can distribute applications in many ways within an organization, for example, by using Group Policy, IntelliMirror, Microsoft System Center Configuration Manager, or a customized distribution method. Reviewing your software distribution–system policies with your application inventory enables better application coverage and narrows your data-collection package deployment targets.

Geographic Distribution of the Organization

Your organization's geographic distribution must be considered when planning for your data-collection package deployment (for example, if you have branches in North America, Asia, and Europe). You must then consider the application-usage patterns across each geographic region. You will need to account for divisional applications, localized versions of applications, and applications specific to the geographic location and export restrictions. We highly recommend that you consult with technical and business leaders from each region to understand these differences.

Computer Types and Usage

Computer types and usage patterns can play an important role in your data-collection package deployment. The following sections describe some of the most common computer types and usage patterns.

Mobile and Laptop Computers

Mobile users frequently work offline, occasionally synchronizing with the corporate network through either a LAN or VPN connection. Because there is a high possibility of a user going offline for long periods of time, you must consider the odds of the user being online for the data-collection package to be downloaded and installed, and then online again for the logged data to be uploaded.

Multi-user Computers

Multi-user computers are typically in university computer labs, libraries, and organizations that enable job sharing. These computers are highly secure and include a core set of applications that are always available, as well as many applications that can be installed and removed as necessary. Because these computers typically have a basic set of applications assigned to users or computers, you can narrow the application coverage and usage to identify destination computers to receive the data-collection package.


AppStations running vertical applications are typically for marketing, claims and loan processing, and customer service. TaskStations are typically dedicated to running a single application, such as on a manufacturing floor as an entry terminal or in a call center. Because both of these types of computers do not commonly enable users to add or remove applications and may be focused for specific users and job roles, the application coverage and usage can be narrowed to identify what should receive the data-collection package.


Kiosks are generally in public areas. These computers run unattended and are highly secure, generally running a single program using a single-use account and automatic logon. Because these computers typically run a single application, the application coverage and usage can be narrowed to identify what should receive the data-collection package.

See Also


Phase 1: Collecting Your Compatibility Data
Creating a Data Collection Package
Deploying a Data Collection Package