githubEdit

Account Usage Investigation Workflow & Cheatsheet

Windows Enterprise DFIR - SOC Analyst Reference


🎯 Investigation Objectives

When investigating account usage, determine:

  • WHO: Which accounts were used (local vs domain)

  • WHEN: Timeline of authentication events

  • WHERE: Source and destination systems

  • HOW: Authentication method and logon type

  • WHAT: Actions performed and resources accessed

  • WHY: Legitimate business need or suspicious activity


πŸ“‹ Quick Reference: Critical Event IDs

Authentication Events (Security.evtx)

Event ID
Description
Protocol
Priority

4624

Successful Logon

Both

πŸ”΄ Critical

4625

Failed Logon

Both

πŸ”΄ Critical

4776

Credential Validation

NTLM

🟠 High

4768

TGT Granted

Kerberos

🟠 High

4769

Service Ticket Requested

Kerberos

🟑 Medium

4771

Pre-auth Failed

Kerberos

πŸ”΄ Critical

4634/4647

Logoff

Both

🟒 Low

4648

Explicit Credentials (runas)

Both

πŸ”΄ Critical

4672

Admin Rights Logon

Both

πŸ”΄ Critical

4778

RDP Session Reconnect

N/A

🟠 High

4779

RDP Session Disconnect

N/A

🟑 Medium

4720

Account Created

N/A

πŸ”΄ Critical

4697

Service Installed

N/A

πŸ”΄ Critical

Service Events (System.evtx)

Event ID
Description
Priority

7045

Service Installed

πŸ”΄ Critical

7034

Service Crashed

🟠 High

7036

Service Start/Stop

🟑 Medium

7040

Service Startup Changed

🟠 High


πŸ” Investigation Workflow

Phase 1: Initial Triage (First 15 Minutes)

Step 1.1: Identify the Scope

Step 1.2: Quick Account Profiling

Document:

  • Account type (Local/Domain/Cloud)

  • Account status (Active/Disabled/Locked)

  • Group memberships

  • Account age and last password change


Phase 2: Authentication Analysis (30 Minutes)

Step 2.1: Collect Authentication Events

On Workstation (Local Auth):

On Domain Controller (Domain Auth):

Step 2.2: Analyse Logon Types

Logon Type Decision Tree:

Red Flags:

  • Type 8 (cleartext credentials)

  • Type 10 from unusual IPs/countries

  • Type 3 during off-hours to sensitive servers

  • Multiple Type 4625 (failed logons) followed by Type 4624 (brute force)

  • Type 9 with service accounts


Phase 3: Timeline Construction (45 Minutes)

Step 3.1: Build Authentication Timeline

PowerShell Timeline Script:

Step 3.2: Correlate with Other Activity

Check for:


Phase 4: Deep Dive Analysis (1-2 Hours)

Step 4.1: RDP Investigation

If Type 10 logons detected:

Artifact Collection:

Step 4.2: Registry Analysis

SAM Hive Analysis:

Look for:

  • Last login timestamps

  • Password last set dates

  • Login counts (high counts = automation/service account)

  • Failed login attempts

  • Cloud account indicators (InternetUserName value)

  • Unusual RIDs or account creation times

NTUSER.DAT Analysis:

Check for:

  • Recent documents accessed

  • Typed URLs (web activity)

  • UserAssist (program execution)

  • Run/RunOnce keys (persistence)

  • MRU lists (recently used files)

  • WordWheelQuery (search terms)

Step 4.3: Service Analysis

Query Service Events:

Red Flags:

  • Services installed during suspicious logon timeframe

  • Service names with random characters

  • Services running from temp directories

  • Services with unusual account contexts

  • Services that crash immediately after suspicious activity

Step 4.4: User Access Logging (Server Only)

For Windows Server 2012+:

Extract:

  • Source IP addresses

  • Accessed services

  • Access timestamps

  • Total access counts

  • Authentication types

  • User accounts used


Phase 5: Pattern Analysis (30 Minutes)

Step 5.1: Identify Anomalies

Statistical Analysis:

Step 5.2: Check for Attack Indicators

Common Attack Patterns:

Attack Type
Indicators

Password Spray

Multiple accounts, few failed attempts each, Type 3

Brute Force

Single account, many 4625 events, then 4624

Pass-the-Hash

Type 3 logons, NTLM auth, no Type 2 on source

Pass-the-Ticket

Kerberos auth without initial 4768, unusual SPNs

Golden Ticket

Long ticket lifetimes, unusual encryption types

Lateral Movement

Type 3 chain across multiple systems

Privilege Escalation

4672 events, Type 9 logons, new admin access

Persistence

Service installs (7045), scheduled tasks, Run keys

RDP Hijacking

4778 without preceding 4624, session transfers


Phase 6: Lateral Movement Tracking (1 Hour)

Step 6.1: Map Authentication Chain

Build Network Map:

PowerShell Lateral Movement Detector:

Step 6.2: Correlate with Process Execution

Check what was executed after authentication:

Red Flags:

  • PowerShell execution immediately after Type 3 logon

  • cmd.exe with suspicious command lines

  • psexec, wmic, mmc, sc.exe usage

  • Mimikatz or other credential dumping tools

  • Remote management tools (TeamViewer, AnyDesk)


πŸ› οΈ Tool Quick Reference

Built-in Windows Tools

Registry Analysis Tools

RDP Artifact Analysis

UAL Analysis (Server)

SAM Analysis


πŸ“Š Investigation Checklist

Initial Assessment

  • [ ] Identify affected account(s)

  • [ ] Determine account type (Local/Domain/Cloud)

  • [ ] Verify current account status

  • [ ] Establish investigation timeframe

  • [ ] Identify affected systems

Data Collection

  • [ ] Security.evtx from affected workstation

  • [ ] Security.evtx from Domain Controller

  • [ ] System.evtx from affected systems

  • [ ] Terminal Services logs (if RDP used)

  • [ ] SAM registry hive

  • [ ] NTUSER.DAT from user profile

  • [ ] RDP Bitmap Cache (if applicable)

  • [ ] UAL databases (if server)

  • [ ] Network traffic logs

  • [ ] EDR/AV logs

Authentication Analysis

  • [ ] Timeline of 4624/4625 events

  • [ ] Analyse logon types distribution

  • [ ] Identify source IPs/hostnames

  • [ ] Check for failed logon patterns

  • [ ] Verify authentication protocols used

  • [ ] Review explicit credential usage (4648)

  • [ ] Check for privilege escalation (4672)

Artifact Analysis

  • [ ] RDP session artifacts reviewed

  • [ ] Registry analysis completed

  • [ ] Service events examined

  • [ ] Process execution correlated

  • [ ] File access patterns checked

  • [ ] Network connections mapped

Lateral Movement

  • [ ] Authentication chain mapped

  • [ ] Pivot points identified

  • [ ] Affected systems documented

  • [ ] Attack timeline constructed

Pattern Analysis

  • [ ] Baseline behaviour established

  • [ ] Anomalies identified

  • [ ] Attack patterns matched

  • [ ] IOCs extracted

  • [ ] Risk assessment completed

Documentation

  • [ ] Timeline created

  • [ ] Evidence preserved

  • [ ] Screenshots captured

  • [ ] IOCs documented

  • [ ] Report prepared


🚨 Quick Win: High-Value Queries

Detect Potential Compromise

1. Find after-hours admin logons:

2. Detect password spray attempts:

3. Find Type 10 (RDP) logons from external IPs:

4. Identify explicit credential usage (runas):

5. Find service installations during suspicious timeframe:


πŸŽ“ Pro Tips

Efficiency Tips

  1. Use FilterHashtable instead of Where-Object for faster queries

  2. Narrow timeframes - don't query entire logs if you know the window

  3. Query remote systems in parallel using PowerShell jobs

  4. Export to CSV for analysis in Excel/Timeline Explorer

  5. Use date math: (Get-Date).AddDays(-7) for relative dates

Analysis Tips

  1. Start broad, then narrow - overview first, deep dive on anomalies

  2. Follow the data - let artifacts guide your investigation

  3. Trust but verify - logs can be cleared/modified

  4. Look for absence - missing logs are suspicious

  5. Context matters - one odd event might be normal, patterns aren't

Documentation Tips

  1. Screenshot everything - you may need it for reports

  2. Note your commands - reproducibility is critical

  3. Preserve original evidence - work on copies

  4. Chain of custody - document who, what, when, where

  5. Timeline format - use ISO 8601 (YYYY-MM-DD HH:MM:SS)

Common Pitfalls

  1. ❌ Only checking Security log (also check System, Application, specialised logs)

  2. ❌ Ignoring logon type (Type 3 vs Type 10 context is critical)

  3. ❌ Not checking Domain Controller (domain auth happens there)

  4. ❌ Forgetting about log rotation (events may be archived)

  5. ❌ Tunnel vision on one indicator (look for corroborating evidence)


πŸ“ˆ Escalation Criteria

Escalate Immediately If:

  • βœ… Admin account compromise confirmed

  • βœ… Domain Controller authentication anomalies

  • βœ… Evidence of credential dumping tools

  • βœ… Lateral movement to multiple critical systems

  • βœ… After-hours access to sensitive data repositories

  • βœ… Service account used interactively

  • βœ… Cloud admin account suspicious activity

  • βœ… Evidence of golden ticket or similar advanced attack

  • βœ… Data exfiltration indicators

  • βœ… Ransomware/malware execution correlated with logon


πŸ“š Additional Resources

Microsoft Documentation

  • Windows Security Log Encyclopedia

  • Advanced Security Audit Policies

  • Account Logon Events Reference

Tools

  • Eric Zimmerman Tools Suite (KAPE, RECmd, Timeline Explorer)

  • Volatility Framework (memory analysis)

  • Chainsaw (Sigma rule detection for Windows Event Logs)

  • DeepBlueCLI (PowerShell threat hunting)

Training

  • SANS FOR500 (Windows Forensics)

  • SANS FOR508 (Advanced Incident Response)

  • MITRE ATT&CK Framework (Credential Access, Lateral Movement tactics)


πŸ“‹ Report Template Structure


Remember: The best investigation is methodical, documented, and reproducible. Take your time, be thorough, and let the evidence tell the story.

Last updated