What We Do With Coffee That Doesn't Sell (and Why We Don't Run Clearance Sales)

We don't run sales. That's a decision we've made more than once and explained in more than one post, and the short version is that discounting a coffee teaches people the coffee was worth less.

But underneath that decision there's an operating question nobody writes posts about: what happens to the bags that don't sell?

Every roastery has this. It's the boring half of the business. A coffee gets a slot on the shelf, someone in the shop falls out of love with it, or a competitor's coffee shows up at a lower price, or one batch just tastes flatter than the last one, and after eight weeks the bag has moved four units. Then what?

Here's the honest version of what we do, including the part we haven't solved.

First: the size of the problem

We're small, and small roasters actually have a specific version of this problem that big ones don't.

We roast to order, Wednesday, ship Thursday. We don't hold a warehouse of finished product. So we don't get stuck with a lot of roasted coffee that ages on a shelf going nowhere. What we get stuck with is green coffee we committed to but didn't sell — which is a bigger problem than it sounds, because green coffee doesn't spoil the way roasted coffee does, but it absolutely stops tasting good, and it ties up money that's sitting in a warehouse somewhere instead of in your account.

The trigger is usually one of three things:

The coffee is fine and we just under-bought. A coffee that reliably sells 60 bags, we bought 90 bags of, and we sold 58. No operational failure. Just a bad estimate.

The coffee is fine and we under-marketed it. A new arrival that got two blog posts and four pins while the one below it got six posts and eleven pins. Allocation of attention, not quality.

The coffee isn't fine. A lot that cupped well and then didn't, or a bag that arrived with a defect, or an origin we bought enthusiastically and are now finding is not as good as the sample suggested. This happens and it's uncomfortable.

What we actually do, in order

1. We wait, honestly. For the first six to eight weeks we do nothing except keep roasting it available. Some coffees are slow starters — they need a few people to taste them and talk about them before they catch on. The bag that sat for eight weeks and then sold forty units in three weeks is a common outcome, not a fluke.

2. We re-roast the profile. This is the move that works more often than people expect. A coffee that tasted flat on its first pass might just need a slightly different development time, a different charge temperature, or simply to be roasted a bit lighter. We re-cup and re-profile before we decide anything else. It's cheap and it's honest, because if the coffee can be fixed by roasting it better, the problem was us.

3. We put it in our own cups. E drinks a lot of our own coffee. So do we, when we're at the roastery. There is a real volume here — a few pounds a month across the team — and it is genuinely useful to have coffee that we ourselves would happily drink and can't sell. That is not a solution. It's a bridge.

4. We bundle it. This is the one that takes real thought. Rather than discounting the slow coffee on its own, we bundle it with something that sells. A mixed bag for people who don't care which origin, an add-on included with a subscription order, a sample included in an order of a faster mover. The slow coffee gets consumed by people who weren't going to think hard about it, and the fast coffee carries the margin.

We don't do a lot of this. Bundles are easy to fake — everyone else is doing bundles too — and we'd rather keep the catalog legible.

5. We work with our rescue partners on it. Coffee that's still perfectly good and just not moving is a reasonable thing to send to an organization. We support our rescue partners with a monthly share of our profits, and that is a separate thing from the coffee. If we have coffee that's a few months old and still good, we're glad for it to go somewhere useful rather than into the trash. We don't publish a number for this and we don't plan to start — the specifics live on our impact page — but it happens and it's one of the better parts of the week.

6. We drop it. Some coffee just doesn't work for us. It sold three bags in three months and it wasn't the coffee's fault or ours, it was a mismatch with this catalog at this moment. We take it out of the lineup and don't re-buy it. Every catalog has to have a way to end a relationship, or the shelf fills up with things nobody wanted.

What we don't do, and why

We don't run a discount. Not a clearance, not a flash sale, not a "last bags" fire sale. This is the decision people ask about, so let's be precise about the reasoning.

A discounted bag tells the customer two things. One: this coffee is worth less than the price on the bag. Two: there's something to be gained by waiting for a discount. The second one is the expensive one. Once a customer learns that waiting is rewarded, every sale you ever run costs you money on the customers who would have paid full price.

We'd rather a bag cost us the $12 in coffee and $3 in packaging than teach a repeat customer to wait for the next drop. If you buy something from us at full price, you're never going to wonder if you overpaid.

We don't put it in a mystery bag. Mystery bags are just discounts with extra steps and a surprise element.

We don't lie about why it's discounted. No "manager's special," no "roast error." If we ever sell coffee below cost, and we don't, we'd tell you exactly why.

The part we haven't solved

The honest answer is that our dead-stock problem is better than it was and worse than we'd like.

Better, because roasting to order means we're not sitting on finished inventory. Most of what we're stuck with is green, and green coffee that's been stored correctly for a few months is still good. We can come back to it next season if the coffee deserves another shot.

Worse, because we don't have a real answer for the case that actually matters: a bag of roasted coffee that we roasted last Wednesday for a coffee that isn't moving, sitting in a bin with maybe six weeks of freshness left in it. For that six weeks, it's a coffee we're happy to sell and nobody wants. After that, it's not a coffee anymore, it's inventory with a problem.

We could roast less of it. We could buy smaller lots of the slow movers. Both are real options and we've argued about both. The tension is that smaller green lots cost more per pound, and the whole reason we buy from a given origin at all is volume.

So this is an open question in our operation, not a solved one, and I'd rather tell you that than write a post pretending we have it handled. When we figure it out we'll say so.

Why we're writing this at all

Because "how we do sales" gets covered everywhere and "what happens to the stuff that doesn't sell" almost never does. It's the unglamorous part of running a small company and it's most of what actually takes the time.

If you've bought a coffee from us and thought "why isn't this one always in stock," the answer is sometimes that it sold out and sometimes that we stopped roasting it, and this post is about the second one.

The takeaway

We don't run discounts because they'd cost us more in trained waiting customers than they're worth in recovered inventory. Slow coffee gets more time, a re-profile, our own cups, a spot in someone else's bundle, or a trip to a rescue partner before it gets dropped. The roasted-version of that problem — coffee we already made that nobody wanted — is the part we still haven't solved properly.

If you've got a good system for it, we would genuinely like to hear about it.

Brew slow.

— E