Skip to content

Parallel Track Investigation

Why It Matters

Sophisticated deception is consistent within a single narrative but breaks down across independent information channels. If you only investigate through financial records, fraudsters can create consistent fake records. If you only interview sources, coordinated lies can be consistent. But reconciling financial records with source testimony with regulatory filings with competitor observations surfaces contradictions that reveal deception.

The Key Move

For a single question ("Did Company X misuse investor funds?"), pursue simultaneously:

  1. Financial record analysis (bank statements, invoices, expense reports).
  2. Source interviews (employees, investors, vendors).
  3. Public records (regulatory filings, court documents, property records).
  4. Digital footprints (emails, calendar invites, social media).
  5. Expert analysis (forensic accountants, industry specialists). Don't finish one track before starting another. Work them in parallel. When they contradict - financial records claim expense X, but no one remembers it and no vendor confirms receiving payment - you've found deception. The contradiction points you toward truth.

Classic Application

Lance Armstrong doping investigation. Journalists pursued parallel tracks:

  1. Testing data and lab results.
  2. Teammate testimony.
  3. Financial records showing payments to doctors.
  4. Email and communication records.
  5. Expert analysis of performance data. Armstrong's denials were consistent until parallel tracks revealed contradictions: teammates saw doping, financial records showed suspicious payments, performance data was statistically improbable for clean athlete, and communication records showed he knew about anti-doping investigations. No single track was conclusive, but contradictions between tracks revealed the truth.

Surprising Application

User research validation. When researching user needs, parallel track:

  1. User interviews (what they say they need).
  2. Usage analytics (what they actually do).
  3. Support tickets (what breaks or confuses them).
  4. Sales call analysis (what they ask about before buying).
  5. Competitor analysis (what alternatives they consider). Contradictions are revealing: users say they want feature X, but analytics show they never use it when offered. They claim Y is essential, but support tickets show they can't figure out how to use it. Parallel tracks reveal actual needs vs. stated preferences.

Failure Modes

Starting too many tracks simultaneously and making no progress on any (focus is required - "parallel" doesn't mean "infinite"). Not documenting contradictions between tracks (they're easy to rationalize away if not explicitly noted). Assuming contradiction means someone is lying (can also indicate misunderstanding, different contexts, or measurement error). Abandoning tracks that initially seem unproductive (sometimes the contradictions appear late). Not knowing when you have enough tracks (more isn't always better - track 6 and 7 might be redundant with 1-5).

Go Deeper

Meyer, "The New Precision Journalism" on multi-method verification. Houston, "Computer-Assisted Reporting" on triangulation across data sources. ICIJ, "Paradise Papers Methodology" on parallel investigation across countries and data types.

  • Deception - Military Strategy & Strategic Thinking: A Map of the Territory