Month Two: Picking a Roaster, and What Due Diligence Looks Like

Month two of the build-in-public log. In month one we committed to a brand, bought a domain, and opened the waitlist. This month the question was bigger and less obvious: who is actually going to roast this coffee?

HexRoast is not (yet) a roaster. We do not own a drum roaster. We do not have a production space. What we have is a brand, a design system, a small audience, and strong opinions about what we want in the cup. The job in month two was to find a roasting partner who could execute those opinions at a scale we could afford and a quality we could defend.

/ 01The shortlist, and where it came from

We started by asking three questions and letting the answers narrow the field:

• Who is currently making coffee I would drink every day? Fifteen roasters on the initial list, drawn from our own tasting history.
• Of those, who operates at a scale that can handle a small private-label partnership without either absorbing us or being overwhelmed? Down to nine.
• Of those, who is close enough geographically for realistic logistics and site visits? Down to seven.

Seven is a manageable number. We grouped them roughly:

/ 02The tasting rubric

We designed a simple scoring sheet before visiting anyone. Every roaster on the list would be evaluated on five axes:

1. Cup quality, at our target roast levels. Each roaster brewed us their interpretation of the three products we care about: a dark roast, an espresso blend, and a washed single origin. Scored 1-10 on each.
2. Transparency of sourcing. Did they know the farm? Could they show us the green coffee invoices? How much did they pay per pound of green?
3. Production flexibility. Minimum batch size, turnaround time for a custom roast profile, willingness to run our specific blends as recurring products.
4. Pricing, landed. Not just per-pound cost. Including packaging, labeling, and fulfillment if they offered it.
5. Human chemistry. We explicitly included this as a scored axis. Working with someone for years requires actually enjoying working with them.

Each axis weighted equally. Scores tallied per visit, then compared across the seven roasters at the end.

/ 03What actually happened on the visits

Each visit was a cupping session at the roastery, followed by a 45-minute conversation. We visited five of seven in person; the remaining two sent samples and did a video call. (Roaster 6, the farthest, offered to fly samples overnight, which we appreciated but ultimately lost a point on "human chemistry" because it made them feel remote.)

A few observations across the visits that we did not expect:

• The most expensive roaster was not the best. Roaster 1, the high-end established option, scored highest on cup quality but lowest on flexibility — they were already over-committed and treated our project as a scheduling problem rather than a partnership.
• The cheapest was not the worst. Roaster 7, emerging, scored surprisingly well on cup quality (a real Ethiopian natural that outperformed every other one we cupped) but struggled on production flexibility. Minimum batch was larger than we could commit to.
• Human chemistry was a better predictor than we expected. The roasters we left visits feeling excited about correlated almost perfectly with the ones who turned out to have thought deeply about their craft. The one visit where chemistry was flat, the cup was technically fine but the conversation revealed they had never thought about their roasting choices the way we wanted our partner to.
• Transparency was the easy filter. Roasters either knew their green coffee invoices or they did not. Two of seven did not want to show them. Both left the shortlist.

/ 04The spreadsheet, in summary

The final scoring, averaged:

/ 05The decision

We signed a letter of intent with Roaster 3 this week. They are the mid-career option who topped our spreadsheet — a seven-year-old operation of four people with an unusually strong point of view about washed Ethiopians and a drum roaster they have clearly spent time dialing in. The cupping session produced the best espresso blend we tasted, and the conversation afterward was the one we most wanted to continue.

The commercial structure: private label, our brand, our packaging, their roasting. 500-pound minimum per blend, which we can afford given our current waitlist numbers. 3-week turnaround on the first production run, 2-week on subsequent runs. Pricing lands us at a healthy margin at our target retail price.

We will not name them publicly until the first production run ships, which is standard for private-label partnerships. But they will get credit on the bag (small, clean, near the roast date stamp), which we think is the right answer both ethically and commercially.

/ 06The thing we almost missed

In the second-to-last interview, we asked a question that was not on the rubric: "What is a roast you would not do for us, even if we paid you?"

Roaster 3's answer was the one that won us over. "I would not do a dark-to-burnt French roast, the kind that tastes like ash and sells well in the diner tradition. I think the demand for that exists because people were raised on it, and I get it, but I think a well-developed medium-dark is better at everything a French roast is trying to do. I would rather lose the account than roast bad coffee."

That is the answer of someone with a point of view. The answer of someone we want to work with for years. The criteria we did not put on the rubric but should have: does this partner have opinions they are willing to defend?

Next month: we figure out crowdfunding. Kickstarter vs direct pre-orders. More numbers. More decisions in public.