THE SHORT ANSWER · Decide the hours first. Then pick a channel that fits them.
Almost every solo founder does this the other way round: read about a channel, decide it sounds right, and then try to find time for it. That ordering guarantees the channel loses, because the time was never there and the channel was chosen without reference to it. Two hours a week, chosen in advance and defended, beats an ambitious plan that survives until the first bad Tuesday.
What makes solo marketing hard is not that the tactics are secret. They are extensively documented and mostly free. It is that you are simultaneously engineering, support, billing and the person who has to stop for a while, and marketing is the only one of those with no external party waiting on it. Nothing breaks when you skip a week. That is exactly why it is what gets skipped.
The interrupt is the enemy, not the workload
Consider a normal week. You plan to write on Wednesday. On Tuesday afternoon a customer hits a bug that eats the rest of the day and half of Wednesday. You do not write. Nothing terrible happens. The following week the same thing happens for a different reason.
The total work involved was never the problem — writing one thing a week is genuinely achievable. The problem is that the work was scheduled into time that could be claimed by something more urgent, and marketing will lose every one of those contests, correctly, because a customer with a broken account really is more important than a post.
So the plan has to survive the interrupt rather than assume it away. In practice that means two things: everything is produced in batches, well ahead of when it is needed, and the amount produced per batch is sized for a bad month rather than a good one.
THE PATTERN · Day-of production is the first thing to fail, and it fails silently.
Anything scheduled to be made on the day it goes out will stop the first week something goes wrong, and there will be no alert when it does. A queue that is three weeks deep absorbs a bad fortnight without anyone noticing, including you.
Budget the hours honestly
Pick the row that matches the time you would still have in a difficult month, not the time you have this week because things happen to be calm.
What fits, by how much time genuinely exists
| Time per week | What fits | What does not, whatever you read |
|---|---|---|
| Under an hour | One channel, produced monthly in a single batch | Anything daily, anything conversational, anything live |
| Two hours | One compounding channel plus replies, or near-daily short video if production is automated | Two channels made by hand |
| Four hours | One compounding channel and one direct channel, batched fortnightly | Three channels; the third always dies first |
| A full day | You are not marketing solo any more, you are running a content operation | Assuming this is sustainable alongside support and shipping |
The most common error is choosing the four-hour row while living in the two-hour one, then concluding that marketing does not work.
Two hours is the honest number for most solo founders, and it is enough. It is not enough for three channels, which is the actual reason most solo marketing plans collapse — not insufficient effort, but a plan built for staffing that does not exist.
Where automation helps, and where it quietly costs you
This is the decision that matters most when there is one of you, and it is usually framed too bluntly. The useful line is not between human and automated work. It is between production and judgement.
The line worth drawing
| Task | Automate? | Why |
|---|---|---|
| Deciding what the product is for and who it is for | Never | It is the input everything else depends on, and it takes one afternoon a year |
| Coming up with the week's ideas | Yes | Repetitive, and the constraint is volume rather than insight |
| Writing captions and cutting video | Yes | Pure production; this is where most of the hours actually go |
| Scheduling and publishing | Yes | There is no upside to a human doing this and a real downside when they are busy |
| Replying to comments and messages | Never | The entire value is that it is you. Automated replies are worse than silence |
| Anything reactive or sensitive | Never | The failure mode is public and disproportionate to the time saved |
Read down that table and the shape is clear: the parts that consume nearly all the time are the parts with the least judgement in them, and the parts that need you are short. That asymmetry is the whole opportunity for a one-person company. Automating production is not lowering your standards. Automating the reply is.
Getting three weeks ahead, once
The difference between a channel that survives and one that does not is almost always the depth of the queue. Building it is a one-off cost, and it is the single highest-value thing a solo founder can do for their own distribution.
- Block one long session, not four short ones — Three hours in one sitting produces far more than three separate hours, because most of the cost of this work is starting it. Treat the session the way you would treat a customer call: it is in the calendar and it does not move.
- Produce three weeks in that session — Not one week. One week of buffer is consumed by a single bad fortnight and then you are back to making things on the day. Three weeks is deep enough that you can lose a fortnight entirely and nothing goes quiet.
- Queue it all, then stop looking at it — Scheduled and out of your head. Checking on it daily reintroduces exactly the attention cost the queue was built to eliminate.
- Top the queue back up monthly — One shorter session a month keeps the depth roughly constant. If you miss one, the buffer is what absorbs it — which is what it is for, not a sign the system failed.
Judge it against yourself, not against a team
You will compare your output to companies with a content hire, and conclude you are behind. You are, on volume, and volume is not what you are competing on. A solo founder can say things a company cannot: what the product costs to run, what it cannot do, why you rejected a feature people keep asking for, what broke last week. None of that survives a review process, which is precisely why it is yours.
The practical version of this: when you have one hour and two options, write the thing that would need approval at a larger company. It will outperform the safer thing, and it is the only structural advantage you have.
THE ADVANTAGE · Being one person is a content position, not just a constraint.
Specificity, candour and speed are all things a team is worse at. If your posts read like they came from a marketing department, you have given up the only edge you had and taken on a fight you cannot win on volume.
Your own account or the product account?
This is the question almost every solo founder asks and almost no guide answers directly, because the honest answer is uncomfortable: your own account will do better, and it is the one you are most reluctant to use.
A product account starts at zero with no reason for anyone to follow it, and every post from it reads as marketing, because it is. A personal account carries whatever credibility you already have, and the same sentence posted from it reads as a person describing their work. On the platforms where founders and developers actually are, that difference is most of the reach.
The practical arrangement is both, with different jobs. The personal account carries the narrative — what you are building, what went wrong, what you decided and why. The product account carries the product: the feature clips, the demos, the things a customer would want to see, and it is the one you point new users at. The product account is also the one that can run on a queue without you, which matters when the whole point is surviving a bad month.
The failure mode to avoid is running the personal account as a second marketing channel. If every post from it is a launch announcement, you have spent your credibility to gain nothing, and people stop reading. Post from it as yourself, mention the product when it is genuinely the subject, and let the product account do the selling.
How to decide when to stop something
Solo founders abandon channels too early and cling to them too long, usually in that order and often the same channel. A decision rule written down in advance is worth more than a judgement made in a discouraged week.
- Set the review date when you start the channel, at three months minimum for anything compounding. Write it in the calendar. Do not evaluate before it.
- Decide now what evidence would count. Rising impressions, a handful of trials, one post that clearly outperformed — pick something observable rather than a feeling.
- At the review, check whether the trend moved, not whether the numbers are large. Small and rising is working; large and flat is not.
- If there is no movement at all after two full review cycles, stop properly and pick one replacement. Do not add the replacement and keep both.
- If it is working, change nothing for another quarter. The most common way a working channel dies is being redesigned in month four.
One caveat worth being honest about: if a channel produced nothing, the usual cause is not the channel. It is that it ran for six weeks rather than six months, or that it went quiet for a month in the middle. Before concluding a channel does not work, check whether it actually ran.
FAQ
How much time should a solo founder spend on marketing?
Around two hours a week for most people, chosen in advance and protected, with the work batched rather than spread across days. The number matters less than sizing it to a difficult month rather than a calm one — plans built for the good weeks fail in the first bad one, and the failure is silent because nothing external depends on it.
What should a solo founder automate, and what should stay manual?
Automate production — idea generation, writing captions, making images and video, scheduling and publishing. Keep judgement manual: what the product is for, who it is for, replies to comments and messages, and anything reactive or sensitive. Production is where nearly all the hours go and where the least judgement is required, which is what makes the split work.
How many marketing channels can one person actually run?
One, if it is made by hand. Two, if production is automated or the second channel is a one-off like a launch. Three is where solo plans reliably break, and the third channel is usually the one that dies first and takes the other two down with it, because failing at something weekly is demoralising in a way that makes the whole effort easy to drop.
How far ahead should I schedule content?
Three weeks. One week of buffer is consumed by a single bad fortnight, which puts you back to producing on the day — and day-of production is the first thing to fail when a customer escalates. Three weeks absorbs a lost fortnight without the account going quiet, and rebuilding it takes one session a month.
Is it worth marketing before the product is finished?
Yes, because distribution has a delay that building does not. Search takes months to return anything and a social account takes weeks to compound, so starting when the product is ready means starting the clock late. The smaller risk is talking about something unfinished; the larger one is launching into silence.
How do I know whether a channel is working?
Set the review date when you start — three months minimum for anything compounding — and decide in advance what evidence would count. At the review, look at whether the trend moved rather than whether the numbers are big: small and rising is working, large and flat is not. Before concluding a channel failed, check whether it genuinely ran without a gap.
Related reading
- Social media for indie hackers — Built for one person: keeps posting through the weeks there is no time for it.
- Marketing a SaaS with no marketing team — Choosing the two channels, and where the material comes from.
- Promoting an app without paid ads — Why the store converts demand but never creates it.
- Social media on autopilot — The full picture of what can and cannot be automated.
Originally published at postotter.app. I build PostOtter, which writes and publishes a brand's Instagram, TikTok and X posts automatically. There is also a free MIT package, @postotter/video, that renders short-form video from your own app screens and plans a week of posts in your repo.
