Mike The Finance Nerd

GPT-5 is here. ~8% more accurate on hard problems. Up to 45% less back-and-forth to get the right answer. For finance pros, that means faster models, cleaner data, and reports that write themselves. #GPT5 #FinanceAutomation

Excel just leveled up. ⚡️ With Agent Mode, you don’t just ask for a chart—you ask Excel to: 👉 Clean data 👉 Build the model 👉 Update the dashboard 👉 Email the report All from one prompt. Finance pros: this is your new junior analyst. #Excel #AI

Year-end reviews don’t have to be pain + coffee + copy/paste. ☕📊 Use AI to: 👉 Pull your biggest wins from emails/Teams 👉 Turn KPIs into clean charts 👉 Draft self-review narratives that don’t sound robotic Stop scrambling, walk in with receipts. #FinanceCareers #AI

How I automate Excel with AI (hands-off): 1. Save my table to OneDrive/SharePoint 2. Power Automate: for each row → call Azure OpenAI (“Standardize vendor name”) 3. Write result back as a new column 4. Run daily Outcome: clean data, zero grunt work. #PowerAutomate #Excel

I use AI + Excel in 10 min: 1. Ctrl+T (make a Table) 2. Copilot: “Pivot revenue by region by quarter” 3. Ask: “Explain it” 4. Flash Fill names (Ctrl+E) 5. Add hidden “Prompts” sheet (log inputs/outputs) Result: faster reports, fewer errors. #Excel #AI

Finance pros: Excel isn’t career insurance anymore. Data science is. • Learn SQL • Pick up Python basics • Think in datasets, not tabs You don’t need a PhD, just enough to stay relevant when CFOs start asking for predictive, not just historical. #FinanceCareers #DataScience

Stop scrolling like a maniac in Excel Add your key formulas to the Watch Window (Formulas → Watch Window). Now you can track KPIs across tabs without leaving your spot. One tiny feature. Massive sanity saver. #Excel #FinanceAutomation

I killed copy-paste consolidation with one script: 1. Add a sheet named “Consolidated” 2. Automate → New Script 3. Loop all sheets (skip “Consolidated”), append UsedRange, skip headers 4. Run 5. Analyze #Excel #OfficeScripts #FPandA

Excel is both hero & villain. I taught it to do its own grunt work with Office Scripts. Quick start (5 min): 1. Open in Excel for web 2. Automate tab 3. Record Actions: bold header, widths, currency 4. Stop → name “FormatReport” 5. Run #Excel #OfficeScripts #FinanceAutomation

AI isn’t replacing all finance jobs — it’s cutting off the bottom rung. 📉 Stanford study: entry-level workers (22–25) in AI-exposed jobs saw a 13% drop in employment since 2022. If your only value is grunt work, you’re on borrowed time. 👉 Build skills in analysis + storytelling.

Finance pros: stop re-explaining the same thing to AI every time. 👉 With Project Memory, you can teach AI your close templates, commentary style & “always flag marketing overspend” rules once… and reuse forever. ⚡ Less typing, more analysis.

Copilot hack I use for exec comms (10–15 min): 1. Paste your variance table into PowerPoint 2. Prompt: “Turn this into a 5-slide deck: Revenue, Expenses, Margin, Top 5 drivers, Next steps. Tone: concise, board-ready.” 3. Tighten headlines, add insights, ship. Reconcile first.

Copilot hack I use at close: GL dump → anomaly shortlist (2 min) 1. Open the GL in Excel (Copilot on) 2. Prompt: “Scan this export; flag anomalies >$25K by account, group by dept, suggest likely causes.” 3. Triage the table; chase only red flags. Always reconcile.

Most “slow” Power BI reports aren’t slow. They’re bloated. Cut the junk, and the model flies: ✅ Drop audit columns no one uses ✅ Move text out of facts ✅ Turn off auto date tables Small tweaks = dashboards execs actually trust.

If your Power BI dashboards are crawling, it’s not the data, it’s your model. 👉 Flatten your tables. 👉 Kill unused columns. 👉 One clean date table. I’ve taken refresh times from 90 min → 12 min with just these moves.

“Automation saved 120 hours” = nice. “Automation freed $9,000 of analyst time every quarter” = budget approved. 👉 Translate hours into $$ if you want skeptical execs to buy in.

Your automation isn’t failing because the tech is bad. It’s failing because you’re automating a mess. 👉 Apply Occam’s Razor: cut steps first, then automate. Simple > complex. Always.

You don’t need to code to be a tech leader anymore. In the AI age, it’s not about Python, it’s about knowing what to automate and why. Finance pros who master tools + strategy (not syntax) will own the next decade.

Google Sheets just got wild → you can now type =AI("Write commentary for this variance") and it’ll draft it for you. No more blank-page paralysis at month-end. Use it for variance notes, formula help, and data cleanup.

FP&A Custom GPT (copy this): 1. Upload: variance template + sample actuals vs budget CSV 2. Rules: “Give 3 drivers w/ % or $, end with 1 action. Bullets only. ≤150 words.” 3. Starters: “Draft revenue variance”, “Explain gross margin” Test -> tighten -> save & share

Build a Custom GPT for finance in 6 steps (no code): 1. ChatGPT → Explore GPTs → Create 2. Role: “Helpful financial analyst; bullets; slide ready” 3. Upload: policy PDF, templates, COA CSV 4. Starters: “Draft OpEx variance”, “Summarize lease policy” 5. Test + tighten rules Save + share