Roast #30: Reflection on V4
const subtitle = null; // I can't think of a good subtitle...
If you have any broad feedback or answers to questions below, I’m collecting suggestions in this google form. Also, if you bought V4, please consider filling out the V4 feedback survey. Thank you!
Some quick numbers:
Final sold V4 box count: 54 boxes (vs. 46 boxes in V3) // 17% increase
Ran 52 roasts for V4 (vs. 92 roasts in V3) // 43% decrease
Delivered 58lbs of roasted coffee (vs. 40lbs of roasted coffee in V3) // 45% increase
Note — the V3 figures differ from those shown in other posts due to a simpler way of interpreting the data.
Onto the fun stuff.
Things that could’ve gone better:
1. I need to upgrade my setup
When roasting, other than the actual drum roaster itself, there are a few additional components that all need to be running as well.
These other components are:
A fan underneath the circuit board to keep it cool
An external inline fan controller to exhaust the fumes from the roaster
Proper alignment of the roaster exhaust port and the external fan.
The first point is insignificant as it doesn't affect the roast. If the fan underneath the circuit board isn't cool, the roaster will emit error codes and won't run.
But for the other two points, if neither were properly set up, the resulting roast would be different than what was intended.
There was an instance when the external fan wasn't aligned with my exhaust port, so the fumes spewed all over my apartment. For a 730g batch, the amount of smoke (during development) is a lot; you could visibly see the thick grey smoke spewing out of the roaster.
Then, there was an instance where the external fan was aligned but on an ultra-low setting. As the roast progressed, I noticed the temperature at key intervals was higher than it should've been. I couldn't figure out why and assumed it was just due to increased ambient temperature. Then, when I dropped the beans, I noticed the drum was way smokier than usual. That's when it occurred to me that maybe the increase in heat was due to poor ventilation. And then I looked at the external fan setting... and yeah, that was it.
For both instances, I had to toss out the batch or recall the batch. Fortunately, these instances were rare, but reducing that to zero would be best.
Action Items:
I'm not sure as some of these issues are procedural. Ideally, I'd like to have my ventilation components connected to my roaster and enforce a dependency that one can't run without the other. But I'm not savvy enough to figure that out with my little setup.
The way I see it, this is a temporary problem. When I move to a #bigboi production roaster, everything (mainly ventilation) will be hooked up, and all should be well.
2. Shipping experience could've been better
Life got hectic for me in August. The buildup of non-coffee priorities made it hard to plan and execute thoughtfulcoffee milestones. This resulted in service degradation, specifically for folks who opted to have their orders shipped.
I analyzed my fulfillment period, and the average fulfillment period specific for shipping orders was 31.36 days. This was the average, meaning orders on the upper bound were even more delayed.
This metric likely is worse than it seems since I opened orders in late July, and for most of August, I was still dialing in and planning, so I wasn't ready to fulfill orders.
Nonetheless, having such a long delay wasn't a great experience.
Action Items:
I'm going to be stricter with myself. I'm going to be more explicit with the process. I'll give clearer dates and deadlines for when you should receive your order.
I'll also release orders closer to delivery. This time, I opened delivery in late July, which may have been too early. I thought opening V4 early could pull in more orders due to the extended buying period. This thinking might've been true-ish, but the tradeoff was not worth it. More than half of the orders came in the first week. So, a shorter but more intensive buying period is the better path to go down for future versions.
3. Better communication on my side.
There were three instances of people reaching out and asking for clarification about the specifics of V4. There probably were a lot of others who were also confused and never asked.
I'd be open to help on this one (anonymous form for suggestions). If anyone has ideas or thoughts on what I can do to fix up the text, or anything additional I can add to make the text clearer, I welcome your thoughts/ideas. Thanks!
Action Items:
Be more explicit in my product description/marketing/communications about what's offered and for how much.
Give clearer timelines for when items should be expected to be delivered.
4. More care in designing the 240g packaging:
Not a huge thing, but I could've selected a better box for the 240g batch. The box was too big, and the bags tossed around in the box. It was an awkward experience.
This error happened because I assumed the coffee bags would fit snuggly in the box. I discovered they didn't when I had to pack the first 240g order.
Action Item:
Better planning. Once the materials come in, don't assume that things will automatically fit and look good.
Things that went well:
1. Growth:
The numbers again (same as in the opening):
Final sold V4 box count: 54 boxes (vs 46 boxes in V3) // 17% increase
Ran 52 roasts for V4 (vs 92 roasts in V3) // 43% increase in efficiency
Delivered 70lbs of roasted coffee (vs 40lbs of roasted coffee in V3) // 75% increase
Number of boxes delivered:
We saw some growth in quantity sold, but I wish we saw more.
This signals that I'm hitting an upper bound with how I'm marketing the product. I think I can fully confirm this hunch in the next round if I hit another ~50 orders.
I've heard some feedback that people are hesitant to buy because there aren't any product photos (very valid). I don't want to spend too much time thinking about this as I feel it doesn't directly contribute to the end product, but it is something to work on if I want to move this project forward.
I’ll try and think through new channels to market this. I’m sharing the box offering through neighborhood subreddits, 2 Discord servers (EAF + NYCoffee), and on my Instagram. I guess partnerships might be a reasonable next step. Anyway, I’d love some thoughts on better marketing this project.
Total weight roasted:
The total weight roasted is 70 lbs this time vs. 40 lbs. This increase was a big surprise to me, and I am very thankful to all of you who mentioned "bigger quantity" in the V3 feedback form. The idea of a 240g honestly never popped into my mind.
I was pigeonholed in my way of buying coffee — buying for one. If I had expanded my thinking to consider buying for a household of 2, it would’ve been obvious a larger quantity makes a ton of sense. Assuming each cup requires 15g, and both individuals drink two cups daily (15g * 2 people * 2 cups = 60g), they'd finish a 120g bag in two days, which is a super short time.
Increase in efficiency:
So, this was a massive win for me. Fulfilling orders now doesn't mean I must block out three full days. Now, it's kind of just one to two nights. For this round, most roasts were 730g batches. Enforcing this practice doubled my output.
I dream of running only three roasts, one for each bean, and then being done with that version's coffee club.
With more consistent, predictable demand, determining when to make this jump becomes clearer. Suppose I could sign up enough subscriptions/wholesale orders to fulfill 150 lbs. In that case, the decision becomes clear as I could deliver the coffee immediately with no risk of roasted coffee going stale/unsold. As is, demand could still fluctuate, so investing in the training/logistics of a co-roasting facility doesn't make sense yet.
2. The process is coming together:
Things feel more stable, and execution feels less variable.
Some notable processes that I instilled this time around:
I came up with a system to stamp my boxes: I came up with this out of frustration. For the first three boxes... my stamping attempts were all misaligned and were too ugly to use. They were so misaligned that I couldn't bear the embarrassment of handing them to someone. From this came the push to figure a system out. I came up with using tape to act as markers to help dictate how I should position the box and where the "center" (to stamp) should be.
Coming up with a versioning system for my batches: If you look at the bottom of the bag, you'll see a number. At the moment, it's:
`{target_roast_number}.{batch_number}` (ie 6.11)The `{target_roast_number}` is the profile I was targeting, and the `{batch_number}` was what batch that roast was. This change helped provide an identification system for each bag of beans.
Greens were vacuum-sealed until usage: I didn't feel any of my greens started to fade for V4. If there were any inconsistencies, it was usually due to roast level rather than fade. Resealing the greens into 4-6lb increments after arrival was a hassle, but the effort was worth removing the fade factor.
Coffee sourcing was better: I think better sourcing comes with experience. I now have a profile that I’ve found to work well, and I feel better adjusting my sample roast profile depending on the green. I also have a better gauge of which exporters have coffees that fit my profile. Coffee green buying is coming together, and I hope to continue building relationships with green exporters.
3. The Villager Popup:
I really like this photo. I love the visual chaos of this photo, the countertop littered with caraffes, cups, and scattered papers. And I love the context of what's happening — Joseph (@curiosocoffee) sharing with us the last of his coffee, Subrat (@astroploidy) elaborating on a coffee experience, and me just laughing. This photo is a good portrayal of a side-effect of this project: bringing people together through coffee.
We sold 39 pour-overs over 3 hours! Before ending, I/Toni (@tonidelsorbo) /Maria (@kitsmaria) guessed how many pour-overs we sold, and I believe our answers ranged from 50-70. So all three of us definitely felt that it was busy.
There were moments when we'd have 4-5 drinks backed up. I remember Ben coming by and saying they would start letting customers know the wait time would be upped from 5-7 minutes to 15-20 minutes. Then there was also the moment when we ran out of carafes to brew into, as they were all being used by the customers, so we started to brew into a milk pitcher.
Overall, the event was a lot of fun, and it was great being back behind the bar again!
Thank you Toni and Maria for helping me serve thoughtfulcoffee that day. I know I wasn't the best barista that day (I kept forgetting to keep my area organized/clean 🤡) so I appreciate you both dealing with me.
Thank you to the Villager (@villager_ny) staff for supporting us throughout service — pointing us to where things were, helping us bus out the drinks, and helping keep the glassware consistently washed and stocked.
I treasure our relationship, and I can't wait to run it back again in the future. <3
4. Side-effects of thoughtfulcoffee:
I'm not sure how to best title this section, but I really enjoy hearing how thoughtfulcoffee has affected you and how you all share thoughtfulcoffee with your networks.
Some notable examples:
Learning that I helped some of you get deeper into coffee. It was really cool to hear that some of you went out and bought a Hario Switch brewer so that you could try the brewing recipes I came up with. It's awesome to know that I can help usher some of you deeper into coffee. 🥲
My coffee making its way into the workplace. I know of... 2 instances of this so far. And in one instance, a coworker of the original buyer reached out and bought a box!
Then, I think the most interesting point is a developing increase in presence on the West Coast. More on this below.
So, Mason, an internet stranger bought a box because he saw my post in #advertisements in EAF. I think he then shared V4 with his group of coffee friends in the Bay Area, and recently, I've seen an uptick of photos of my coffee in West Coast kitchens.
Which is awesome!! I grew up on the West Coast (specifically Fremont, CA), and setting up a presence there has always been a goal. Since starting something in NYC was more realistic (as I live here), I've never put any effort into starting something on the West Coast, but a West Coast presence seems to be organically happening.
Anyway, I'm secretly hoping that I'll be able to serve more "West Coast"-ers in the future, as then it'll give me another excuse to visit more often. #510forlife
Final words:
And lastly, just hanging out and hearing about how you all are doing was a real joy.
David was in Seattle recently, Edmund is enjoying his new role at Datadog, Sylvan was getting surgery (I hope it went well!), and the list of updates goes on.
It's kinda sappy, but I feel an attachment to all of you (my customers). You all support me on this "start a cafe/roastery" journey, and that means a lot to me. I’ll treasure this attachment while I still can since I know I won’t be able to build a similar relationship if/when this project expands (not scalable :|).
It was a pleasure fulfilling coffee club V4 for you all, and I hope to see most of you back for V5. It’ll be another fun one. 🥲
hey that's me in the screenshot!
- fellow Brian :)