What Makes a Gelato Recipe Work: Six Numbers to Check
Table of contents
Most gelato recipes cannot tell you whether they will work, because they were never measured. They are a list of quantities somebody liked once. Six numbers separate a recipe that repeats from a recipe that got lucky, and you can check all six before you buy a single ingredient.
Why most recipes cannot tell you anything
A recipe is a set of instructions. A formula is a set of proportions with targets. Almost everything published online is the first thing pretending to be the second.
Here is the practical consequence. When a recipe fails, and it will, you have no lever. Was it too hard because the sugar was low, because the total solids were low, because you froze it too slowly, or because you served it at freezer temperature? A recipe cannot answer that, because it never told you where any of those values landed. You are left changing one thing at random and hoping.
A balanced formula answers it immediately. It tells you the mix contains, say, 38 percent solids and a PAC of 255, so if the result is still hard the problem is not the sugars and you go look at the freezing instead. That is the entire difference, and it is worth more than any specific recipe.
The good news is that you do not need the author to have done this work. You can do it yourself, on any recipe, in about five minutes.
The six numbers, and the range each has to hit
Every professional gelato formula reports the same handful of values. These are the ranges for a milk-based gelato, the gelato al latte family that covers fiordilatte, vanilla, chocolate, nut pastes and most of what a gelateria sells.
Quick reference. Milk gelato targets: total solids 36 to 42 percent, sugars 16 to 22 percent, fat 6 to 10 percent, MSNF 8 to 12 percent, PAC 240 to 280, sweetening power 180 to 220. A recipe that lands outside two or more of these will not behave.
| Number | Target | What it controls | What goes wrong outside it |
|---|---|---|---|
| Total solids | 36 to 42% | How much of the mix is not water | Low: icy and thin. High: pasty, heavy |
| Sugars | 16 to 22% | Sweetness and softness together | Low: hard and flat. High: soupy, cloying |
| Fat | 6 to 10% | Body, melt, mouth coating | Low: lean and cold. High: greasy, muted flavour |
| MSNF | 8 to 12% | Protein and lactose that block ice | Low: coarse. High: sandy from lactose |
| PAC | 240 to 280 | How much water stays liquid when frozen | Low: rock hard. High: will not hold a scoop |
| POD | 180 to 220 | Sweetness, independent of softness | Low: tastes flat. High: sweetness masks flavour |
These are not arbitrary house preferences. They are the windows in which the physics cooperates. Below 36 percent solids there is simply too much free water available to freeze, and it freezes into crystals the tongue can find. Above 42 percent the mix runs out of water to dissolve everything and the texture turns pasty and dull. The same logic holds for each of the other five: every window is bounded on one side by ice and on the other by gum.
The pair that surprises people is PAC and POD. Sugars do two separate jobs, and different sugars do them in different proportions. Sucrose is the reference at 100 for both. Dextrose has a PAC near 190 and a POD near 70, so it softens far more than it sweetens. That is why a recipe that lists one single sugar has already given up one of its two levers, and it is the fastest tell in the whole audit.
Fruit gelato and sorbetto run their own ranges, lower in fat and higher in sugar. The method here is identical; only the target windows move.
The four tells that a recipe was never balanced
Before you run any arithmetic, four signals tell you almost everything.
It is written in cups, or in whole units. Four egg yolks is not a quantity. Yolks range from roughly 15 to 20 grams, so four of them is anywhere from 60 to 80 grams, and at 32 percent fat that swing alone moves the fat content of a 1 kg batch by more than half a percentage point. Grams to a fixed batch weight is not fussiness. It is the only way the numbers can exist at all.
It uses exactly one sugar. As above: sucrose alone means the author had no way to move softness without also moving sweetness. Almost every recipe that produces a brick is a single-sugar recipe with the sugar set by taste.
There is no milk powder and no mention of solids. Whole milk is roughly 88 percent water. A recipe built on milk and cream alone often lands with MSNF around 5 or 6 percent, well under the window, because skim milk powder is the only practical way to raise it without adding fat or water. Its absence usually means solids were never considered.
The fourth is softer but almost as reliable: the recipe gives a churning time rather than a temperature, and no storage advice. Gelato is finished at a draw temperature, not after a fixed number of minutes, and a recipe written by someone who tracked that number would say so. Silence about storage usually means nobody tasted the batch on day four, which is when an unbalanced mix reveals itself.
None of these four make a recipe unusable. They make it unaudited, which means you have to audit it yourself.
Run the audit on a real recipe
Take the most common shape of internet gelato recipe: 500 g whole milk, 250 g heavy cream at 35 percent fat, 150 g sugar, 4 egg yolks. Total batch, about 968 g. Using standard composition values for each ingredient, here is roughly where it lands.
| Number | This recipe | Target | Verdict |
|---|---|---|---|
| Total solids | about 35.9% | 36 to 42% | Just under |
| Sugars | about 15.5% | 16 to 22% | Just under |
| Fat | about 13% | 6 to 10% | Far over |
| MSNF | about 5.8% | 8 to 12% | Well under |
| PAC | about 190 | 240 to 280 | Far under |
| POD | about 160 | 180 to 220 | Under |
Read that table and the recipe stops being a mystery. Not one of the six lands inside its window, and total solids miss by so little that the recipe looks plausible right up until you check where those solids came from. They are cream fat, not milk solids: fat is half again over the ceiling while MSNF sits at barely half the floor. The mix is simultaneously too rich and too coarse.
The sharpest number is PAC. The recipe carries a perfectly respectable amount of sugar, and its PAC is still 188 against a target of 240, because every gram of it is sucrose. Sugar quantity and anti-freezing power are not the same measurement, and this is what that difference looks like on a spoon: it will freeze hard enough to bend one.
That is not a bad recipe. It is a French-style ice cream recipe, and it would be perfectly good if it were labelled as one. It is simply not gelato, and it will never behave like gelato no matter what machine you put it through. The numbers said so before you turned the stove on. Run the same recipe through the free total solids calculator and the PAC calculator and you will get the same answer in about two minutes.
What to do when a recipe fails the audit
Failing the audit does not mean discarding the recipe. It means you now know which lever to pull, and each of the six numbers has an obvious one.
- Fat too high. Swap cream for milk, gram for gram. Cream is the only concentrated fat source in most home recipes, so this is a single substitution.
- MSNF too low. Add skim milk powder, typically 30 to 60 g per kilogram of mix. It raises solids and protein without adding fat or water, which is why professionals reach for it first.
- PAC too low. Replace 20 to 40 g of the sucrose with dextrose. Softness climbs sharply, sweetness barely moves.
- Sweetness too low after that swap. Nudge sucrose back up, or accept it, because most people find properly balanced gelato tastes cleaner than they expected.
- Total solids too low. Milk powder, more sugar, or a nut paste, in that order of preference.
- Total solids too high. Add water or milk. This feels wrong and is usually correct.
Order matters more than people expect. Fix fat and MSNF first, because both change the total solids figure and would invalidate any sugar work you did beforehand. Set the solids, then set the sugar blend, then check PAC and POD last. Working the other way around means recalculating everything twice.
Change one thing, note it, and taste the next batch cold rather than warm. Two or three passes is normally enough. The full sequence, in professional order, is set out in how to balance a gelato recipe.
When a recipe breaks the rules on purpose
An honest caveat, because rigid rule-following produces its own failures. Several respected styles sit deliberately outside these windows.
Sicilian bases often use starch instead of cream and run 3 to 5 percent fat, well below the window, and they work because the starch supplies the body that fat would have. Alcohol-based flavours run PAC deliberately high because the alcohol itself contributes anti-freezing power that the sugar no longer has to. Chocolate gelato often runs higher solids because cocoa contributes both solids and its own fat.
The distinction is intent. A formula that sits outside a window and says why has made a choice. A recipe that sits outside four windows and says nothing has made a mistake. After a dozen audits you will tell them apart instantly, and at that point you have stopped needing anyone else's recipe to be correct.
Related Concepts
- How to balance a gelato recipe - the full professional sequence
- Gelato total solids - the number that builds body
- Sugar selection guide - choosing the blend, not just the amount
- How to calculate MSNF - the step-by-step arithmetic
- How to make gelato creamy instead of icy - the same numbers, from the symptom side
