Publishing makes a programme available; it does not make anyone aware of it. The first audience comes from people you can name — contributors, the organisations in the piece, and communities where the subject already lives — approached individually rather than by broadcast posting. After the first fortnight, almost everything comes from being findable and from making the next one.
Flow diagram of 5 steps: Write the one sentence first, Send it to everyone who is in it, a day ahead if you can, Make the piece findable before you drive anything to it, Write ten individual messages, not one group message, Post where the subject lives, then stay for the replies.
Kind Channel is new, nothing has aired, and there is no audience here to release a finished piece into. There is no promotion desk, no mailing list with subscribers behind it, and no front page with traffic to pass on. Everything below is the general practice of getting work seen, written down because a maker publishing anywhere has to do this themselves — including, for now, here.
Uploading the finished file feels like an ending, and it is the point at which a great many first programmes quietly stop. What the upload actually does is change the piece from something on a hard drive into something at an address anyone could reach. Nobody is reached by that. Awareness is a separate job with its own hours and its own materials, and the reason it catches people out is that it looks as though it ought to be automatic — the platform has a homepage, a recommendations rail and a search box, and it is a natural assumption that some part of that machinery is now pointed at you.
It is pointed at signals you do not yet have. Recommendation systems are amplifiers of a response that already exists: before they show a piece to strangers they want evidence that people who started it stayed, finished it, came back. A piece with no initial audience produces no evidence, and no evidence behaves identically to poor evidence. This is the mechanism behind the otherwise mystical advice about the first few dozen viewers. They matter not as a number but as the only information the system has to go on, and they have to be found by hand.
The practical shape of the job surprises people more than the principle does. A reasonable working ratio for an independent producer is to spend something like a fifth to a quarter as long on distribution as on production: a twenty-hour edit earns four or five hours spread over the following fortnight. That feels wrong the first time, because the editing hours are obviously the real work and the distribution hours feel like self-promotion. The alternative is a good piece nobody has encountered, which is the most common outcome in community media and almost never a quality problem.
It is also worth deciding in advance what would count as this having gone well, because the default measure — a view count — is the one least connected to why most community pieces are made. If the purpose was that a particular neighbourhood, or the people in a specific situation, would see it, then whether they did is the question, and forty of the right people is a success that three thousand of the wrong ones is not. Writing that down before publication protects you from judging the piece by a number you never actually cared about.
The instinct after publishing is to post the link somewhere large and wait. This almost always disappoints, and the reason is structural rather than a matter of the wording: a broadcast post is shown to a fraction of an audience that is not expecting it, arriving among a hundred other things, from an account with no track record of this kind of work. The response rate is low because everything about the format makes it low.
The higher-yield move is unglamorous and slightly uncomfortable. Write to people individually, by name, with a sentence about why this particular piece is for this particular person. Twenty such messages will out-perform a post to a thousand followers, consistently, and they are also the only route that produces replies you can learn something from. Start with the people in the programme and the organisations they belong to: they already care about the subject, they frequently have a newsletter or noticeboard that you do not, and they have a reason to want the piece seen that is independent of you.
Whatever the route, give people something to forward rather than a request to do work. A line they can paste, a still image, and — where it applies — a link that starts at the moment they are in. "Please share" asks somebody to compose the message themselves, which is why it so often produces warm agreement and no action.
Routes to a first audience, what each realistically delivers, and its cost
| Route | What it realistically does | The catch |
|---|---|---|
| A personal message to one named person | By some distance the highest response rate available, and the only route that reliably produces a reply rather than a metric | It does not scale, and it degrades the moment it is automated — a personalised template reads as exactly what it is, and it spends goodwill you will want later. |
| Contributors and the organisations in the piece | Reaches a group that already cares about the subject, often through a mailing list, staff channel or noticeboard that is closed to you | Send them the finished piece before you ask for anything, and accept a refusal without pressing. Their willingness to be associated with it publicly is a separate consent from their agreement to take part. |
| A community where the subject already lives | Puts the piece in front of people who sought the subject out, which is a far better-qualified audience than any general feed produces | Most such spaces treat link-dropping as spam, and are right to. The only version that works is turning up beforehand and posting the piece as a contribution to a conversation that was already happening. |
| Your own followers or subscribers | A dependable trickle, and the easiest thing to do | Smaller than the follower count implies — organic reach on a single post is commonly a low single-digit percentage — and it is the same people every time. Posting more often to one list mostly wears the list out. |
| A local newsletter, noticeboard, station or parish magazine | Reaches people entirely outside your network, and small local outlets are chronically short of material | Needs a reason to run it now and a paragraph an editor can use unedited, not a link. Lead times are weeks, so this has to be started before publication rather than after. |
| Search and on-platform discovery | The only route still delivering a year later, and the one that compounds as more pieces accumulate | Slow to begin — months rather than days — and entirely dependent on the text work: an honest title, a real description, a corrected transcript. It rewards patience and punishes nothing else. |
The order below is built around one idea: the effort that produces most of the result happens in the first three days, and almost all of it is directed at individuals. It takes four or five hours in total for a single piece, and it is worth doing in sequence because several of the steps depend on the one before.
Before any link goes anywhere, write a single sentence that describes the piece to somebody with no prior reason to care, containing no adjectives about how moving or important it is. This sentence goes in the message, the post, the newsletter pitch and the description, and writing it once prevents six inconsistent versions. If the sentence will not come, that is usually diagnostic: the piece has not settled on what it is about, and more distribution effort will not fix that.
Contributors should not first encounter the finished piece at the same moment as the public. Giving them a day serves three purposes: it catches factual errors and misheard names while correction is still cheap, it is the decent way to treat somebody who trusted you with their account, and it makes them your most motivated distributors on the day. Include a link that starts where they appear, and a sentence they can pass on without having to write one.
A real title that says what it is rather than being clever, a description written for a reader, a corrected transcript on the page, a chosen still rather than whatever frame the platform grabbed. This is dull and it is the only part of the work that keeps returning results after the fortnight ends. Doing it after the traffic arrives means the arriving traffic sees the worse version.
Ten people, chosen because you can articulate why this piece is relevant to each of them, each message written separately and honestly short. A realistic expectation is that most will read it, several will reply, and two or three will pass it to somebody you could not have reached. A group message to the same ten converts to roughly nothing, because a message addressed to everybody is addressed to nobody.
One or two places where people have already gathered around this subject, posted as a participant rather than as an announcement, with the context in the post itself so it stands up even for people who never click. Then stay for forty-eight hours and answer what comes back. Replies are what keep a thread visible, and an unanswered thread from somebody who posted a link and left is the exact pattern those communities have learned to ignore.
Most people who eventually engage with a piece did not see it the first time, and availability rather than interest is the usual reason. The second outing should not be the same link with "in case you missed it" attached — pull out a ninety-second clip, quote one thing a contributor said, or raise the question the piece ended on. It is a new entry point to the same work, which is legitimate, where the identical post repeated is merely noise.
Ten minutes at the end of the fortnight: where the traffic actually came from, which messages got replies, which organisation passed it on, what produced nothing despite taking an hour. After three pieces this becomes a distribution map specific to your subject and your area, which is worth considerably more than any general advice, this page included.
Analytics at small scale are mostly noise, and treating them as signal is a reliable way to make bad decisions confidently. With forty views, the gap between twenty and thirty per cent completion is four people, any one of whom might have been interrupted by a phone call. Percentages need a base big enough to mean something, and below roughly a hundred sessions almost nothing is stable enough to act on.
That does not make the data useless, it makes some of it far more useful than the rest. Four things are worth reading even when the totals are tiny, because they answer questions that change what you do next rather than merely describing what happened.
The retention curve is the single most informative thing available to a small producer, and it is readable at any sample size because the shape repeats. A steep fall in the first thirty seconds means the opening is not doing its job — usually too much throat-clearing before anything happens. A steady decline throughout is normal and mostly reflects length. A sharp drop at a specific point is a gift: go and look at what is there, because something is losing people at an identifiable moment and you can hear what it is.
Referrer data answers the only strategic question a first-time publisher has, which is where to spend the next fortnight. It is common to discover that the exhausting social posting produced eleven visits while one message to a local organiser produced sixty, and that pattern tends to hold across pieces. This is the number to record after every release, because three data points turn it from an anecdote into your own working knowledge of where your audience actually is.
Returning visitors are the first evidence that something is accumulating rather than merely happening. The absolute figure will be small and genuinely encouraging anyway: a dozen people who came for a second piece are a nascent audience in a way that a one-off spike of two thousand from a single popular post is not, because the spike leaves nothing behind. If the returning figure never moves across several releases, the issue is usually cadence or the absence of any reason to expect a next one.
For community work the thing that matters is frequently invisible in analytics: somebody wrote, somebody volunteered, a contributor told you their family finally understood what they do, an organisation asked to use the piece in a training session. Keep a note of these alongside the numbers. They are the outcomes most pieces are actually made for, they cannot be counted automatically, and a run that produced three of them and modest view figures is not the failure the dashboard implies.
It cannot make people want something they do not want. If viewers leave at forty seconds, more of them arriving simply means more of them leaving at forty seconds, and the fix is in the edit rather than in the posting. This is worth testing honestly before concluding that the problem is reach, because distribution work is legible and satisfying in a way that recutting an opening is not, and it will happily absorb every hour you have.
A related group of techniques does not work either, and some of them carry real risk. Purchased views and followers inflate a figure without producing a single person who will watch anything; engagement pods and follow-for-follow arrangements manufacture signals that platforms are demonstrably good at recognising; mass unsolicited messages, repeated identical link posts and link-dropping in unrelated comment threads are what platform rules and search guidelines describe as spam, and they are treated accordingly. The plain objection is not only that this can get an account restricted or a site demoted. It is that the audience these produce is not an audience — it does not watch, does not return, and does not tell anybody. Every hour spent on them is an hour not spent on the ten messages that would have worked.
Nor can promotion substitute for cadence. A single piece, however good, is a hard thing to build anything on, because a visitor who liked it has nowhere to go and no reason to expect more. The second and third pieces do more for the first than any amount of additional posting about the first, partly because they give returning visitors something to return to, and partly because a body of related work is what makes search discovery start functioning at all. If a choice has to be made between another fortnight of promotion and starting the next piece, the next piece is usually the better investment.
And it cannot be outsourced here, which should be said plainly given how much of this page describes what a channel would normally do for a maker. Kind Channel is new. Nothing has aired, there is no audience to release into, no mailing list, no promotion service and no front page with traffic to pass on, so a piece proposed here is not a piece that acquires a distribution department. If you publish your work elsewhere, the fortnight described above is yours to do, and it will be yours to do for some time yet.
The reason to do it anyway is that this is the only part of the process that teaches you who your work is for. A maker who has written the ten messages knows, specifically and by name, which people responded and why — and that knowledge shapes the next piece more usefully than any audience research, because it came from actual people reacting to actual work rather than from a guess about who might be out there.
About a fortnight of active effort, then stop and make the next thing. The first three days carry most of the result, the second week is for the second airing and any local outlet with a slower lead time, and after that the returns fall away sharply for anything you do by hand. A topical piece has a shorter window still, because the news it attaches to moves on. What continues working is the passive layer — an honest title, a real description, a corrected transcript — which keeps delivering to people searching for the subject months later without any further effort from you.
It is worth doing once, briefly, and it is not where the audience will come from. Organic reach on a single post is commonly a low single-digit percentage of an already small list, so the direct yield is close to nothing. The indirect value is real but modest: it creates a public, linkable record of the piece that somebody else can forward, and it is where people who hear about you will look to check that you exist. Treat it as ten minutes of housekeeping rather than as the distribution plan, and put the remaining hours into individual messages, which convert at a rate no feed can match.
Yes, with three conditions. Send them the finished piece first and give them a day with it, because seeing yourself in an edit for the first time is disorienting and nobody should experience that simultaneously with the public. Make it effortless — a sentence they can pass on unchanged, an image, and a link that starts where they appear. And accept a refusal immediately and without discussion: agreeing to take part is a separate thing from agreeing to promote, some people have entirely private reasons for not wanting a piece circulating among colleagues, and consent to appear must never be treated as conditional on helping it travel.
Lower than almost anyone expects, and the number is the wrong measure in any case. First pieces from unknown makers commonly reach a few dozen people, occasionally a few hundred if an organisation with a list passes it on, and this is ordinary rather than a sign of failure. Two things say more. First, completion: forty people who watched to the end is a stronger result than four hundred who left in the first minute, and it is the figure that recommendation systems respond to. Second, whether the specific people the piece was made for saw it, which you decide before publishing and can often simply verify by asking.