Skip to content

Leadership

Evaluate a request for employee monitoring data

Before producing a report, clarify the work question, test whether the measure fits it and establish the boundaries of the proposed use.

Illustration of a woman weighing a folder and shield.

A senior colleague asks for a weekly list ranking employees by keyboard activity. The monitoring data already exists, and the request is framed as a quick way to identify low productivity. As the team manager, you still need to establish what decision the report would support, whether the measure fits that decision and whether this use is authorized.

The ability to produce a report is not a reason to use it. Before requesting or sharing employee monitoring data, examine the purpose, the evidence and the effects on the people being measured.

Start with the decision the requester wants to make

Ask which problem the report is meant to resolve. Is the concern a missed deadline, an overloaded service queue, a security incident or an individual performance issue? Those questions need different evidence. A general request to find out who is working is too vague to establish which data is necessary.

For a delayed service queue, start with the work: demand, staffing, task complexity, blocked cases and completed outcomes. A rise in keyboard activity might reflect efficient processing, unnecessary rework or a change in the task. It cannot, by itself, tell you which explanation is correct.

GAO’s 2025 review of digital surveillance, drawing on stakeholder interviews and published studies, identified risks from flawed productivity benchmarks and measures that miss parts of a worker’s responsibilities. That finding supports a basic question for any proposed ranking: what important work would this measure fail to see?

Check what the evidence can support

Ask how the data was generated, what gaps it contains and whether people doing different work are being compared as if their tasks were identical. Consider meetings, customer conversations, planning, accessibility needs and work performed outside the monitored application.

A meta-analysis first published in 2022 by Daniel Ravid and colleagues found no evidence that electronic performance monitoring improved worker performance across the samples reviewed, while monitoring was associated with increased stress. It also found more positive worker attitudes with more transparent, less invasive monitoring. The findings do not settle every safety or security use, but they challenge an automatic productivity justification.

Distinguish a signal worth investigating from a conclusion about an employee. If a report appears to show a problem, check it against the actual assignment and give the person an appropriate opportunity to explain context before drawing a performance judgment.

Establish the boundaries of the proposed use

Find out why the information was originally collected, who is allowed to see it and whether the proposed report changes its purpose. Data gathered for system security should not quietly become a productivity league table simply because it is convenient.

The UK’s Information Commissioner’s Office emphasized clear purpose, transparency and less intrusive approaches in its 2023 monitoring guidance announcement. Its current guidance is under review following legislative changes. Use the relevant privacy, HR or legal owner to check current local requirements and your organization’s rules, especially when a request changes the use or recipients of data.

Make the request concrete enough for that review: the question, data fields, people covered, time period, recipients, intended action and proposed retention. Do not send a broad export first and decide its boundaries afterward.

Consider a narrower way to answer the question

If the aim is to understand a service bottleneck, an authorized team-level analysis of waiting time and blocked work may answer more than individual activity histories. Remove detail that does not contribute to the decision. Aggregation can reduce exposure, but a tiny group or a distinctive case may still identify someone.

Explain the approved use in language affected employees can understand, including what the measure cannot establish. A report described as workload planning should not drift into another purpose without review. Treat concerns and correction requests as information about the system’s quality, rather than evidence that someone has something to hide.

Record the decision and revisit it

Document whether the request was approved, narrowed, redirected or declined, together with the reason and responsible owner. If a report is produced, check whether it actually informed the intended decision and whether its access and retention remain appropriate.

A manager’s contribution is judgment: connecting evidence to a legitimate work question while recognizing the limits of the measure and the responsibilities attached to using it.

More to read

Browse insights