The three major cyber security job families are security analyst, security engineer, and penetration tester. A person in these roles triages alerts and investigates incidents, builds and maintains defensive systems, or attacks authorized systems to find weaknesses before criminals do. Analysts watch and respond, engineers build and harden, and pen testers probe and report β and the shortest honest answer to "what does a person do" is that they spend the day between dashboards, change tickets, and written findings that other people act on.
The field also splits cleanly by employer type: a bank analyst, a hospital engineer, and a consultancy pen tester hold similar titles but see wildly different work weeks. According to the U.S. Bureau of Labor Statistics Occupational Outlook Handbook, information security analysts earned a median annual wage of $124,740 in May 2024, with about 16,600 openings projected per year over the 2024β2034 decade β numbers worth keeping in mind because they describe the largest of the three families, not all of cyber security.
One boundary up front: no job title outrunning the actual duties is worth trusting. A "cyber security specialist" at a 40-person firm may run everything from phishing training to firewall rules, while the same title at a defense contractor maps to one narrow function. Check the duties, not the label.
Security Analyst: The Alert Triage Job Nobody Sees From Outside
A security analyst is the person who owns the queue. Alerts land from endpoint detection tools, firewalls, email filters, and identity systems, and the analyst's job is to decide, quickly and defensibly, which ones are noise and which are an active intrusion.
A typical day starts with an alert queue in a SIEM β security information and event management β platform. The analyst sees flags like "impossible travel" (a login from Tokyo two hours after a login from Chicago), a burst of failed logins against a service account, or an endpoint that started running an unfamiliar script. Reading each one takes minutes; deciding what to escalate takes judgment built from months of seeing what normal looks like on that network.
When an alert clears the noise threshold, the analyst investigates: pulling sign-in logs, checking the machine's process history, asking the employee whether they clicked something, then opening a ticket and writing the timeline. If malicious activity is confirmed, the escalation goes to an incident responder or a dedicated threat-hunting team, depending on company size.
The failure case defines the job. A 2024 report from IBM put the global average cost of a data breach at $4.88 million β a figure that explains why analysts write everything down. The common miss is an alert closed as "benign, user error" that turns out to be a first-stage compromise; that's why mature teams require a second set of eyes on closes involving privileged accounts.
Tiering shapes what a person actually does hour to hour:
| Tier | Typical daily work | Time split | Common title |
|---|---|---|---|
| Tier 1 | Monitoring queue, triaging alerts, escalating | Roughly 80% queue, 20% documentation | SOC Analyst |
| Tier 2 | Deep investigation, containment, tuning | Half investigation, half detection tuning | Security Analyst II |
| Tier 3 | Threat hunting, incident lead, detection engineering input | Project-driven, fewer live alerts | Senior Analyst / Threat Hunter |
Alert volume determines whether an analyst's shift feels like a conveyor belt or a detective story. A managed security provider might push thousands of alerts per day through a team; an in-house bank team of six might see fifty. Either way, the analyst's output is the same currency: clean dispositions and written evidence.
Security Engineer: The Job That Builds What the Analyst Watches
A security engineer designs, deploys, and maintains the technical controls the rest of the company depends on β firewalls, endpoint protection, identity and access rules, encryption, logging pipelines, and the automation that keeps alerts flowing.
Days split between tickets and projects. Morning might be a change ticket to tighten a firewall rule after a new application went live; afternoon could be rolling out multi-factor authentication to a business unit, testing the policy in a staging environment first so nobody gets locked out of payroll on a Friday. Engineers spend real time in infrastructure consoles and configuration tools.
Engineers also write the glue code. Pulling logs from three vendors into one pipeline, deduplicating alerts, and auto-enriching them with asset data is engineering work, and it determines whether an analyst sees one useful alert or forty useless ones. That's the hidden link between this role and the analyst role β bad engineering shows up as analyst burnout.
The failure mode is the outage. A misconfigured group policy or an aggressive endpoint rule can brick laptops or block a revenue system, and the incident review afterward is unkind. The mitigation is the same everywhere: test in a pilot group, schedule a rollback window, and document the change so the on-call engineer at 2 a.m. knows what changed.
Where the role sits depends on the employer:
| Employer type | Security engineering focus | Typical constraint |
|---|---|---|
| Bank or insurer | Compliance-driven controls, audit evidence | Change windows, regulator review |
| SaaS company | Cloud identity, infrastructure-as-code guardrails | Developer velocity, on-call rotation |
| Hospital system | Medical device segmentation, legacy patching | Clinical uptime, vendor lock |
| Government agency | Baseline hardening, accreditation | Procurement cycles |
The engineer's visible output is usually invisible when working: nothing breaks, nobody notices. That's the point, and it's also why the role suits people who like fixing systems more than chasing incidents. If you want variety and firefighting, analyst work delivers more of it.
Penetration Tester: Authorized Attacks and Reports That Get Read
A penetration tester gets paid to break into systems on purpose, with a signed scope document that states exactly which systems, when, and how far the test may go β and then writes a report that tells the client how to fix what was found.
Engagements follow a rhythm. Reconnaissance maps the target's public footprint and internal structure; scanning and enumeration find candidate weaknesses; exploitation attempts confirm whether a weakness is real; post-exploitation shows how far an attacker could travel if they got in. On a typical web application test, that might mean hunting for an injection flaw, confirming it against a staging copy of the app, and documenting the exact request that exposed it.
The report is the deliverable, not the break-in. Clients don't pay for the trophy; they pay for the remediation path β a finding phrased as "attacker can read other users' invoices by changing an ID in this request" with a severity rating, a reproduction path, and a recommended fix. Junior testers often underestimate how much of the job is writing clearly under deadline.
The edge scenario that trips teams up is scope drift. Testing a production environment during business hours can knock over a fragile service, and testing an out-of-scope host β even accidentally β can void the engagement and create legal exposure. Good testers confirm the rules of engagement in writing before a single packet goes anywhere, including a named emergency contact on the client side.
Common engagement types vary by how much the tester knows:
| Test type | What the tester knows before starting | Typical duration | Client signal |
|---|---|---|---|
| Black box | Nothing beyond the target's public presence | 1β3 weeks | Realistic outsider view |
| Gray box | Some credentials or documentation | 1β2 weeks | Faster coverage, less realism |
| White box | Full architecture and source access | 2β4 weeks | Deepest findings |
| Red team | An objective, with stealth required | 4β12 weeks | Tests detection, not just defenses |
Pen testing suits people who like puzzles with receipts. The hours can be long near a deadline, and the travel varies by whether you work at a consultancy, an in-house product company, or as an independent contractor. Note the legal boundary: unauthorized testing is a crime in most jurisdictions, and the signed authorization is what separates the profession from intrusion.
What Does a Person Actually Do All Day in This Short Statement
The one-sentence version the query asks for: a cyber security professional monitors, builds, or tests, and then writes down what happened so someone else can act on it. Everything else is detail β and the "writes it down" part is the detail most beginners skip.
Across all three roles, a person's day breaks into five activities:
1: Watching β reviewing alerts, dashboards, or test output for anything outside the norm. 2: Investigating β chasing a signal until it's explained or escalated. 3: Changing β configuring a control, deploying a fix, or executing a test. 4: Documenting β tickets, timelines, remediation advice, change records. 5: Talking β to developers, employees, clients, or management, all with different vocabularies.
Time splits vary by role, and this table is a rough planning estimate, not a measured figure from a study:
| Activity | Analyst | Engineer | Pen tester |
|---|---|---|---|
| Watching and triaging | Highest | Rare | Rare |
| Investigating | High | Occasional | High during engagements |
| Changing and building | Low | Highest | Occasional |
| Documenting and reporting | High | Moderate | Highest near deadlines |
| Talking to stakeholders | High | High | High |
Notice that documentation appears near the top for two of the three roles. A person who dislikes writing is going to have a hard time in analyst or pen testing work, no matter how much they enjoy the technical side.
The practical way to test fit: ask each interviewer how much of the week is writing versus hands-on work, and whether their team keeps a tracked backlog of tuning and detection ideas. Teams that say "we mostly close alerts" are running a Tier 1 machine; teams that say "we tune weekly" are giving analysts somewhere to grow.
The Skill Stack Each Role Demands, From Networking to Reporting
Skills map closely to daily duties, and the overlap across the three roles is smaller than most career guides suggest. Networking fundamentals β TCP/IP, ports, DNS β sit under all of them, but what you do with that knowledge differs.
For the analyst, the core skills are log reading, triage discipline, and clear escalation writing. Familiarity with a SIEM query language, a ticketing system, and MITRE ATT&CK technique names helps because alerts and investigation notes get tagged against them. Comfort with identity and email investigation matters more than most people expect, since phishing and suspicious sign-ins drive a large share of queue volume.
The engineer needs systems knowledge: operating system hardening, cloud identity and networking, scripting in Python or PowerShell, and increasingly infrastructure-as-code. A useful check when job hunting is to ask how much of the team's control against a specific threat is automated versus manual β if the answer is "we script it," the role is engineering, and if it's "we have a checklist," it may be closer to compliance work.
Pen testers need exploitation tooling fluency, web and network protocol knowledge, and report writing that a developer can read without a translator. Certifications are common entry signals β CompTIA Security+ for fundamentals, OffSec's OSCP as a practical benchmark for testing, and GIAC or ISC2 credentials for defense roles β but they cost money and time, and the certification is not the skill.
What none of the three should skip is business context. A finding that a server runs an outdated service matters more if that server processes payments and less if it's an internal lab box, and the ranking of severity is the difference between a report that gets fixed and one that gets filed.
Hiring Reality: Titles, Certifications, and What Actually Gets People Paid
Titles vary more than duties, and the first job is often the hardest to land. Entry paths include a help desk or IT support role that touches security, a SOC Tier 1 seat at a managed security provider, or an internship that turns permanent. Pure "entry-level pen tester" jobs are rare; most testers spend a year or two in a related role first.
Certifications function as filters, not proof. Security+ clears HR gates at many government-adjacent employers, and a CISSP is common for senior roles, but hiring managers will still ask what you've built, triaged, or tested. Practical portfolios β a homelab SIEM you've tuned, a documented vulnerability on a practice range like Hack The Box or TryHackMe, a public write-up β usually beat a certificate wall in interviews.
Employer type drives pay and pace as much as title does, and public-sector or regulated-industry roles tend to have more defined tiers and slower promotion rhythm than startups. The BLS median of $124,740 covers information security analysts; engineer and pen testing roles often sit at or above it, but title inflation means the median is a weak anchor β compare the duties, not the label.
The repeatable answer to "how do I get in" is: get any role that touches systems or incidents, then specialize. Help desk plus a home lab plus a few certifications is a slower path than a SOC job with a good mentor, and the mentor is a variable worth weighing when choosing offers.
Common Mistakes, Scope Failures, and the Cost of Getting One Detail Wrong
The mistakes that actually hurt cluster around scope and documentation. In pen testing, testing beyond the authorized target β or failing to confirm a production host is a mirror rather than the live system β can break something and create legal exposure. In engineering, shipping a control change during peak hours without a rollback plan turns a security improvement into an outage.
In analyst work, the riskiest mistake is closing too fast. False negatives β real compromises marked benign β are often invisible until the next incident review, and the documented disposition is what the team later measures. Someone who escalates too much gets coached on tuning; someone who escalates too little creates audit findings.
Cross-team failures are common too. A pen test finding that never reaches the engineering backlog is functionally a report, not a fix. A detection rule an analyst requests that never gets built is noise the team keeps drowning in. The fix is a tracked work item with an owner, not a chat message.
Edge scenario worth knowing: insider risk. A tester given credentials must treat the client's data as confidential, not just the findings, and an analyst who suspects an employee account is misused should follow the established process rather than investigate independently. That's the line between security work and workplace surveillance, and it's defined by company policy and, in many jurisdictions, employment and privacy law.
Career Path: How Each Role Grows, and Where the Jobs Sit by Region
Growth paths differ by starting role. Analysts often move to threat hunting, incident response, or detection engineering; engineers move toward security architecture, cloud security, or platform engineering; pen testers move to red teaming, application security, or independent consulting. Lateral moves are common, and the liberal-arts reality is that few people stay in the same lane for a decade.
Geography changes the opportunities more than any individual skill does. Washington, D.C. area, Colorado, and Texas consistently host a large share of U.S. openings, and U.S., U.K., and Australia government agencies each maintain their own qualification frameworks and vetting β a candidate moving between countries should check the local accreditation body's requirements before assuming a certificate transfers directly.
Pair a successful first year with visible output, not just tenure. An analyst who cuts false positives on one rule family and documents the change has something to discuss at review; a tester who runs engagements to deadline with clean reports builds a client-facing reputation. Those are the details that move people into senior tiers.
FAQ
Are cyber security jobs all night-shift alert watching?
No β 24/7 availability demands night shifts in most organizations, but the majority of day-to-day roles run office hours. A company with a 24Γ7 security operations center rotates shifts or hires follow-the-sun coverage, while many internal security teams work standard business hours and outsource overnight monitoring to a managed provider.
Do you need a computer science degree to become a penetration tester?
Not legally, and not in practice β job postings frequently accept equivalent experience or a practical certification. That said, degree requirements vary widely by employer and country, so confirm the standard at your target companies before assuming a portfolio alone will clear a screening filter.
Why does a low-severity pen test finding matter if it's marked "low"?
Severity ratings are context-sensitive; a low finding on an internal test box can be critical if the same flaw exists on a payment system. Ask for the finding's exposure context and a reproduction path, and rank remediation by the data and access the vulnerable system actually touches.
Which of the three jobs pays the most, analyst, engineer, or pen tester?
There's no fixed order β it depends on employer type, seniority, and location, not job family. The BLS median of $124,740 for information security analysts is a broad reference point; compare specific postings at your target employers rather than assuming one family always pays more.
What's the biggest misconception about cyber security work?
That it's mostly movie-style hacking β most time goes to reading logs, checking configurations, and writing findings. The technical thrill exists, but it's a minority of the week, and someone expecting constant break-ins will find the documentation side of the job surprisingly large.
How much of the role is writing versus hands-on technical work?
More than newcomers expect, especially for analysts and pen testers, where tickets, timelines, and reports are the deliverable. A useful rule is to ask at interview how a team tracks fixes and detection tuning; if there's no backlog, the writing side may be healthier than the technical side.
Can insider risk be part of these roles?
It can appear, but it should be handled through established process rather than ad hoc investigation. Company policy and employment law determine what an analyst may do with employee activity data, and independent snooping creates legal exposure for the employer and the employee.
This article is general educational information about cyber security careers, not career counseling, legal, or employment advice for your specific situation; requirements and regulations vary by country, employer, and role, so confirm details with a qualified local advisor, the relevant accreditation body, or the specific employer before making decisions. No outcome, salary, or hiring guarantee is implied.