Your POS system knows more about your restaurant than you do.

Every transaction, every modifier, every void, every discount, every time-stamp — it's all sitting in a database that you probably never export. Most restaurant owners use their POS to take orders and run end-of-day reports. That's like buying a car and only using the radio.

Here are the things your POS data is trying to tell you.

1. Your real peak isn't when you think it is

Most restaurants say their peak is "around 7:30 on Fridays." But when you actually plot your transaction timestamps, the picture is usually more nuanced.

Maybe your real peak is 7:15-8:00 — a 45-minute window, not a vague "around 7:30." Maybe your Saturday peak starts 20 minutes later than Friday. Maybe your Sunday brunch has two peaks: one at 10:30 for the early crowd and another at 12:15 for the late risers with a dead valley in between.

Why this matters: Your staffing, your kitchen timing, and your table-turn strategy should be built around your actual peak, not your assumed peak. If you're scheduling your strongest line cook to start at 4pm but your real peak doesn't hit until 7:45, you're paying for 3.75 hours of underutilisation.

2. Some dishes are secretly losing you money

Revenue per dish is not the same as profit per dish. Your POS tells you what sells and for how much. Combine that with your recipe costs and you'll find:

The popular loser: a dish that sells 40 portions a week at $24 but costs you $10 to make. That's $14 margin — sounds decent until you realise your chicken dish sells 35 portions at $22 but costs you $4.50 to make. The chicken earns you $17.50 per plate. Your "best seller" is actually your fourth most profitable dish.

The hidden gem: a dish that only sells 12 portions a week but has an 82% margin. Nobody orders it because it's buried on page two of your menu. Move it to the top right of page one (where eyes go first) and watch what happens.

The loss leader you didn't mean to create: that appetiser you priced at $11 because it "feels right" but actually costs you $6.50 to plate. At 41% food cost, every table that orders it is dragging your overall margin down.

Your POS has the sales data. Your recipe cards have the costs. Put them together and you'll see your menu completely differently.

3. Your discounts are more expensive than you think

Every void, comp, discount, and manager override is logged in your POS. Most restaurants glance at the total and move on. But the pattern matters more than the total.

Questions to ask your data:

Which server gives the most discounts? Is it generosity, poor training, or are they covering mistakes?
What time do most voids happen? If it's during the rush, it's probably miscommunication between FOH and kitchen. If it's at the end of the night, someone might be gaming the system.
Which dishes get sent back most often? This is quality control data hiding in your POS. If the salmon comes back twice a week, there's a consistency problem on that station.
What's your discount rate as a percentage of revenue? Industry average is 2-3%. If you're above 4%, you have a process problem.

4. Your table turn time is costing you covers

If your POS tracks table assignments (most modern systems do), you can calculate your average table turn time: the gap between one table's first order and the next table's first order at the same seat.

The benchmark: casual dining should turn tables in 45-60 minutes at dinner. Fine dining, 90-120 minutes. If your casual restaurant is averaging 75 minutes per turn, you're losing 20% of your potential covers every night.

Where the time goes:

Pre-order dwell time: how long after being seated before they order? If it's more than 8 minutes, your service approach needs work.
Course gaps: how long between courses? If mains are landing 25 minutes after starters, your kitchen is either backed up or not firing on time.
Post-dessert linger: how long after the bill is settled before the table is cleared and resat? This is often the biggest leak.

Your POS timestamps tell you exactly where the bottleneck is. Most restaurants never look.

5. Day-of-week patterns you've never noticed

Export your daily revenue for the last six months and plot it by day of week. You'll see patterns that are invisible when you're living inside the business:

Your actual slowest day might not be the one you think. Many restaurants assume Monday is dead, but the data sometimes shows Tuesday or Wednesday is actually weaker.
Your weekend variance — is Saturday consistently 30% above Friday, or does it fluctuate wildly? High variance means external factors (events, weather) are driving your business more than your regulars.
Seasonal shifts — does your Thursday-to-Friday ratio change in summer versus winter? Patio restaurants often see Thursday become nearly as strong as Friday in warm months.

6. Your menu changes aren't working

If you changed your menu six weeks ago, your POS can tell you whether it worked:

Did average cheque go up or down?
Did your sales mix shift toward higher-margin dishes?
Did any new dishes underperform expectations?
Did removing a dish cause guests to trade down to a cheaper option instead of up to the replacement?

Most restaurants change their menu based on gut feel and never measure the impact. Your POS has the answer — you just have to ask.

How to start

You don't need a data scientist. You need one export and one hour.

Export your last 90 days from your POS — transaction-level data with timestamps, dish names, and revenue
Open it in a spreadsheet (or upload it to Mise — Claude will parse it for you)
Look at three things: your actual peak times, your dish-level profitability, and your day-of-week pattern
Make one change based on what you find — move a dish on the menu, adjust a staffing schedule, or reprice one item
Measure the impact 30 days later

The data has been sitting there the whole time. The only thing missing was someone looking at it.