May 19, 2020

Sample Roasting: What You Should Be Looking For


Sample roasting will help you decide whether or not to buy green coffee beans, how much they’re worth, and what roast profile you should use. Yet no matter why you’re doing it, it’s important that you know what to look for. 

After all, there’s a lot resting on this tiny sample. You’ll want to make sure that you understand all the available data so that you can be confident in your final decision.

To find out more, I spoke to experts from sample roaster manufacturer ROEST, specialty coffee roasters, importers, and more. Here’s what they told me.

Lee este artículo en español Sample Roasting: What You Should Be Looking For

Credit: ROEST

How to Sample Roast: A Step-by-Step Guide

Sample roasting is an essential step for quality analysis. Stephen Cornford, Coffee Hunter at coffee importer Mercanta, says, “Sample roasting is a fundamental activity in our business, used in every step from selecting delicious coffee to buy, checking its quality on arrival, tracking performance over the season, and promoting them to our customer base of specialty roasters through cupping events and roasted samples in the post.”

Yet effective sample roasting starts long before the coffee roaster is warmed up. It starts with an examination of the green beans.

1. Recording The Green Beans’ Physical Attributes 

Grab a pen and pencil, or load up your software, because there’s a lot of data to be recorded. 

Stephen says, “When we receive offer and approval samples, they are typically very small – just 400g. And so, before sample roasting, it’s important to record all of their physical attributes, such as screen size, moisture content, etc., since you will have lost the chance once you’ve roasted them.” 

Even if you’ve specified these details in the contract, you should still measure it in order to keep a record of it. Screen size can potentially influence the roast outcome, as the heat will transfer differently through different-sized beans. Knowing the proportion of each screen size in your sample will help you to not only ensure your contract is being fulfilled but also better execute your roast. You can use grading screens, often made of metal sheets, to classify the green coffee by size.

Morten Wennersgaard, Founder of Tropiq and Nordic Approach, stresses the importance of knowing the density, moisture level, and water activity. The softer the coffee bean, the gentler and longer you’ll want the roast to be. Meanwhile, the greater the moisture level, the higher the risk of lost profits and the longer your drying phase will need to be. And water activity measures the likelihood of mold growth.

Read more in our roasters’ guides to coffee bean density and moisture content!

The coffee’s variety and processing method will also give you an insight into the beans’ physical characteristics, as well as the flavor and aroma profiles you might expect. 

Honduran coffee producer Marysabel Caballero tells me, “You need to know the variety of the coffee because that tells you the shape and size of the beans, and also gives you information about how they behave during the roast. Second, the process the beans went through, i.e. washed, natural, or honey, also gives you information about the coffee’s behavior during the roast.” 

You might also like: Practical Tips For Quality Control in The Coffee Roastery

Credit: Mercanta

2. Measuring Green Bean Defects

Just a couple of defects can significantly reduce the coffee’s cup quality. Whether your aim is to check the sample quality or remove defective beans, this is an important step.

A dark background (e.g., black paper) and a neutral light will help defects stand out. Using UV lighting is also helpful since it illuminates defects that are otherwise invisible to the naked eye, such as quakers.

While there are different ways that organizations and individuals grade their green beans, the most common method is with the SCA’s green grading form. You’ll want to use 350g of green coffee and count the number of Category 1 and 2 defects.

Just one Category 1 defect (e.g. full black, full sour, or a foreign object) in a sample can mean the coffee is not considered specialty.

Discover more in 7 Green Bean Defects Roasters & Producers Need to Recognize

Credit: Neil Soque

3. Selecting The Roast Method

The SCA has published protocols for roasting samples, which dictate the ideal roast time and color. However, how you get there is up to you. 

Stephen says, “When it comes to approaching sample roasting, many roasters I know use different profiles for different types of coffees; however, here at Mercanta, we believe in using a single profile which doesn’t favor any coffee more than another. Fewer variables, fewer mistakes.”

Not only is there less potential for mistakes, but you’ll know that you’re comparing the coffee rather than the roast method. Morten tells me. “My tip sometimes would be for roasters to make it simple… Small details can make a big difference. But when you have rules for buying coffee, you should have actually a couple of standardized profiles and not make it too complicated.” 

That’s not to say that you can use identical roast profiles for every coffee. If you’re planning to compare them fairly, a low-elevation, washed processed coffee with 12% moisture content will need a different approach than a dense, natural processed lot with 10.5% moisture content. 

You’ll also need to adjust for environmental factors, e.g. the temperature in the roastery, and any changes you make to your brewing equipment.

However, Morten recommends having standard profiles for similar coffees. You can then use these to make sure that it’s the differences between the coffees that stand out – not the roast profile.

Credit: ROEST

4. Roasting Your Sample

The better you understand your roaster, the more control you’ll have over it. The more control you have over it, the better your roasting consistency. This means you need to understand how your sample roaster behaves.

Morten, who worked closely with the ROEST founders to develop a sample roaster matching the importing coffee company’s needs, stresses the importance of consistency and generating measurable data points. “You need measurable things and to be able to control all the variables.”

Pay attention to environmental factors like temperature and humidity in the roastery, as well as all the data points available to you, such as bean temperature, rate of rise, environmental temperature, and environmental rate of change. Study how they interact and what impact they have on the coffee’s cup profile.

Read more in How to Use Common Roast Data

Smart sample roasters make it easier to measure, control, and compare variables. For example, the ROEST sample roaster allows you to control several variables ranging from drum speed to the airflow at the cooling tray, from the digital interface. Getting the drum speed right can improve the consistency of the heat transfer within the beans, and help you adjust to different sample sizes. Trond Simonsen, ROEST CEO and Co-Founder, says “You can learn a lot about coffee when operating a sample roaster that doesn’t limit your control”.

Sverre Simonsen is CTO and Co-Founder of ROEST. He tells me, “It mustn’t be hard to control [your roast]. It must be easy to replicate your results.” He stresses that roasters cannot afford to waste even small portions of the sample on poorly controlled roasts. “It’s sample roasting… you might only have one or two tries,” he stresses.

With ROEST, you can automate your roasting, but Sverre recommends seeking a balance between automation and manual control, as “even though you save a lot of time roasting automatically, the operator should have an option to manually interact with the roaster at any time during the roast. ” 

He adds that with automated roasting, “It should be easy to fully develop each coffee with less focus. Especially when you have loads of samples to get done.” In addition, it will free you up to pay attention to other tasks needing attention. “You need to evaluate green beans, weigh beans, print labels…therefore it’s better to invest in a roaster that doesn’t require your full attention, saving your time for other tasks.”

Credit: ROEST

5. Analysis

And finally, be ready to analyze the roasted coffee. Don’t rush straight to the cupping table: Marysabel recommends starting with a visual inspection. “Once you’ve visually checked your roast, you can make changes to your profile. If the coffee is not homogeneous or well developed, it might be because the roast was too fast and/or too hot, so the bean didn’t have time to roast completely.”

When you’re ready to cup the coffee, remember your goals. Are you looking for an exceptional single origin or a good coffee for a blend? What types of coffees do your audience favor? Morten says, “We are buying coffee that goes into different markets with different preferences. So it’s more about trying to take out the potential of the coffee, so we can clearly see the flavor spectrum that can be developed… like flavors and sweetness.”

Remember to follow best practices for cupping. Include as many members of the team as possible to calibrate your scores. Blind cuppings will help make sure you’re being objective.

Ideally, your team will be cupping regularly to hone their skills. Stephen says, “It’s a cliché but the good roasters are also good cuppers. It’s not a bad idea to regularly go back and cup roast defects just to remember the tastes.

You might also like A Beginner’s Guide to Cupping Coffee & Improving Your Palate

Credit: Nordic Approach

Sample roasting is an essential skill for roasters, traders, and even coffee producers. It gives us insight into the coffee’s quality, as well as its potential for different markets. So, it’s important to have as much control as possible over the roast. Analyze your green beans. Use the same profile for similar coffees. And take advantage of all the available data.

Enjoyed this? Read Creating a Safe & Healthy Coffee Roastery

Feature photo credit: ROEST

Please note: This article has been sponsored by ROEST.

Want to read more articles like this? Sign up for our newsletter