Financial Reports & Profit Analysis
Most restaurant owners are good operators but uneasy financial readers. The P&L arrives monthly, you glance at the bottom line, file it, and keep going. AI changes the relationship. You can paste in your numbers and have a thoughtful CFO sitting across from you in seconds, ready to explain what's strange and what to do about it.
What You'll Learn
- How to use AI to interpret your weekly and monthly P&L
- How to spot food cost, labor cost, and prime cost issues quickly
- How to model "what if" scenarios on price changes, menu cuts, and labor
- How to prepare for tough conversations with your accountant or investor
The Weekly P&L Read
A 5-minute Monday morning ritual. Pull your weekly P&L from your accounting platform (Restaurant365, MarginEdge, QuickBooks). Paste:
Act as my restaurant CFO.
Below is last week's P&L for my 40-seat Italian
restaurant in Park Slope. Average weekly sales over
the prior 8 weeks: $34,800.
[paste this week's P&L]
Tell me:
1. Three things that look notable (good or bad)
2. Any line item more than 15% off the 8-week
average — call it out specifically with $ and %
3. The single biggest issue I should investigate
today
4. The single biggest opportunity I might be missing
5. A 4-bullet summary I can text to my partner
Tone: direct, short, no hedging.
You'll have a CFO-level read in 20 seconds. Most owners get 2–3 actionable insights per week from this prompt alone.
Prime Cost Analysis
Prime cost (food + labor as a % of sales) is the single most important number in a restaurant. Healthy independents target 60–62%. Above 65% and the math stops working.
Act as my prime cost analyst.
8 weeks of weekly data:
[paste sales, food cost $, labor cost $]
For each week:
- Food cost %
- Labor cost %
- Prime cost %
Then:
- Trend over the 8 weeks (improving / declining)
- The single biggest swing factor
- Whether the issue is variable (cost of goods) or
fixed (over-scheduled labor)
- One specific recommendation to bring prime cost
toward 60%
A real CFO would charge you $400 for this analysis. AI does it in 30 seconds and gets it right.
Spotting Theft, Variance, and Mistakes
Variance — the difference between what your inventory says you sold and what your POS rang up — is where shrinkage hides. AI is great at flagging it.
[paste house context]
Below is theoretical food cost (based on POS sales
and recipes) vs actual food cost (based on weekly
inventory) for the last 6 weeks.
[paste data]
Calculate variance per week. Flag any week with
variance over 1.5%. For weeks above the threshold:
- Suggest the most likely root cause (over-portioning,
comps not entered, theft, waste, broken recipe spec,
inventory count error)
- Suggest the single best diagnostic step
This single weekly habit catches a surprising amount of shrinkage. Don't skip it.
"What If" Scenario Modeling
Owners avoid scenario modeling because it's slow in spreadsheets. AI does it in conversation:
[paste house context]
Current state: weekly sales $34K, food cost 31%, labor
30%, fixed costs $4,200.
Model these scenarios:
Scenario A: Raise menu prices 4% across the board.
Assume 5% drop in covers.
Scenario B: Cut 10 hours/week of FOH labor.
Assume no change in covers. 1% NPS-style risk.
Scenario C: Add a $42 prix-fixe Mon–Wed.
Assume it captures 15% of those nights' covers and
lifts average check from $48 to $52.
For each scenario, show:
- New weekly sales
- New food cost % and $
- New labor cost % and $
- New net profit $
- Notable risks
End with: which scenario produces the highest
durable profit increase?
You just ran three financial models in 90 seconds. Pick the winner. Plan the next 30 days around it.
Preparing for the Accountant Meeting
Once a quarter you sit with your accountant. Most owners go in cold. Show up with this:
Act as my accountant prep coach.
Below is the last 12 months of P&Ls (paste).
Build me a one-pager for my quarterly meeting:
- Top 3 trends to highlight (with $ context)
- Top 3 risks (with $ context)
- 3 questions I should ask my accountant
- 1 tax planning angle to discuss
- 1 question that will surprise her into respecting
me as a serious operator
Walk in prepared. Walk out with better advice.
Daypart and Channel Profitability
Most operators don't know which dayparts and channels actually make money:
Below is 30 days of sales by daypart and channel:
brunch / lunch / happy-hour / dinner / late-night
across in-house / takeout / delivery / catering.
[paste data]
For each daypart-channel combo, compute:
- Total revenue
- Estimated contribution margin (food cost + variable
labor)
Rank by contribution. Tell me:
- Which combos to grow
- Which combos to cut, reduce hours, or de-prioritize
- One specific action to take this week
You may discover delivery is barely break-even, or that brunch is your best margin hour. Either insight is worth thousands.
What You Should NOT Paste Into AI
AI prompts about your business are sensitive. Don't paste:
- Bank account numbers
- Employee SSNs or full payroll details
- Investor cap tables with names
- Identifiable customer data
Aggregate first. AI works fine on summarized data — it doesn't need the raw payroll register.
A Friday Afternoon Financial Workout — 30 Minutes
- Weekly P&L read → 4 minutes
- Prime cost trend prompt → 4 minutes
- Variance flag prompt → 5 minutes
- One scenario model on a planned change → 6 minutes
- Daypart profitability check → 5 minutes
- Note 2–3 actions to take next week → 6 minutes
The most expensive consultant you'll never hire.
Key Takeaways
- A weekly 5-minute "P&L read" prompt surfaces 2–3 actionable insights every week
- Run prime cost trends every 8 weeks; target 60–62% for an independent
- Use AI to flag food cost variance; investigate any week over 1.5%
- Scenario model price, menu, and labor changes before committing
- Walk into your accountant meeting with an AI-prepped one-pager — get better advice