The operational difference between phishing and spear phishing calls for different metrics and controls. NIST, PCI DSS, and CIS all agree on continuous training and the ability to recognize and report social engineering, but the evidence in the material does not support measuring only clicks or treating traditional training as effective by default. To make the right decision, a CISO has to separate mass campaigns from targeted attacks, compare results while controlling for lure difficulty, and prioritize delivered credentials, reports, and response time.
Placed side by side, the sources show something that is not visible when they are read separately. The standards describe awareness obligations and evidence, the guidance suggests measurement, and the benchmarks and studies make clear that clicks alone are not enough to understand human risk.
| Topic | Attack type | Definition | Goal | Distinctive trait | What is measured | Decision it supports | Source |
|---|---|---|---|---|---|---|---|
| Phishing | General family of deception | A technique used to obtain sensitive data through a fraudulent request by email or website, and also the use of a forged email to collect private information, deliver malware, or commit fraud | Broad audience, usually with generic messages | Seeks to persuade a person to hand over sensitive data to an impostor | CISA describes a six-week assessment that measures the workforce's tendency to click on phishing lures. Fortra reports click rate, credential submission rate, and report rate. The NIST Cybersecurity Framework 2.0 Concept Paper gives as an example the number of phishing attempts prevented and 80% of users reporting known attempts | Guide antiphishing training and awareness and evaluate effectiveness with more than one metric, not just clicks | NIST, NIST SP 800-177 Rev. 1, CISA, Fortra, NIST Cybersecurity Framework 2.0 Concept Paper |
| Spear phishing | Targeted phishing variant | A highly targeted phishing attack, usually aimed at one person or department within an organization, where the adversary knows the victim's profile and uses contextual information to make the message seem legitimate | One specific person, group, organization, or department | Target selection, prior research, and personalization using contextual information or known personal data. ENISA also adds frequent use of OSINT | Not specified as a single benchmark in the standards. NIST Phish Scale is designed to score the human difficulty of detecting a phishing email and make simulations comparable across lures. INCIBE describes follow-up actions if credentials were submitted | Segment by role or criticality and do not compare campaigns with different difficulty levels without normalizing the lure | NIST, NIST SP 800-177 Rev. 1, ENISA, NIST Phish Scale, INCIBE |
| NIST SP 800-53 Rev. 5 | Does not define a separate requirement for phishing vs spear phishing, but does mention spear phishing in exercises | Requires security and privacy literacy training for users, managers, and contractors. It also requires training to recognize and report potential and real instances of social engineering and social mining | All system users, including managers, senior executives, and contractors | Requires practical exercises that simulate events and incidents, including unannounced social engineering attempts and the simulated impact of opening malicious attachments or activating malicious links through spear phishing | Does not specify a single numerical metric. AT-4 treats training records as evidence separate from the training activity itself | Separate compliance evidence from actual effectiveness and use practical simulations, including spear phishing, not just initial training | NIST SP 800-53 Rev. 5, NIST |
| PCI DSS v4.0 | Phishing and related attacks within awareness, with a separate technical control for phishing | Training must include awareness of threats and vulnerabilities that may affect the security of the CDE, including the words phishing and related attacks and social engineering | All personnel | PCI DSS v4.0.1 states that security awareness education is an ongoing activity. It also separates requirement 5.4.1 for technical and automated controls to detect and protect personnel from phishing from requirement 12.6.3.1 for training | Not specified as a KPI. It does require a formal program, and the v4.0 change summary says to review and update the program at least every 12 months | Do not treat phishing as addressed by awareness alone or by technology alone. Both fronts must be maintained separately | PCI DSS v4.0.1, PCI DSS v4.0, PCI Security Standards Council |
| CIS Controls v8.1 | Phishing as an example of social engineering to recognize | Requires establishing and maintaining a security awareness program and training the workforce to recognize social engineering attacks such as phishing, BEC, pretexting, and tailgating | Workforce members | The policy template makes clear that an effective program should not be limited to a generic annual video paired with periodic phishing tests. In addition to annual training, there should be more frequent themed messages and notifications | No numerical KPI specified in the cited control | Avoid minimum-compliance programs based only on an annual course and periodic tests. Maintain more frequent interventions | CIS Controls v8.1, Center for Internet Security |
| Operational metrics observed | Simulated phishing campaigns | Benchmarks of human behavior in simulations and response measurement | Recipients of simulation campaigns | Fortra reports 5.42% clicks, 1.99% credential submission, and 10.50% reporting across 13.9 million recipients and 7,504 campaigns. Proofpoint reports a 4.93% average failure rate, 6.59% in attachment-based campaigns, and 2.46% in data-entry campaigns. It also reports that 10% of users accounted for 100% of clicks within a given wave | Clicks, credential submission, reporting, lure type, and risk concentration by user | Prioritize submitted credentials and reporting, and focus controls on higher-risk groups or users rather than reading click rate as a sufficient indicator | Fortra, Proofpoint |
Do phishing and spear phishing require the same measurement?
No. NIST defines phishing as the general family of deception and spear phishing as a highly targeted variant against a specific person or department. NIST SP 800-177 Rev. 1 adds that in spear phishing the adversary knows the victim's profile and uses contextual information. Whalemate's reading is that this difference changes measurement: mass campaigns can be tracked with aggregate KPIs, while targeted campaigns require segmentation by role, exposure, criticality, and degree of personalization.
What changes when the attack is mass and not targeted?
According to NIST, general phishing usually targets a broad audience and uses generic messages. CISA describes the Phishing Campaign Assessment as a six-week evaluation that measures the workforce's tendency to click on phishing lures. Whalemate's reading is that, at this level, measurement helps reveal aggregate behavior, but it is still incomplete if reduced to clicks and not paired with delivered credentials, reports, and time to report.
What changes when the attack is spear phishing?
NIST and ENISA agree that spear phishing includes target selection, prior research, and personalization. ENISA adds that this personalization often relies on OSINT. Whalemate's reading is that it makes no sense to compare a generic campaign with a targeted one without controlling for the human difficulty of the lure, and that is why NIST Phish Scale shifts from a methodological detail to a condition for interpreting results without self-deception.
Do the standards call for continuous awareness or is one course enough?
One isolated course is not enough. NIST SP 800-53 Rev. 5, in AT-2, requires security and privacy literacy training as part of initial training and at a frequency defined by the organization, and it also requires training to recognize and report potential and real instances of social engineering and social mining. PCI DSS v4.0.1, requirement 12.6, literally states that security awareness education is an ongoing activity. Whalemate's reading is that compliance requires continuity, not a once-a-year event.
Do the standards distinguish awareness from technical controls?
Yes, and that point is often lost in practice. PCI DSS v4.0 separates requirement 5.4.1 for technical and automated controls to detect and protect personnel from phishing from requirement 12.6.3.1 for training. NIST SP 800-53 Rev. 5 also separates the training activity from its records, since AT-4 treats training records as evidence distinct from the training itself. Whalemate's reading is that there are three separate layers: training, technology, and documentation evidence.
Does the evidence support using clicks as the main KPI?
Not as the only indicator. CISA uses clicks for a specific assessment, but Fortra also reports credential submission and reporting, with 5.42% clicks, 1.99% credential submission, and 10.50% reporting across 13.9 million recipients and 7,504 campaigns. The NIST Cybersecurity Framework 2.0 Concept Paper gives examples of measuring prevented phishing attempts and users reporting known attempts. Whalemate's reading is that the click is an initial signal, not a sufficient reading of risk.
What does practice show about the real risk behind a low click rate?
Proofpoint reports that even best-in-breed companies still registered more than 1% clicks in its measurements, so a low rate does not equal zero risk. The same Proofpoint report says that, in one malicious attack measurement, 10% of users accounted for 100% of the clicks within a given wave, and that 40% of clicks often comes from users who clicked only once. Whalemate's reading is that the average can hide risk concentration and critical users.
Does continuous training always work?
Not consistently. A large-scale experimental study available on arXiv reported a general click rate of 10.4% and found no statistically significant differences between the control group, the training group, and the training plus exercise group, either in click rate, with p=0.450, or in report rate, with p=0.417. Whalemate's reading is that the regulatory duty to train remains in force, but it does not justify assuming behavioral effectiveness simply because traditional training was delivered.



