Most businesses that stop producing content do not stop because it failed. They stop because production depended on someone having ideas and spare time in the same week, and that stopped happening.
This is an operations problem wearing a creative costume. The businesses that publish consistently are rarely more creative than the ones that don't. They have a process that does not require inspiration to function.
Separate the four jobs
The single most useful change is recognising that content production is four distinct activities, usually collapsed into one:
Deciding what to make. Choosing the subject, based on something other than what occurred to someone that morning.
Making it. Writing, filming, designing.
Preparing it. Editing, formatting, adding whatever the channel requires.
Publishing and distributing. Getting it out, and then getting it in front of people.
Collapsed together, the whole thing stalls if any one part stalls. Separated, they can be done by different people, at different times, in batches — and a delay in one does not stop the others.
This is the same logic as any operational process. The failure is that content is treated as an act of expression rather than as work with steps.
Deciding what to make: a source, not a mood
The most common reason production stops is running out of ideas. That is a symptom of having no source.
Reliable sources, in rough order of usefulness:
Questions customers actually ask. Before buying, during, and after. These are already validated demand — someone asked. Keep a running list; anyone customer-facing can add to it.
Objections you handle repeatedly. The reasons people hesitate. Answering them publicly does the work in advance.
Decisions customers must make. Comparisons, sequencing, what to do first. These are searched heavily and answered badly.
Things you know that are not obvious. Genuine expertise from doing the work. This is the category that earns links and gets referenced.
What is already working. A post that performed, a page getting traffic — extend it rather than starting fresh.
A business that maintains a list from these sources does not run out of subjects. It runs out of capacity, which is a different and more solvable problem.
Batch the making
Producing one piece at a time is the most expensive way to work, because every piece pays the full cost of context-switching, setup and getting started.
Batching by activity — a session for filming, a session for writing, a session for editing — reduces that overhead substantially. For video especially, the difference between filming one and filming six in a session is not six times the effort.
The practical consequence: a business that can commit to one focused half-day a month can produce meaningfully more than one attempting something every week and managing it half the time.
Publishing is not distribution
The most common gap. A piece goes up and the work is considered done.
Publishing is making it available. Distribution is getting it in front of people — and it is usually where the effort should be weighted, because a good piece nobody sees performs identically to a bad one.
For most businesses this means: sending it to people who would find it useful, linking it from relevant existing pages, using it in sales conversations where it answers a question, and re-sharing it more than once over months.
The related failure is treating every piece as disposable. Something that answers a real question is useful for years. Businesses that publish and move on are rebuilding their audience from zero every time, which is the rented-versus-owned distinction in distribution is a system, not a budget.
Volume is not the goal, and it is not the metric
A commitment to weekly output produces forty pieces and, frequently, negligible results — because volume was the target and relevance was not.
The businesses that get results from content usually publish less than they expected to, aimed more carefully. Three pages that answer real questions outperform thirty that answer none, and they keep working.
The corollary: measure whether pieces get found and whether they produce enquiries, not how many were published. Output is an activity metric. Which marketing numbers are worth tracking covers the distinction.
Where AI fits, and where it does not
Language models change the cost of producing text substantially, and it is worth being precise about what that does and does not solve.
It helps with: drafting from an outline you supplied, restructuring, editing for length, producing variants for different channels, and getting past a blank page.
It does not help with: knowing what to write about, or the original knowledge that makes a piece worth reading. A model does not know what your customers ask you, or what you have learned doing the work.
The failure mode is using it to solve the capacity problem while ignoring the source problem — which produces volume with no original value, at scale. That is a large amount of content that will not rank and would not deserve to.
Our position, applied to our own work: use it to draft and restructure; never to supply the substance. Everything that makes a piece worth reading has to come from actual experience.
What we have seen
Truffaire builds its own products and had to distribute them, which meant solving this before offering it to anyone.
The finding that mattered most: the constraint was almost never ideas. It was that production depended on the same people who were doing the primary work, so it was always the thing that slipped. Separating deciding from making — so that the list of subjects was maintained continuously and production could be batched — is what made it sustainable.
The second finding is that the pieces which performed were consistently the ones answering a specific question plainly, not the ambitious ones.
Frequently asked questions
How often should we publish?
At a cadence you can sustain for a year. Monthly consistently beats weekly for two months and then nothing. Consistency matters for compounding; frequency matters much less than people assume.
Should we hire a writer or do it internally?
The knowledge has to come from inside. A writer can handle the making and preparing if someone internal supplies the substance — which is usually the workable split, since the constraint is rarely writing ability.
How long should pieces be?
As long as needed to answer the question and no longer. Word-count targets produce padding, and padding is visible.
What about repurposing across channels?
Efficient and worth doing, provided each version fits its channel rather than being the same text pasted everywhere. One idea, several appropriate forms.
How do we know if it is working?
Whether pieces get found, and whether they produce enquiries. Give it months rather than weeks — content compounds slowly, and judging it on a four-week window will lead you to stop just before it starts working.
Where to start
This week, start the list. Every time a customer asks something, write it down. Do not filter it.
In a month you will have more subjects than you can produce, sourced from actual demand rather than from guessing. Then block one half-day and batch-produce three of them.
That is the entire process. Everything else is refinement.
If you want help building the production system rather than the pieces, talk to us.