Walk into a typical independent café on a Tuesday morning, ask the owner to name their top five sellers in order, and watch what happens. They will get the flat white right. After that it gets shaky. By number four they are guessing. By number five they are guessing about an item they were sure was a hit but actually sells twice a week.

Their menu has 70 items on it. Maybe 80 if you count the modifiers. They can describe every single one in detail, where the supplier is, what the prep is, what the margin should be in theory. What they cannot tell you, with any confidence, is which 10 of those items are actually carrying the business.

This is the gap the Pareto cut closes. And it takes about 40 minutes.

The 80/20 in cafés is usually more brutal

Vilfredo Pareto noticed in 1896 that 80% of the land in Italy was owned by 20% of the population. The pattern shows up everywhere once you look. In hospitality it shows up hard.

In a typical UK independent café, the distribution is rarely a clean 80/20. It is more often:

Top 10-15 items = 70-85% of revenue
Bottom 30 items = roughly 5% of revenue but 30% of prep complexity, fridge space and ordering decisions
A handful of "ghost" items that sold twice last quarter and somehow never came off the menu

Some cafés are running closer to a 10/70 split. A few are at 8/80. The longer your menu, the more skewed it usually is, because you have padded out the bottom with items that felt important when you wrote them but never earned their keep.

This is not a problem to solve with a menu engineering matrix and a four-quadrant chart. Stars, plough horses, puzzles and dogs are useful frameworks, but they ask you to plot every item by popularity and contribution margin and then make a judgment call. That is a Sunday afternoon job, and most owners never do it.

The Pareto cut is the version you can actually do on a Tuesday morning between the breakfast rush and lunch prep.

How to actually pull the data

Every modern POS will give you a sales mix report. The names vary, the menu paths vary, but the data is there.

Square: Reports > Item Sales. Set the date range to the last 90 days, sort by Net Sales descending, export to CSV.
Lightspeed K-Series / L-Series: Reports > Sales by Item. Same idea, 90-day window.
Toast: Analytics > Menu Item Sales. Pick gross sales, 90 days, export.
Epos Now: Reporting > Product Sales Report. Filter to last quarter.
Zettle / SumUp: Sales > Items. Smaller toolset but the export still works.

Why 90 days? Because seven days is too noisy (one wet weekend tanks a category), 30 days catches a single bank holiday weirdness, and 12 months smooths over seasonal shifts you actually want to see. Ninety days is the sweet spot - long enough to be stable, short enough to reflect what your menu is doing now.

Sort by revenue, not units. This is the bit most owners get wrong on the first pass, and we will come back to why.

A worked example: 15 items, one café

Here is a fictional but very plausible Oxfordshire café. Sixty-eight items on the menu. Ninety days of POS data. Top 15 items by revenue:

Rank	Item	Units sold (90d)	Avg price	Revenue (90d)	% of total
1	Flat white	4,820	£3.80	£18,316	11.2%
2	Smashed avo brunch plate	1,140	£12.50	£14,250	8.7%
3	Cappuccino	3,210	£3.80	£12,198	7.5%
4	Full breakfast	820	£14.50	£11,890	7.3%
5	Bacon and egg roll	1,640	£6.95	£11,398	7.0%
6	Latte	2,890	£3.80	£10,982	6.7%
7	Cheese and ham toastie	1,210	£8.50	£10,285	6.3%
8	Brownie	1,950	£4.20	£8,190	5.0%
9	Sourdough toast and jam	1,310	£5.50	£7,205	4.4%
10	Americano	1,820	£3.40	£6,188	3.8%
11	Pesto chicken focaccia	580	£9.50	£5,510	3.4%
12	Mocha	1,180	£4.20	£4,956	3.0%
13	Carrot cake slice	1,120	£4.20	£4,704	2.9%
14	Granola bowl	480	£8.95	£4,296	2.6%
15	Iced latte	1,080	£3.90	£4,212	2.6%
	Top 15 total			£134,580	82.4%

Total 90-day revenue across all 68 items: roughly £163,300.

Fifteen items doing 82% of the work. The other 53 items - every variation of tea, every speciality drink, every "we should probably have a vegan option for that" plate, every cake the owner felt loyal to - share the remaining £28,720 between them. That is around £540 per item over 90 days. About £6 a day each. A quarter of those items are doing under £3 a day.

Why revenue, not units, matters

Look at that table again. By units, the flat white is the obvious king at 4,820 cups. But by revenue it sits at 11.2%, and the brunch plate at one quarter of the unit volume is doing 8.7%. The full breakfast, sold less than a sixth as often as the flat white, is at 7.3%.

This is the menu reality that unit-volume reports hide. Coffee is the soundtrack of the café. Food is the album sales. If you optimise your menu around what you sell most cups of, you end up with a brilliant coffee operation and a food menu nobody is paying attention to, even though the food is paying the rent.

A £14.50 brunch plate sold 30 times a week is doing £435. A flat white sold 30 times a week is doing £114. They demand the same level of attention from you on quality, prep and consistency, but only one of them moves the P&L.

The next layer: gross profit contribution

Revenue ranking is the right starting point. Gross profit ranking is where the real decisions get made.

A £4 flat white at roughly 50p in milk and beans contributes around £3.50 net before labour. A £4.50 brownie at £1.30 in butter, sugar, eggs and chocolate contributes £3.20 net. The flat white wins on contribution per unit, but the brownie does not lag as far behind as you might think, and it requires no barista to make it on the spot.

Now flip it. The £12.50 brunch plate sounds like a star. But the avocado is £1.10, the sourdough is 60p, the eggs are 70p, the feta is 80p, the chilli oil and herbs another 30p. You are at £3.50 in food cost before you have plated it. Net contribution is around £9, which is excellent, but it ties up your hot section for six minutes during the rush. The cake contributes £3.20 with zero kitchen time.

I covered the deeper version of this in POS data vs recipe costing: what actually makes money. The short version: rank by revenue first to find the items that matter, then rank by gross profit contribution to find which of those are actually pulling their weight, and which are just busy.

What to do with the bottom 50

You have three options for the long tail, and you need to apply one to each item.

Cut. If an item sells fewer than three times a week, has its own ingredient SKU not used elsewhere, and is not strategically essential (one vegan plate, one gluten free option), it goes. No discussion. The fridge space and prep time are worth more than the £4 a day in revenue.

Simplify. Some long-tail items are there because a customer asked once and you said yes. Roll them into a modifier, or onto a daily special board, or fold the ingredients into another dish. The "halloumi wrap" that sells four times a week probably uses the same halloumi as your brunch plate. Keep the ingredient, lose the SKU.

Feature. A small number of bottom-tail items underperform because nobody knows they exist. They are buried at the bottom of a 70-item menu in 9pt font. Move them to the top of the board, talk about them, train the team to recommend them, and they may climb. Give them 30 days. If they do not move, see option one.

I wrote the full case for shorter menus in why smaller menus make more money. The Pareto cut is the evidence base for it.

What the Pareto cut actually changes

Once you have your top 15, everything downstream gets easier.

Menu board design. The top 15 go in the prime real estate (eye level, top of the board, photographed if you do photos). The middle tier fills out the rest. The cuts come off entirely.

Prep list. Your morning prep list should be ordered by what is going to sell. If sourdough toast is item nine on revenue, your sourdough par should reflect that. If the granola bowl is item fourteen and only sells 5 a day, you need 5 portions ready, not 20.

Ordering. Your weekly ordering should be driven by the top 15. If smashed avo plates are number two, that is the supplier conversation you want to be having: better avo pricing, consistency, backup supply. Not the one-off chutney you use on the four sandwiches a week of an item ranked 47th.

Supplier negotiations. This is the underrated one. When you walk into a supplier conversation knowing you sell 1,210 toasties in 90 days and the cheese and ham represent 30% of food cost on each, you have a real number. "Can you do me better on the cheddar?" lands harder when it is followed by "I am going through 38kg a quarter on this one item."

Staff training. Your barista should be able to make your top 15 in their sleep. Calibration, consistency, presentation, allergen knowledge. If a new starter learns those 15 cold in week one, they are useful immediately. The other 50 items can come later, or never, if half of them are about to be cut.

Pricing. This is where the Pareto cut overlaps with how to set menu prices and protect margins. Your top 10 items deserve a price review every quarter. The other 50 can wait until the annual menu refresh, because a 30p price tweak on an item that sells twice a week is not worth the time to print the new label.

The menu density trap

There is a tempting argument that a long menu signals choice and abundance. In practice it signals indecision. When customers see 70 items they default to whatever they ordered last time, or whatever the person in front of them ordered, or whatever has a photo. Your bottom-tail items are not getting picked because they are buried.

Worse, a long menu dilutes attention from your winners. The brunch plate doing £14,250 in 90 days could be doing £18,000 if it had the room to breathe on the board, instead of being the second item in a brunch section of seven plates that all blur into each other.

This is the same argument that runs through the hidden cost of modifier extras in cafés. Every additional choice is a small tax on the customer's attention and your team's prep complexity. The Pareto cut tells you which choices are worth keeping in the budget.

Where Brikly fits

The whole exercise above is about 40 minutes if you know your way around a spreadsheet, plus another two hours if you want to layer gross profit contribution on top. Most owners never get to the second pass because the data is in three places (POS for sales, recipe book for costs, supplier invoices for current pricing) and joining them up is the friction.

MenuBrik pulls your POS sales mix automatically, joins it to your CostingBrik recipes and your latest supplier pricing, and ranks every item by revenue, units, and gross profit contribution in one view. It also flags the items where margin has been quietly eaten by ingredient cost creep since you last priced them, so the Pareto cut becomes a live dashboard rather than a quarterly project.

If you are on Square specifically, the Square category analyser will run a free first-pass Pareto on your last 90 days of sales without any signup. Good way to see whether the pattern in this post matches your own café before you commit to anything bigger.

The point is not the tool. The point is knowing your top 10. Once you know them, every other menu decision gets easier, faster and more profitable. You stop running a 70-item café in your head and start running the 15-item café you actually have, with 53 items of optional padding.

The big chains have known this for decades. It is why the menu at a Pret looks the way it does. There is no reason an independent café in Witney or Burford or Carterton cannot run on the same logic, just with better food and a soul.

Ed O'Brien has run Hunters Cake Company for 17 years across cafés in Witney, Burford, and a bakery in Carterton, Oxfordshire. He's building Brikly - modular tools that give independent café owners the same data the big chains have, without the big chain price tag.