LearnCyrenza

4.3

Walkthrough — Financial Analyst

45 min

Meet David, FP&A Analyst at a $200M manufacturing company

  • Prepares monthly financial reports for executive team
  • Tasks: variance analysis, forecasting, ad-hoc analyses, board deck preparation
  • Pain points: explaining budget variances, translating financial data into narrative, creating multiple scenario models

The Before State

  • Spends 2-3 days per month writing narrative explanations of financial results
  • Executives ask follow-up questions that require re-running analyses
  • Presentation formatting takes longer than the actual analysis
  • Historical context gets lost (hard to remember why Q2 2023 had that spike)

Knowledge Base Setup

  • Folder structure: FinancialReporting/[Year]/[Period]/[Actuals|Budget|Variance|Narratives|Presentations]
  • File naming: Period_ReportType_Version_Date (e.g., 2024-Q1_VarianceAnalysis_v3_2024-04-15)
  • Context documents:
    • Business model overview (revenue drivers, cost structure, seasonality)
    • Monthly financial narrative log (explaining major variances and one-time items)
    • Executive reporting preferences (what they care about, how they like data presented)
    • Historical context notes (major events, strategic initiatives, timing issues)

Tool Integration Choices

  • Excel with AI-assisted formula building and data cleaning
  • Connected AI that reads from Excel but drafts in separate platform
  • No direct ERP integration (finance system security restrictions)
  • PowerPoint with AI for slide content (not design)

Three Core Workflows

Workflow 1: Monthly Variance Analysis Report

  1. Export actuals vs. budget from ERP to Excel
  2. Calculate variances, identify material differences (>5% or >$50k)
  3. Use AI to draft initial variance explanations: "Explain this variance table [paste]. Context: [link to business model doc]. Focus on: revenue miss in Product Line B, cost overrun in Manufacturing."
  4. Review AI draft, verify against source data
  5. Add business context AI couldn't know (timing of specific orders, one-time events)
  6. Use AI to suggest implications: "Given these variances, what questions might executives ask?"
  7. Prepare response talking points
  8. Final review, format in standard template

Workflow 2: Building Financial Forecast Scenarios

  1. Executive asks for scenario planning: "What if revenue drops 10%?"
  2. Update Excel forecast model with new assumptions
  3. Use AI to check logic: "Review this forecast formula [paste]. Does it correctly account for: fixed vs. variable costs, timing of layoffs, severance costs?"
  4. Run scenario, export results
  5. Use AI to draft narrative: "Compare these three scenarios [paste tables]. Summarize: cash flow impact, breakeven timeline, key risks for each scenario."
  6. Review AI summary, add strategic implications
  7. Create executive summary slide deck
  8. Present to CFO for feedback

Workflow 3: Ad-Hoc Analysis Request

  1. CEO asks during meeting: "Why is our gross margin down 2 points YoY?"
  2. Pull relevant data from ERP and historical files
  3. Use AI for initial analysis: "Analyze this gross margin data [paste]. Break down variance by: volume, price, product mix, input costs."
  4. Verify AI's math and logic
  5. Dig deeper into surprising findings (AI may identify patterns you didn't expect)
  6. Use AI to draft response email: "Draft a concise explanation of 2-point gross margin decline. Audience: CEO. Tone: direct, data-driven, actionable."
  7. Add chart or visual if needed
  8. Review and send within 2 hours of request

Quality Control Checklist

  • Does the AI's variance explanation match the actual numbers in my Excel file?
  • Have I verified the AI didn't confuse favorable vs. unfavorable variances?
  • Are the scenarios mathematically consistent (do the assumptions flow through correctly)?
  • Does the narrative make sense to someone who doesn't live in this data daily?
  • Have I clearly separated facts from AI-generated interpretations?
  • Would I bet my credibility on this analysis?

The After State

  • Saves ~10 hours per month on report writing and scenario narratives
  • Can respond to ad-hoc requests 3x faster
  • Executive feedback: reports are clearer, more actionable
  • Still manually builds financial models (AI not reliable for complex Excel logic)
  • Historical context log prevents repeating explanations (AI can reference past variances)

Common Mistakes for Financial Analysts

  • Not verifying AI's math (it can misinterpret financial formulas)
  • Letting AI write conclusions about financial performance (analyst judgment required)
  • Using AI-generated forecasts without validating assumptions
  • Not maintaining historical context documents (AI can't remember prior periods)
  • Assuming AI understands company-specific accounting policies