[S1:E2] Interview with Neal Wilson

Podcast feed
or download torrent file

Neal Wilson is owner and roaster at Wilson Coffee and Tea and is a long-time instructor for the SCA. In this episode, I ask Neal aobut the rate of change, Typica software, and instructing at the Roasters Guild Retreat.


This audio was recorded at the Coffee Roasters Guild retreat in Stevenson, Washington in 2018 and was later transcribed from the recording.

Neal Wilson
Wilson’s Coffee and Tea
Recorded: 25 August, 2018
Transcribed: 9 September, 2018

My newest coffee is from Oaxaca, Mexico and it's a chocolate-bomb! Check it out: Shop.OilSlickCoffee.com/products/mexico-oaxaca


[Beginning of recorded material]

Michael: Neal, thank you for taking the time and sitting down with me. I appreciate it.

Neal: Sure, no problem.

Michael: You are a roaster, you own a cafe, you’re a co-owner of a cafe, and you author Typica roasting software.

Neal: That’s correct.

Michael: Why don’t you tell me a little bit more about you, how you got into coffee. It’s an interesting story.

Neal: Sure. Well my parents started the company when I was ten. You know, they wanted to start a business and initially they didn’t really know what sort of business they wanted to start. I remember as a kid they were bouncing around different ideas and eventually my mother noticed my dad was driving to Milwaukee every week to buy coffee, because there wasn’t a place in Racine with good coffee. And it kinda clicked and that’s when they started really exploring that idea. It was always their intention, in whatever business they wanted to start that that was something that the whole family could be involved in. So I just sorta grew up in the business.

Michael: Very cool. And now you’re roasting for the business as well?

Neal: Right. We expanded out of our original location. We moved a couple of blocks down the street to the other side of the street. We bought the building that we’re in now and in 2000 we got that opened up and that’s when we started roasting coffee and we also put in the commercial kitchen and started doing our own baking at the same time.

Michael: Very nice. And this was Wilson’s Coffee and Tea in Racine?

Neal: That’s correct. It’s not a very creative name.

Michael: It works! It works! And you’re roasting on a Diedrich now. Is that correct?

Neal: Yes.

Michael: And what were you roasting on before?

Neal: I wasn’t really. Because we had a little tiny home fluid-bed thing that, this was not really a thing that we were roasting on but more of a, it’s kind of cool to be able to pull a thing out and show customers that ‘before we get this roasted coffee, it’s this seed.’ And you can see what’s happening to that. And the nice thing about those is that you can see you’ve got, it’s a glass bowl, basically so you can really well see those changes that are happening in the coffee a little bit easier than like in a drum where you’ve just got a little sight glass going.

Michael: It’s quicker for a demo.

Neal: Yeah. It is. But we weren’t really trying to produce coffee that way. It was more of a way to say…It’s important to remember that back in the 1990’s, the consumer understanding of specialty coffee was very different from what it is today. Back in the early 90’s, when we started, if you asked people what they thought a specialty coffee was, they would answer like ‘hazelnut,’ or ‘French vanilla,’ or something like that. And these days, I can’t sell flavored coffee as a specialty roaster because people who are buying those, you know, they’re getting it at the grocery store, or whatever. It’s sort of…so we’ve been trying to spread that education as an industry, really, not just us in our shop. But a lot of roasters or even just coffee retailers have been trying to point out that ‘hey, there are all these coffees that are growing all over the world and they have all of these different flavors inherent in the coffee. You don’t need to add these other things to it.’

Michael: And do you get to educate your customers with that, and show them that?

Neal: Yeah. We used to do that a lot more than we do now, because again, earlier that awareness wasn’t really there and these days you can find that information in so many different places. And the level of awareness…the people who are interested in that can find that information. And I’m happy to talk with them about that, but I’ve had a lot less need to do specific workshops or things like that in the shop. So we’ve dialed those back quite a bit.

Michael: Yeah. Very cool. Now you’ve been roasting for 18 years! Production roasting…

Neal: Yeah. Right. It was when we expanded in 2000, that’s when we put in the roaster and I took over that aspect of the business, pretty much immediately.

Michael. Nice. And it’s a good fit for you?

Neal: I like it, yes. I really enjoy roasting coffee. And more so than even just the day-to-day roasting, what I really enjoy is the product development aspect of it; getting new coffees and figuring out what do I want to do with this coffee.

Michael: And you and I talked a little bit earlier about the rate of change and that’s one of the things I’ve been focusing on in some of my videos. And you had a lot of really interesting views on that. Can you talk a little bit more about how you use rate of change and how you find it important?

Neal: Right. What this is, is you’re really looking at how the temperature of coffee is changing. It’s not…sorry, I’m tripping over my tongue. The really useful thing about rate of change is that when the rate of change changes, that calculation that you see as you’re graphing that out, that changes in a way that is more quickly obvious than if you’re just looking at the bean temperature. So if you think about, if you’re going along at, you know, nine degrees every thirty seconds and something changes and now you’re going at about nine-and-a-half degrees per thirty seconds, it takes a while, if you’re just doing, if you’ve got your profile following your plan, it takes a while before you really notice you’re getting off on the bean temperature. But on that rate of change calculation, it becomes obvious a lot faster. And so you’re able to pick up on that change and make appropriate adjustments much more quickly.

Michael: Hold that thought…I’m sorry, this, this stopped. I heard it stop. It’s been doing that. Okay, we’re back on. Resync the audio. So go ahead.

Neal: I was pretty much done with the thought there, so…

Michael: So rate of change is more of a tactical, immediately useful piece of information?

Neal: Right, and what I’ll do is I’ll have both the bean temperature and that rate of change in my roasting plan and so I’m trying to get both of those lines to match up as I’m roasting.

Michael: Okay. That’s very cool. And that segues nicely into Typica, which is the roast software that you author and and maintain. So tell us a little bit more about that.

Neal: The way this started, was, it actually starts with some hardware failures, because I used to do all this logging manually. I’d look at a timer. I’d look at my temperature display. I’d write down in a notebook every thirty seconds what the measurement is. So the timer that I had, it was a big Radioshack, dual digital timer clipped up to the hopper. And it broke to the point where it was no longer reliably keeping time and replacing the battery didn’t fix it. And I looked around and nobody cooks anymore. So I couldn’t find anyplace locally that had an appropriate timer to sell me. And after going to a number of stores, I thought ‘you know, I write software, I can just make my own timer.’ It takes more time to look at all of the timer programs that people have written to see if those might be appropriate than it does to just write one. It really is that trivial.

Michael: For you maybe!

Neal: But then when once you have a computer at the roaster you may as well start using more of that, right?

Michael: Sure.

Neal: So the next thing to go was I had a panel display and the numbers on those are the seven-segment LEDs and some of those started to go out and it got to the point where it became ambiguous. You couldn’t tell necessarily, just from looking at that, what the measurement was because there would be a couple of possibilities.

Michael: Is it a ‘3’ or is it an ‘8’?

Neal: Right. And if you have to stare at that panel before you know what the measurement is that slows things down and it’s kind of obnoxious. So I replaced that with, they of-course no longer made that exact model of panel display anymore, so the manufacturer recommended something else and it was, it had this weird quark to it, where it didn’t bother to do all of the math need to convert from the tiny little thermocouple signal into the temperature, because it’s this big giant polynomial. It wasn’t doing that on every, single, measurement. It would do it once, and then it would estimate. And once the error got large enough it would do it again. And so I’m looking at the panel display seeing the tenths-of-a-degree go up and I’m thinking ‘okay, when it hits this temperature I want to make this control change.’ And what would happen is on its way there, it would then suddenly jump up four degrees, clear where I wanted to make that control change. And while trying to wait for that to happen, I missed my mark. So I started looking into, ‘well is there something that I can hook into the computer?’. And it kind of snowballed from there.

Michael: Awesome. And Typica is database-driven. It has a database on the back-end.

Neal: Right.

Michael: And so you’re starting to include more features with that. So why don’t you talk about some of those features.

Neal: Sure. What it’s doing is it’s, well first the database is optional. So if you really don’t want to set that up, you don’t have to. You could save all of your roasting data to files and load things that way. But then you don’t get a lot of the functionality in terms of tracking what your production is doing over time. There are a ton of reports that you can run from that data that you’re collecting. So when you roast a batch of coffee, you’re saying ‘okay, here’s the roasted coffee item that I want to produce’ and it will look up what green coffee or coffees, because you can do pre-blends in it as well, that you used for that roasted coffee before. If you’re out of that coffee, it’ll show up blank. So that’s a clue that maybe you’re switching to a different lot and you might want to pay a little bit more attention to that. You might not actually want to roast that coffee if you want to go and do some product development first, to make sure that you’re using the right plan for that. But on a normal day-to-day thing it’ll pull up the green coffee, you put in the weight, and load your target roast profile, your roasting plan. And then at the end you can put in your roasted coffee weight. It can also hook into the Javalytics Degree of Roast analyzer—it can communicate over the network with that to just pull measurements right from there for whole and ground color.

Michael: Nice.

Neal: If you care to collect any of that data it’ll calculate your percent-mass-loss between your green and your roasted weights. And it will store all of that information in the database so that you can run a lot of different reports to see things like how much of this roasted coffee have I used on a weekly average over the past month, for example. So you can see, you can better decide things like how large your batches should be while hitting your freshness thresholds. So if you, you don’t, we have this advice that you want to roast large batches of coffee, because thats very efficient for, you know, it takes the same amount of time to roast a full batch of coffee as it does to roast a half batch. And if you’re roasting a full batch of coffee you’re getting that, you’re getting twice the as much coffee for the amount that you’re paying your roaster [operator]. Right?

Michael: Right.

Neal: But if you’re not selling that coffee so fast then that’s going stale. That’s going to impact customer-perception of your quality. So having that information really helps to strike that balance of maximizing efficiency while maintaining that freshness goal as well. And there’s a lot of different reports that you can run on different aspects of your production operations to see information that…it’s not about telling you what to do. But it’s about presenting the information that you have in a way that let’s you reason about your information easier and make those decisions more quickly and more accurately.

Michael: Yeah, information is empowering.

Neal: Right.

Michael: And it’s got to be very handy to be able to…

Neal: And the thing is that that information needs to be presentable in a way that you can easily digest it and make sense of it. Because if I had…I was doing a lot of these sorts of analyses when all of that was just on stacks of paper. But then you have to go through all the papers and pick out ‘okay, where are my data points that are relevant to what I’m interested in?’

Michael: And that’s assuming you’re gathering those data-points in the first place.

Neal: Exactly. Yeah. And by having this all in a database you can very quickly query that database for just the things that you are interested in. And that’s also one of the neat things about Typica is that since it is going to a database, if there’s something that you want, that you know you’re collecting the data, you know the data is there, but maybe Typica doesn’t have a report for it. You can connect to the database from any of the popular spreadsheet programs or you can connect and execute the SQL commands directly if you know that language. So I’ve had people who use the software say, you know, ‘here’s some information that would be really helpful for me to be able to get to’ or ‘here’s something that my organic certification auditor is interested in.’

Michael: That’s a good point.

Neal: Yeah. And I can often times go into my database say ‘okay, copy this query…’

Michael: Nice.

Neal: ‘…paste it there, and that’ll get you the info…’

Michael: And enjoy your data.

Neal: Right. But then that’s also helpful for me in knowing, you know, is this another report that I should add into Typica?

Michael: Right, is this a new feature I should be including?

Neal: Right. Right.

Michael: Very cool. Yeah, you might get me to convert. I’m using Artisan.

Neal: Sure, and I, Artisan is a good program. A lot of people use it. It has really fantastic hardware support. It just doesn’t have a lot of that surrounding business-data collection.

Michael: Right. Right.

Neal: So, that, I think they probably consider that to be out-of-scope but, you know…

Michael: Yeah, they’re largely [targeting] home-roasters and…

Neal: Exactly, right. And I would say that for a home-roaster, I know that there are some home-roasters that are using Typica, but it kind of over-kill and possibly a little top-heavy for that application.

Both programs really do have their niche and there’s room for more than one roasting program.

Michael: Yeah. Absolutely! And hopefully more room—get more roasters out there.

Neal: Yeah. Yeah.

Michael: So you are also instructing here at the Roasters’ Guild Retreat. You want to talk about that some?

Neal: Yeah. So I’m teaching a class called Roasting Styles Exploration [it] was under a different name at Expo. But it was really well-received there and this is, it’s based on the old Profile Roasting class, which is what actually got me instructing in the first place. I was coming out of Expo and another person was going in, or the other way around. I don’t really remember. But I was pulled aside by someone who was on the equivalent of Education Committee at the time and he said ‘I’m redoing this Profile Roasting class and I’d like your input on that and can you help me out with that?’. And what that really meant was ‘hey, can you rewrite this class for us?’

Michael: Nice.

Neal: And that class held up remarkably well. So we just went through this unification with SCAA and SCAE and the two, the, are class-based curriculum and they’re more modular, module-based curriculum…we merged those together. So now those individual classes, they kind of don’t really technically exist. But up until that point, the changes that were made between 2012, when I tackled that, and the end of that program were incredibly minor. And I eventually ended up on the Education Committee people would, every now-and-then, bring up possible changes to that class and it was incredibly gratifying to hear other people articulating what was, what they thought was really special about that Profile Roasting class and resisting those changes. But the good thing about this is that now that it’s not part of the certificate program, I have a little bit more freedom to kick it up a notch. And kind of take that material to the next level, which is what I’m doing in that class.

Michael: Yeah, it kind of becomes a Master-class for you.

Neal: Right. Right. And then I’m also helping Anne Cooper with her class; Can You Taste The Roasting System. And that is a really fun class. I helped out with that class at Expo also. And one of the things that I really like about that class is that it covers topics that I’ve been trying to promote in, in the coffee world for a long time. There are features in Typica that directly relate to what she’s talking about in this class.

Michael: Like your profile translation?

Neal: Right. Right. And it’s one thing for me to make a video and talk about the theory and my experience with this, and it’s something else entirely do the experiment yourself, to taste it in the correct way. And this is really important, because a lot of people, when they set up a roasting experiment, they’ll say ‘okay, I’ll put five cups of this and I’ll put five cups of that and I’ll taste what I have.’ And if you do that, you’re going to knit-pick. And even if it’s five cups of this and five cups of exactly the same coffee from the same batch, that you just don’t know it’s the same five cups, you’ll pick out differences.

Michael: Yeah. It has to be truly blind.

Neal: Right and so this is, her class is set up right. It’s blind triangles and we also have triangles where all three cups actually are the same. So you’ve got that test of ‘are you really tasting differences are you just guessing?’.

Michael: Right. It’s a lot like the Q-Grader.

Neal: Because you think there should be a guess.

Michael: Yeah. It’s a lot like the Q-Grader test, because it does something similar. Are you finding, not only are you finding the odd-ball in a triangle, but are you rating consistently the best cups on the table as the best cups?

Neal: Right. Right.

Neal: And so, and having that experience, doing it yourself, and getting the results immediately, it’s a much more powerful presentation. I think.

Michael: Yeah. And I don’t want to spoil it to find out if you can actually taste the difference between the machines.

I’m in that class as well. I’m one of the Station Instructors, so that’ll be a good time.

Neal: Yeah. I think you’ll enjoy it.

Michael: Yeah. Well thank you very much for sitting down with me I appreciate this.

Neal: Sure.

Michael: And we’ll have to sit down again. We can dive into profile translation a little bit more.

Neal: Of course!

Michael: Alright. Thank you.

Neal: Sure.

[End of recorded material]