Toolkit Tuesday: The Roast Profile Is Where Your Flavor Is Actually Decided, And How to Build One You Can Repeat

By Cocoa Craft

The workshop was quiet except for the drum, the cooling fan, and the occasional scrape of a pencil across a notebook.

A maker had spent months trying to fix a flat 60% dark chocolate formula. Sugar percentages changed. Cocoa butter was adjusted. Refining and conching times moved up and down. The recipe was treated as the problem because the finished bar tasted dull.

Then the maker roasted the same beans differently.

A slower drying phase brought out more fruit. A steadier development stage added cocoa depth without bitterness. The recipe did not change. The roast did.

That experience is common in bean-to-bar chocolate. Craft makers and serious home chocolate makers often spend their time refining the formula downstream while treating roasting as a fixed prelude. In reality, the roast profile is one of the first major flavor decisions in chocolate production. It determines which aromas are developed, which remain delicate, and whether the final chocolate has clarity, body, roast character, or a flat center.

"Get free, unlimited access to the full Cocoa Craft software suite in exchange for monthly feedback. Join the beta at cocoa-craft.com!"

Before your next production run, build a roast and batch plan in Cocoa Craft. Use the software to connect the roast decision to the formulation, batch records, and tasting notes that follow.

What a roast profile actually records

A roast profile is more than a final temperature and a total time. It is the repeatable path your beans take through heat, moisture loss, flavor development, and cooling.

These are the main points to track:

• Charge temperature: The air or drum temperature when the beans enter the roaster. It controls how aggressively the batch begins and how much energy the beans receive at the start.
• Charge weight: The quantity of beans loaded into the drum. This affects how quickly the roaster recovers and how much thermal momentum the batch carries.
• Drying phase: The early stage when beans warm and release moisture. Watch the temperature curve, the color shift, and the aroma. Early smells may be sharp, acidic, or vegetal before moving toward toasted grain, nuts, caramel, and cocoa.
• Color and aroma cues: The beans should develop evenly. Look for a consistent color rather than isolated dark spots. Aroma is equally useful: the transition from raw and acidic to browned, nutty, or brownie-like tells you that the roast is moving into flavor development.
• Development time: The period after the main moisture release, when the flavor profile becomes more defined. Too little development can leave a thin or sour result. Too much can mute origin character and push the roast toward bitterness.
• Drop temperature: The bean temperature at which you stop the roast. This is a reference point, not a universal target. The right endpoint depends on the origin, bean condition, equipment, and the flavor you want.
• Cooling speed: Beans continue cooking after they leave the drum. Fast, even cooling helps stop the roast where you intended. Slow cooling can turn a balanced roast into a heavier, more roasted one.

The most useful profile is the one you can describe and repeat. “Roast until it smells good” may be a valid sensory instinct, but it is not enough for a holiday production schedule or a team handoff.

The four levers you can control

When you are tuning a roast, change one of these four variables at a time.

1. Charge temperature

A higher charge temperature gives the beans more initial energy and can shorten the path to the drying phase. It may support deeper roasted, nutty, or caramel notes, but it can also create uneven development if the surface heats faster than the center.

A lower charge temperature generally gives you a gentler start. That may protect floral or fruit notes, but it can also create a long, dull drying phase if the roaster lacks enough heat recovery.

Do not make large changes without tasting. A small adjustment can alter the entire curve.

2. Charge weight

Charge weight is not simply a production quantity. It is a heat-transfer decision.

A heavier batch pulls more energy from the drum and may slow the roast early. A lighter batch may respond quickly and develop faster than expected. Record the actual weight, not just “half batch” or “full batch.”

3. Time to drop

Total roast time matters, but the distribution of that time matters more. A 20-minute roast with a smooth drying phase and controlled development will not taste the same as a 20-minute roast with a stalled middle and a rushed finish.

Track the time to key events, especially the end of drying, the beginning of development, and the drop.

4. Cooling speed

Cooling is the final part of the profile. If the beans remain hot in a deep pile or a slow-moving cooling tray, the roast continues. That can erase the difference between two carefully tuned batches.

Use the same cooling method whenever possible, and record the time from drop to a stable, cool batch.

What you mostly cannot control

Your process needs to account for variables outside the control panel.

Ambient room temperature affects heat loss, airflow, and cooling. A winter roast in a cool production room may behave differently from a summer roast in a warm one.

Bean moisture and size also change the profile. New-crop beans landing now may have different moisture, density, and size distribution from the same origin last year. A profile that produced expressive chocolate in 2025 may taste flat or overdeveloped in 2026.

That is not a failure of the old profile. It is a reason to re-tune.

For Q4 holiday production, start new-crop trials earlier than you think you need to. Change one lever at a time, cup the result, and keep the previous profile available as a reference. If you change charge temperature, weight, airflow, and drop point together, you may get a better batch without learning why it improved.

Scaling from a benchtop roaster to production

A benchtop drum roaster sized for 2–4 lb / 1–2 kg batches is useful for origin trials, recipe development, and small production runs. A 10 kg / 20 lb commercial roaster with profile control provides more capacity and more automation, but it will not behave like a simply enlarged benchtop machine.

The important comparison is charge weight relative to rated capacity, not charge weight relative to the drum’s physical maximum. A 1 kg test batch in a 2 kg roaster is a 50% capacity load. A 5 kg batch in a 10 kg roaster is also a 50% capacity load. Those ratios provide a more useful starting point than trying to copy the absolute weight.

Even then, the curves may differ.

Larger batches carry more thermal momentum. They can recover differently after charging and may respond more slowly to changes in heat or airflow. Air movement through the bean bed also changes with depth. A deeper batch can need different airflow and heat management to dry evenly.

To scale a profile:

1. Match the percentage of rated capacity.
2. Compare time to the end of drying, not just total roast time.
3. Compare the shape of the temperature curve and the rate of rise.
4. Adjust charge temperature and airflow conservatively.
5. Taste the scaled batch before approving it for production.

The goal is not to copy every number. The goal is to reproduce the same flavor-producing pattern.

You can find the two equipment classes currently stocked for chocolate making at the Cocoa Craft equipment shop.

Build a roast log that survives the next season

Roast data becomes valuable when it is connected to batch records and tasting notes. Without that connection, a temperature graph is only a partial memory.

Record at least:

Field	What to capture
Origin and lot	Supplier, origin, lot or harvest reference
Moisture	Measured moisture if available, plus visual observations
Charge weight	Exact weight and percentage of rated capacity
Charge temperature	Air or drum temperature at loading
Drying notes	Color, aroma, timing, and any stall
Drop temperature	Bean temperature at drop
Total time	From charge to drop
Cooling time	Drop to a stable, cool batch
Tasting note	Acidity, fruit, cocoa, nut, roast, bitterness, body
Batch result	Approved, needs adjustment, or rejected

Copyable pre-roast checklist

ORIGIN / LOT:
DATE:
ROASTER:
RATED CAPACITY:
CHARGE WEIGHT:
BEAN MOISTURE:
ROOM TEMPERATURE:

[ ] Beans inspected for size variation and defects
[ ] Roaster and cooling tray clean
[ ] Ventilation and airflow checked
[ ] Charge temperature stable
[ ] Timer and temperature recording ready
[ ] Sample tray and tasting tools ready
[ ] Target flavor direction defined
[ ] Previous profile available for reference

PLANNED CHARGE TEMPERATURE:
PLANNED DROP TEMPERATURE:
TARGET DEVELOPMENT APPROACH:
COOLING METHOD:

Create a Cocoa Craft login so roast profiles, batch records, formulation versions, and tasting notes can be saved against each origin. When next year’s beans arrive, you should be able to see what changed, what worked, and what needs to be re-tuned.

Make roasting part of the formulation workflow

A roast profile should not live in a separate notebook from the recipe. If a 60 cocoa chocolate formula tastes different after a new crop arrives, the first question should be whether the bean changed, the roast changed, or the formulation changed.

Cocoa Craft’s software helps keep those decisions together. Record the roast, move into the formulation, capture downstream process notes, and compare the finished result with earlier versions. That is how chocolate craft becomes a working system rather than a collection of isolated experiments.

It is also the practical meaning of “Built by chocolate makers, for chocolate makers.” The tools should support the decisions that happen in a real workshop: weighing, roasting, tasting, adjusting, and preparing the next run.

For equipment planning, visit the chocolate making equipment shop. For the production workflow, create your free Cocoa Craft account and build your next batch plan before the beans enter the drum.

"Get free, unlimited access to the full Cocoa Craft software suite in exchange for monthly feedback. Join the beta at cocoa-craft.com!"

When the roast is cooling and the next batch is loading, take a short break in the Cocoa Craft Arcade. Cocoa Solitaire is live now; Tempering Tiles, Roast Runner, Grind Gauntlet, Origin Odyssey, and Bean Drop are coming soon.

A repeatable roast will not eliminate the craft from artisan chocolate. It protects the craft by making your best decisions visible, testable, and ready to use again.

The operating system for chocolate production starts with a profile you can repeat.