Saturday takes the most money and costs the most to open. Once you price the staffing, the security and the stock against each night separately, the ranking usually changes — and so does where you should be spending effort.

Saturday took 1.2 million. Tuesday took 320,000.

Saturday is obviously the better night. Everyone knows it, the rota reflects it, and the marketing budget follows it.

Then you subtract what each night cost to open, and the gap narrows enough to be worth a second look. Sometimes it closes entirely.

The short version

Compare nights on what is left after the cost of opening, not on takings
Saturday's costs scale with the room: staff, security, stock, breakages
A quiet night with a thin roster can out-earn a busy night per unit of effort
The useful ranking is contribution per night, and contribution per entertainer hour
Most venues are over-invested in their best night and under-invested in their second best

Why takings mislead

Takings are the top line. Every night has a different amount underneath it.

Saturday	Tuesday
Full roster	Half roster
Two security	One, or none
Full bar staff	One behind the bar
Highest stock draw and breakage	Low
Highest promoter and comp load	Usually none

⚠️ The night with the biggest number on top also has the biggest number underneath. Comparing only the tops is comparing two different businesses.

One — work out the contribution per night

Take four weeks and do this per night of the week.

Takings
  less  entertainer earnings paid out
  less  bar and floor staff wages
  less  security
  less  stock used
  less  promoter and comp cost
= contribution

Saturday …1,200,000 takings
   entertainers 420,000 / staff 160,000 / security 60,000
   stock 180,000 / promoter 90,000
   ⇒ contribution 290,000

Tuesday …320,000 takings
   entertainers 110,000 / staff 45,000 / security 0
   stock 48,000 / promoter 0
   ⇒ contribution 117,000

⇒ Saturday is still ahead — but it is 2.5 times Tuesday, not 3.75 times.

Now per unit of effort

Saturday …8 entertainers × 6 hours = 48 entertainer hours
   ⇒ 6,040 contribution per entertainer hour

Tuesday …3 entertainers × 6 hours = 18 entertainer hours
   ⇒ 6,500 contribution per entertainer hour

⇒ Tuesday is the better night per hour worked. That changes what you do with your effort, not whether you open on Saturday.

Two — what this actually changes

Nobody closes on Saturday because of this. Four other things change:

One   where you put marketing money
Two   which night gets the new entertainer trial
Three which night you extend hours on
Four  which night you protect when you are short-staffed

⚠️ The instinct is to pour everything into the biggest night. But the biggest night is usually the one closest to its ceiling — the room is already full and the roster already deep.

The question that reframes it

If I added one more entertainer, which night would return the most?

⇒ Almost never the night that is already full.

Three — the ceiling test

A night can be busy and still have room, or busy and completely capped.

Ask, for each night:
  Did we turn anyone away?
  Were there empty seats after 11pm?
  Did anyone wait more than 15 minutes for an entertainer?

Turned people away, no empty seats ⇒ at the ceiling. More marketing is wasted
Empty seats after 11pm             ⇒ room to grow. Marketing can work

⚠️ Spending marketing money on a capped night raises your costs and nothing else. It is the most common way a venue works harder for the same money.

Four — the night that is quietly losing money

Run the same arithmetic and look for a negative or near-zero contribution.

Thursday …480,000 takings
   entertainers 200,000 / staff 90,000 / security 40,000
   stock 85,000 / promoter 70,000
   ⇒ contribution −5,000

⇒ A night that takes 480,000 and contributes nothing. It looks respectable from the top line and is pure cost.

What to do with it

One  cut the promoter deal and see what remains
Two  cut the roster by two and see whether takings move proportionally
Three shorten the hours before changing anything else

⚠️ Change one thing and wait four weeks. Changing three at once means you will never know which one mattered.

Five — seasonality will lie to you

Four weeks is enough to compare nights against each other. It is not enough to say a night is improving.

Comparing Tuesday to Saturday over four weeks …fine
Saying "Tuesdays are up" from four weeks       …not reliable

⇒ For trend, compare the same night against the same night twelve weeks back, and know your own seasonal shape before reading anything into it.

Six — the numbers to keep

One   takings per night of week
Two   entertainer earnings paid out, per night
Three staff, security and stock cost, per night
Four  entertainer hours worked, per night

⚠️ Four is the one that is never recorded and it is the denominator for the only ranking that matters.

A sheet for the office

────────────────────────────────────
  Which night is actually best?

  Per night, over four weeks:
    takings …………
    − entertainer payouts …………
    − staff …………  − security …………
    − stock …………  − promoter/comps …………
    = contribution …………

    entertainer hours …………
    contribution ÷ hours …………

  Then ask: did we turn anyone away that night?
  If no → there is room. If yes → it is capped.
────────────────────────────────────

Common objections

"Saturday pays for the week"

⇒ Possibly. Run it and see. Often Saturday pays for Saturday and the midweek pays for the building.

"We can't split stock by night"

⇒ Approximate it from takings share for now. It does not need to be exact to change the ranking.

"Tuesday can never be as big as Saturday"

⚠️ Nobody said it should be. The question is where the next hour of effort returns most, not which night is biggest.

"Our costs are fixed, they don't vary by night"

⇒ Rent does not vary. Almost everything else does. Fixed costs sit outside this calculation entirely — they are paid by the total, not by any one night.

"This is a lot of work"

⇒ Four weeks, once. It will change your rota and your marketing for the next year.

This week

One   take four weeks of takings, split by night of week
Two   add entertainer payouts, staff, security, stock and promoter cost per night
Three work out contribution per night
Four  count entertainer hours per night and divide
Five  for the top two nights, ask whether you turned anyone away

Doing it by hand, or on a screen

The first run is an evening with the till data and the rota. Keeping it monthly is where it dies.

By hand …takings by night, rota by night, payouts by night, assembled manually
On screen …takings and entertainer payouts already attached to the night,
           with hours worked from the shift records

⇒ tasteck holds takings and entertainer payouts against the same night as the shift records, so contribution per entertainer hour can be worked without assembling three sources by hand.

⚠️ Staff wages, security and stock usually live outside the system. Those you will still be adding yourself, and the answer is only as good as those three numbers.

Wrapping up

Rank nights by contribution, not takings
Then rank again by contribution per entertainer hour — it often reorders them
The biggest night is usually closest to its ceiling, so extra effort returns least there
Look for the respectable-looking night with near-zero contribution
Four weeks compares nights to each other; twelve is needed for trend
The missing number is nearly always entertainer hours worked

The best night is not the one that takes the most. It is the one that gives back the most for what you put into it.

Your Best Night Is Probably Not Saturday