If you read the coffee-internet enough, you are guaranteed to be more confused about roasting than you were before reading it. One of the more confusing bits of roasting advice I hear often is to “extend the Maillard phase” of a roast. The internet seems to believe more time in the Maillard phase has predictably beneficial results.

One challenge in roasting is when a roaster makes a single change from batch to batch, tastes a difference, and then declares that one aspect of that change (eg lower gas, lower ET, longer Maillard phase, longer roast time, etc) responsible for the change in flavor. Unfortunately, to be confident in the cause-and-effect relationship, one would have to make that change countless times, under controlled conditions, and make extreme efforts to rule out other possible causes of the new result. It’s painstaking work, which is why scientists often spend years and millions of dollars to study the effects of a small change in a system—often without clear answers despite the time and effort!

A few thoughts on why the standard Maillard-phase thinking is questionable:  

• Extending the Maillard phase necessarily means lowering and flattening the ROR, reaching first crack at a lower ROR, and increasing the risk of an ROR crash. There is no way to change just one variable in roasting— every change creates a chain reaction.

• Extending the Maillard phase results in Maillard reactions happening at a slower rate. If you double the time in the Maillard phase, you may approximately halve the rate of Maillard reactions (this is a guess; I don’t know the exact change, but the relationship of Maillard phase time to Maillard products is not remotely linear, and no one really knows the relationship yet.) So, it’s possible the net change in Maillard products is negligible. I do not believe this issue has been studied thoroughly.

• One can easily create two equivalent-tasting roasts of quite different total durations and Maillard-phase durations. For example, a 7:00 sample roast may taste nearly identical to a 13:00 production roast. This fact alone negates the standard Maillard thinking.

• It is suspicious that no one ever seems to recommend shortening the Maillard phase. The implication seems to be “more is better.” Every good practice in roasting seems to have a “happy medium.” I can’t think of any instance where more is always better. This is a red flag indicating the proponents of longer Maillard phases are just repeating a trendy idea and don’t really know what’s optimal.

• Focusing on balancing the time in various phases is extremely difficult and will almost necessarily lead a roaster to sacrifice curve shape, consistency, and predictability.

The bottom line is that roasting is complex. Most of what one reads online is false. Knowing a factoid about chemistry is not the basis for a roasting system. It’s an excellent practice to form opinions slowly, and only after exhaustive testing. 

Sign up for my ONLINE ROAST COACHING GROUP

Cancel anytime.

The group is incredibly affordable, supportive, and effective. Just having access to the vast archive of posts is worth the price of admission.