Phishing simulations are a powerful tool for improving employee resilience, but poorly designed tests can do more harm than good. Security teams must balance realism with respect for privacy, legal boundaries and workforce morale.
This tutorial lays out a repeatable approach for designing, delivering and measuring phishing simulations that are effective and ethical. It focuses on setting clear goals, building believable templates without crossing sensitive lines, tracking outcomes safely, and using results to coach—not punish—employees.
Define goals and acceptable risk
Start with clear, measurable objectives
Before you draft a single message, decide what the simulation should prove or teach. Common objectives include detecting vulnerable groups, testing a specific control (for example, link filtering or email authentication), or reinforcing a behaviour such as verifying unusual requests.
- Pick one primary learning objective per campaign.
- Choose secondary objectives only if they don’t complicate measurement.
- Document success criteria up front so stakeholders share expectations.
Establish acceptable risk and governance
Simulations expose employees to contrived threats. Treat them as formal exercises with documented scope, approvals and oversight.
- Identify stakeholders: HR, legal, information security, and executives.
- Define prohibited themes and a consent model (explicit, implied, or opt-out where allowed).
- Decide escalation rules for high-risk outcomes, such as disclosure of sensitive personal data or apparent credential compromise.
Create a short policy that spells out purpose, scope, data handling, and escalation so everyone knows the ground rules before the test begins.
Craft believable templates while avoiding sensitive themes
Make messages realistic—but ethical
Realism increases learning. Use familiar senders, plausible subject lines and believable landing pages. But realism must not rely on fear, trauma or manipulation of personal circumstances.
- Avoid themes tied to health, bereavement, legal threats, payroll errors, or other deeply personal topics.
- Do not impersonate the CEO or HR in ways that could authorise real actions or create confusion about legitimate requests.
- Use language consistent with normal corporate communications and include subtle clues that the message is a test for education—for example, an instructional message on the landing page after the click.
Design safe landing pages and assets
If your simulation uses a web form or login page, never collect actual credentials or personal data. Instead, use one of these approaches:
- Show an instructional interstitial immediately on click that explains the simulation and offers guidance.
- Use a non-functional mock login that records the click but does not accept or store submitted credentials.
- Capture only the minimal metadata needed for analysis (for example, timestamp and anonymised user ID) and avoid storing IP addresses or device identifiers unless reviewers need them and clearance is documented.
Technical delivery: options and safe tracking
Choose a delivery method that matches risk tolerance
Delivery can be in-house, via a third-party platform, or through managed service partners. Each option has trade-offs for control, cost and data exposure.
- In-house tools give maximum control over content and data but require development and operational attention.
- Commercial platforms simplify campaign setup and reporting but may store campaign data offsite—confirm data residency and retention policies.
- Managed services reduce internal workload but demand clear contracts that limit data use and prohibit reuse of employee data for other customers.
Track clicks and interactions without exposing data
Logging user behaviour is essential for measurement but must be privacy-aware.
- Log only what you need: anonymised user identifiers, event type (click, open, submit), timestamp, and campaign tag.
- Avoid capturing clear-text credentials, full form data, or any sensitive personal information.
- Use short retention windows for raw event logs and aggregate results for reporting.
- Encrypt stored data and limit access through role-based permissions—separate analytics from HR and keep reviewers accountable.
Technical checklist
- Use an isolated domain or subdomain for simulation assets to avoid interfering with production systems.
- Serve landing pages over HTTPS with valid certificates to model real-world conditions without exposing users to mixed-content warnings.
- Implement strict Content Security Policy (CSP) on simulation pages to prevent unintended third-party content loading.
- Ensure logging systems redact or truncate any free-text inputs and do not log headers that reveal device or location unless justified and approved.
Measuring effectiveness: metrics that matter
Focus on actionable metrics
Not all metrics are equally useful. Prioritise those you can act on to improve training and controls.
- Click rate: the proportion of recipients who clicked a link. Useful as a rough signal but easily gamed.
- Credential submission rate: the number of users who submit data to a mock form. Only acceptable when the form does not capture real credentials.
- Report rate: how many recipients reported the message to the security team or used an authorised reporting mechanism.
- Time-to-report: how quickly users flag suspicious messages after receipt.
Interpret metrics in context
Break metrics down by team, role and device type to find patterns. High click rates in one department may indicate targeted training or technical gaps, such as missing email authentication checks.
- Combine automated metrics with qualitative feedback from coaching sessions.
- Use trend analysis across campaigns rather than comparing single-campaign scores in isolation.
Remediation, coaching and repeating simulations responsibly
Prioritise education over punishment
Phishing simulations are teaching tools. Use results to guide one-on-one coaching and team workshops rather than formal disciplinary action.
- Provide immediate, constructive feedback when a user clicks or submits—an on-screen explanation with tips is effective.
- Offer optional follow-up training modules tailored to the user’s knowledge gaps.
- Coordinate with HR on any policies governing mandatory training or escalation to ensure consistency and fairness.
Schedule and cadence
Run campaigns on a predictable, documented cadence that balances reinforcement with respect for employee time. Vary templates so users do not learn to recognise tests rather than threats. Avoid over-testing the same cohort, which can erode trust.
- Plan a mix of low-, medium- and high-complexity simulations across the year.
- Rotate themes and senders, but keep prohibited topics out of scope.
- Use anonymised benchmarking to set realistic improvement goals for each cycle.
Integrate simulations with broader awareness efforts
Phishing exercises work best when paired with practical guidance on secure behaviours. For example, remind staff about credential hygiene and safe reporting tools.
For background on secure authentication habits you can point employees to a practical guide on password management and passkeys in your learning materials, such as A Beginner’s Guide to Password Managers and Passkeys. For spotting modern lures and cloned pages, consider pairing simulations with the site’s analysis of evolving phishing trends: Phishing in 2026: How to Spot the Attacks That Look Real.
Quick operational checklist
- Define one clear objective and success criteria for the campaign.
- Get written approvals from legal, HR and leadership for scope and sensitive exclusions.
- Use isolated domains and do not collect real credentials or personal data.
- Log minimally, encrypt data and set short retention windows.
- Provide immediate, educational feedback and offer follow-up coaching rather than punishment.
- Measure trends, break down by group, and adjust training based on patterns.
Closing: put this into practice
Designing ethical, realistic phishing simulations requires upfront governance, careful template design and privacy-aware technical choices. Start small: run a single, well-scoped campaign with stakeholder sign-off, measure a few actionable metrics, and use results to coach employees. Iterate with increasing complexity as confidence and controls grow.
Use the templates and checklists above as a blueprint for your next exercise. Keep training constructive, keep data handling conservative, and make learning the primary outcome of every test.