
A useful demonstration looks like the work the computer will do. Choose a few repeatable tasks before testing, so the decision is based on a recognisable workflow rather than a quick tour of the desktop.
Choose a small set of representative tasks
For example, open the usual application, work with a representative spreadsheet and produce a sample document. Describe what needs to remain open at the same time. Keep the trial manageable enough that you can repeat it on another shortlisted configuration.
Use suitable test data
Prepare sample files with realistic size and complexity, but remove customer records, passwords and other confidential material. Arrange any necessary application access through the person responsible for your systems. A demonstration should not require exposing unrelated business information.
Record what the user notices
Note delays, display problems and connection issues against the task being performed. Separate a network or account problem from a computer limitation where possible. If something fails, record the circumstances and ask for an explanation before changing several settings at once.
Keep a decision record
Write down the tested specification, tasks and unresolved concerns. Add the result to the desktop setup plan, together with the other evidence needed to approve the purchase.
Frequently asked questions
Should an office PC trial use confidential live customer files?
Use representative sample data without confidential content wherever possible, and follow your organisation’s rules for any application access.
Why repeat the same tasks on shortlisted desktops?
A consistent task sequence makes the comparison easier to interpret and reduces the effect of different demonstrations.
Does a slow website prove the desktop is too weak?
Not necessarily. Check the connection, service and task conditions before attributing the delay to the computer.