How I Decide What to Write About (Most of It Comes From Actual Emails)
I keep a document I have never shown anybody. It is a running list of questions, most of them half-written, in no order I would defend. About half of what we publish started as one line in that file, and almost none of it started as a keyword.
People ask how a small company decides what to put on its blog and the honest answer is that it is much less systematic than it looks from the outside, and also more systematic than "whatever I felt like." Here is the actual order I work through, roughly from top to bottom.
Where the list actually comes from
1. A question someone actually emailed us. This is the top of the list and it is not close. I answer our customer emails myself — that is the whole operation, not a touchpoint I delegate — so the questions arrive in my inbox in their own words rather than in a marketer's paraphrase of them. If one person asks something, it gets a good email answer. If the same thing comes in from three different people, it becomes a post.
The reason that rule works is that a question someone actually typed is already tested. Nobody invents a question at 6am that they do not have. A keyword tool can tell me what people search for; an inbox tells me what people are stuck on, which is the more useful of the two.
2. A brew problem I hit myself that I could not find a straight answer to. This is the second source and it is where my own brewing history fills in. Sometimes I go looking for an answer I expected to find and simply could not — the forums were confidently wrong, the videos contradicted each other, and the sources all skipped the step I was stuck on. Those gaps are posts, because I wrote the answer I needed to read.
There was a stretch where my pour-over tasted sour and thin no matter what I did to the grinder, and every explanation I found online assumed you had your dose right. I did not have my dose right. I had been eyeballing it long enough that it had become a habit, and the habit was well under where it should have been. The only thing that fixed it was putting the coffee on the scale and realizing I had been patiently adjusting a grind setting to solve a problem that lived in the dose. That is the shape of the posts that come from this bucket. Not expertise I already had — a specific thing I was wrong about that a lot of other people are probably also wrong about.
3. A coffee or tea we just bought that I can taste but not yet explain. This one is straightforward. Something arrives on a Wednesday, I cup it, and there is something in the cup I can feel but cannot yet name. If I cannot say what it is, I am not putting it in a product description either. Writing the post is how I find out. Why we cup every coffee twice is really about this — the second cupping exists partly to make me say the thing out loud.
4. A decision we made that I would want to be held to later. Decisions age better than opinions. Whatever we decided about shipping, about how we roast, about what we will and will not sell, I would rather write down why at the time than reconstruct it a year later from memory while being slightly more favorable to myself. What I would tell someone who wants to start a coffee company and the hardest part of running a small coffee company are both this category.
5. The region, the process, the history. Last, and deliberately. I will get to why in a minute.
The ones that need me to do something first
Some posts cannot be written on the day the question arrives, and I have stopped pretending otherwise.
When somebody asks why their tea tastes like metal, or why two coffees from the same country taste nothing alike, the answer usually requires doing something rather than recalling something. Those questions mean brewing six cups side by side, or roasting the same coffee two ways and tasting them blind days apart, or reading enough about a process to be able to explain it without making it up. That takes weeks, not an afternoon.
The order I have landed on is this: the email goes out the same day, with what I honestly know. The post waits until I have actually done the work. If I published first and researched later, I would be writing the version of the answer I hoped turned out to be true, which is why how we choose which green teas to stock starts at the tasting table and not at the flavor notes someone wrote on the packet.
There is a scheduling consequence too. The editorial calendar has posts on it weeks ahead of where I actually am in the work, so the gap between "the question arrived" and "the post is live" is normal and not a sign of anything going wrong.
The one I over-produce
Origin, process, and history is the category I over-produce, and it is the one that would quietly take over the blog if I let it.
It is the most satisfying thing to research. A washing station, a drying patio, a harvest calendar, a co-op that got its electricity in a particular year — there is no version of that rabbit hole that does not eat an afternoon. And none of it is what makes someone buy a bag.
The rule I have landed on is that it gets written when it connects to something I actually stock, and it gets cut when the connection is thin. If I would be writing about a region for the sake of the region, it waits. It is not that history does not matter — it is that a post nobody lands on from any real question does not do anything except make the archive look impressive.
This is the part of my process I would call the weakness, not the strength, and I would rather say that here than have someone notice it in the post archive.
What we deliberately do not write
1. Listicles built to rank. Not "the 9 best pour-over tips" written for a machine. They do not teach anybody anything and they age badly, because the tips in them stop being true as grinders and brewers change.
2. Anything about coffee that would embarrass us if we were wrong. If we have not done it ourselves, we do not write it. Not "according to some sources." Not a technique a roaster mentioned once. If a claim is going in a post, I have brewed it or measured it or read enough primary material to stand behind it, and where I am still unsure I say I am unsure. That is a smaller blog than the competition. It is also a blog that will still be true next year.
3. Anything where I would have to publish a number I cannot verify. This applies to us as much as to anyone else. I do not publish figures about our own business that I cannot check, and the same rule runs through everything we say about the rescue partners we support — the specifics live at /pages/our-impact where they are maintained, and the blog stays honest about what I know. A number I cannot verify is a liability, not a flex.
The honest part: some of it is SEO
I would be lying if I said the search side did not matter, so let me just describe it accurately.
We plan around keywords. I keep an editorial calendar, I look at what people are typing, and I have a rough sense of which questions have real search volume behind them and which nobody has ever searched for at 6am. That is a real part of how I decide what to write and it always has been.
Where it sits in the decision is the part that matters. A keyword is the tiebreaker, not the reason. If two questions are both worth writing and one of them happens to be the phrase people search for, that one goes first. Neither one gets written only because of a phrase. And I do not chase volume — a post that answers one real person's actual problem at 6am is worth more to me than a post that lands on a head term, and the difference between those two posts is not visible in any ranking report.
The takeaway
The list is emails first, my own stuck points second, then whatever I am standing behind at the time, and origin stories last so they do not crowd out everything else. Some of it is SEO and I am not going to pretend otherwise, but the keyword breaks the tie and never makes the case. If you have ever emailed us a question and wondered if it turned into a post, it might have.
Brew slow.
— E