My Problem with Thermometers
/By Michael Teahan
Back Story
A couple of decades ago, before Starbucks used robots to make coffee, they embarked on a coffee quest. They had figured out how to put stores on 3 corners of the same intersection but they couldn’t wrap their heads around the problem of shot time consistency. Every time there was a shift change, they would have to readjust grinders and toss coffee because Bruce and Bob dosed and tamped a little differently. Before they pulled the trigger on automated machines—they broke a lot—they tried to find a solution for their traditional machines.
A small contingent from Big Green went to Italy to tour some manufacturers and see what could be done and one of their stops was at Rossi in Retorbido. When they asked about how Italians dealt with the problem of tamping consistency, the engineer didn’t seem to understand the problem. He ground the coffee, loaded the portafilter and pulled a shot without tamping anything that was exactly the same as a tamped shot.
The reason was that the machine was fitted, as were many machines in the Italian market, with a preinfusion system that saturated the grounds before applying pressure. What Italians had long taken for granted since the early 1950’s, the largest coffee retailer in the world had never heard of. Apparently.
Two things came from that meeting. The first was a scramble from Starbucks to fit preinfusion to traditional machines as an alternative to jumping all in on super-automatics and the second was a frenzied notion that Marzocco’s lock on Starbucks could be broken.
There appeared to be a mad dash to get a machine into Starbucks to test and what could they build to blow them away. In the midst of all this, Marzocco wasn’t idle. Within months, Bambi filed a patent on a preinfusion system for their saturated group design. While competitors saw an opening to squeeze into Starbucks, Marzocco saw that it was really between traditional and automated machines—which they did not have.
We went to work on a machine built just for Starbucks to play with. It used mechanical preinfusion that could be enabled with a switch so they could compare the effect. They also knew that Starbucks sold more milk that coffee, so we built a machine that never ran out of steam. That part was a surprise.
Road Trip
The machine arrived in Los Angeles where it was bench tested and shipped off to Seattle. They knew it was coming and seemed anxious to see how it worked. This was because the machine got there a couple of days before me and when I walked into their R&D department it was stripped naked.
There were at least a dozen other machines on benches—we were not alone—but at least our was hooked up.
But there was a problem. They had fitted an analogue temperature gauge used to set Marzocco boiler temperature to one of our portafilters and found the machine to be too hot. No matter what they did, they couldn’t get the temperature down to their spec. They turned down the pressure stat, drained water from the boiler and even opened the steam valve for 30 minutes and couldn’t get the steam pressure below 1 bar. They reasoned that if they could just get the pressure down, the temperature would come with it. They were so fixated on temperature that they weren’t paying attention to anything else. When I pointed out that they had done everything they could to kill the machine and it refused to run out of steam, I asked them if that isn’t really what they were looking for. There was a dramatic pause as they looked at each other as that sank in.
They hadn’t even pulled a shot yet. They were so married to their standard that they didn’t even try. I pretended to be puzzled—most HX machines that temperature profile run hot at the start and I didn’t want to have that discussion yet—and suggested that we just make some coffee and see what happens. Of course, the coffee was fine and the preinfusion effect came across like a mysterious magic trick.
I think the writing was already on the wall about a shift to super automatic machines, but we continued to talk with their engineers every couple of days for a few months. There were so many machines on the bench in Seattle, though, that I don’t think they were really interested in another traditional machine. We did manage to sell them a bunch of little pour over machines, though. They may have felt sorry for us.
Another expert with a thermometer.
We weren’t finished with this machine yet. We managed to get a test into a Boston roaster a few months before they were taken over by Starbucks. I arrived the morning after a red eye flight from Los Angeles to find the machine all setup and ready to go. I walked up to the machine to give it the usual rundown to make sure everything was ready, checking the steam and water pressure and opening the valves. As I flushed the groups a young man walked up behind me and pulled a thermometer out of his pocket. He checked the temperature of the water from the group and declared that they could not use this machine, it was incapable of brewing espresso as it was too hot. I had already been through this with the lead engineers at Starbucks and I asked him about his background and expertise (it was a red eye and I was a little punchy).
His singular reply was that he was from Seattle. At the time, merely being from Seattle was an acceptable qualification for quality control positions anywhere east of the left coast. He couldn’t have been 22 years old and he was, in fact, his quality control expert.
I said that since I was already here and the machine was setup, let’s see how bad it could be. After fifteen minutes of pulling shots and steaming milk, someone asked if we were using their coffee for the test (we were) because they had never seen their coffee come out like that.
It had crema.
So . . .
We are sometimes so convinced of what we know that we forget to be open to other ideas and the possibility that we could learn a little more. If you can’t tell a good shot from bad with three of your five senses, nothing will help you.
This is why I hate thermometers.