Starting a group
Start a SkillSwap Group
The original site had a form here. It let anybody propose a new local group, and it is how the Brighton, Bristol and London groups came into existence. There is no form now, because there is no central organisation left to approve anything — and there does not need to be. Nothing about this format requires permission.
What follows is what the record suggests actually matters, drawn from three groups that were set up the same way and got three different results.
The Four Things That Decide It
A room you can have again
Not the best room — the repeatable one. Every documented group used a single venue for its whole run rather than shopping around.
Two speakers, not one
Pairing halves the ask and covers you if somebody drops out. It is also the format the longest-running branch kept, and the long run converged on after trying alternatives.
A date that repeats
The group that never fixed one held nothing at all, despite listing 15 members.
Somebody whose job it is
Every group in the record had one or two named curators. Shared responsibility for the date appears to mean nobody owns it.
What the Three Groups Did
Brighton ran for a decade and 19 evenings. Bristol ran 3 sessions in one half of one year and stopped. London ran none. All three were set up through the same form, in the same period, on the same idea.
The difference was not enthusiasm, since London had members. It was whether a specific evening existed in public with a specific person responsible for it. That is the whole trick, and it is why this page is four points long rather than forty.
For the mechanics of the evening itself — numbers, kit, ground rules, refreshments — the guides go into detail.
Choosing the Community Before the City
The instinct is to think geographically — a town, a neighbourhood — and geography is the second decision, not the first. What determines whether a group works is whether the people in the room share enough context that a specific talk can be pitched without being explained from first principles.
Every group in this archive was a trade rather than a place: people who worked on the web, who happened also to live near each other. That is why a talk on typographic detail or on scripting the browser could be given to a dozen people and be worth their evening. The same talk to a general audience in the same town would have needed twenty minutes of preamble and would have taught nobody anything.
So pick the circle first and then find the room inside it. A trade, a workplace, a professional interest, a craft. Narrower than feels comfortable, because the width you are tempted to add is exactly the width that makes talks impossible to pitch.
The First Three Evenings
Treat them as one decision rather than three, because that is what the record suggests actually happens. Bristol held three sessions and stopped. The group that never fixed a date held none. Nothing in this archive shows a group failing at event four after a healthy first three — the losses happen earlier or not at all.
So before the first evening, have the second room booked and the second pair of speakers asked. It feels like over-planning and it is the single highest-value thing you can do, because the second evening has to run on a habit that does not exist yet, and the enthusiasm that carried the first one will have gone.
For the first night specifically, consider borrowing the format from the one evening in this archive that broke the pattern: eight speakers, a few minutes each, titles only. A short slot is a far smaller thing to ask of somebody who has never spoken, you need no abstracts, and one evening can surface half a dozen people willing to come back and do a full twenty minutes. It is the cheapest way to find out who your speakers are.
From the fourth evening onward the job changes. It stops being recruitment and becomes maintenance: keeping the date, keeping the room, and asking one person each month before they leave the building.
What You Do Not Need
A constitution, a committee, a bank account, a mailing-list platform, a website, a logo, a code of conduct drafted by a lawyer, or permission from anybody. The platform that once existed on this site offered member lists, topic-suggestion forms and speaker-nomination workflows to any group that registered, and the group that used all of it held no sessions while a group with a page and two names held three.
You will eventually want a way to announce the date somewhere permanent, and that is genuinely useful, because it is what lets somebody find you in month eight. But it is a consequence of running events rather than a prerequisite for them, and building it first is the most common way an evening never happens.
The one thing worth doing early is writing down the three ground rules — nobody pays, nobody is paid, talks are short and about the speaker’s own work — because they save awkward conversations later and they tell a prospective speaker what is actually being asked of them.
If It Stops
It probably will, and the record suggests it stops in a particular way: not with a decision but with a month in which nobody fixed the next date, followed by another one. Two of the three groups in this archive ended like that, and the long Brighton run ended like that too — its final evening looks exactly like all the others, with no closing note.
Two things are worth doing about that in advance. The first is to decide, privately, what your own signal is: a number of attendees, a number of months, a point at which you would rather stop than keep pushing. Having that in mind converts a slow fade into a choice.
The second is to leave the record standing. The single most useful thing any of these groups did for anybody who came later was to publish what happened and leave those pages where they were. Dates, names, subjects. It costs nothing at the time and it is the only reason this site can tell you any of the above.
The First Conversation
Almost everything above depends on one exchange going well: the moment you ask somebody to speak. It is worth thinking about how it is phrased, because the difference between a yes and a polite deflection is largely in the framing.
What does not work is an open invitation. "We are looking for speakers" asks the person to decide that their own work is interesting enough to occupy a room, and almost nobody concludes that about themselves. What works is naming the thing: "the way you sorted out that data problem last month — twenty minutes on that would be genuinely useful." You have made the judgement so they do not have to.
Be concrete about the size of the ask in the same breath. Twenty minutes, informal, nothing has to be finished, there will be another speaker so you are not carrying the evening, and roughly a dozen people who all do this for a living. Every one of those clauses removes a reason to say no, and the last one is the most important: the fear is almost always about being judged, and knowing the room is small and friendly addresses it directly.
Ask in person if you can, and ask after a session while they are still in the building. A general call sent to a list reliably produces silence, and it is the single most common way an organiser concludes there is no appetite when in fact nobody was asked.
Two Things That Are Not Your Problem
Organisers routinely spend effort on two things that the record suggests do not matter much, and it is worth naming them so you can stop.
The first is the venue’s quality. Every group here used whatever room it could book repeatedly — an arts centre, a conference room, a broadcaster’s training space — and nothing in the record suggests the choice affected whether the series survived. What affected it was whether the room could be had again. Time spent finding somewhere impressive is time not spent asking the next speaker.
The second is your own expertise. You do not need to be the most knowledgeable person in the room and it is better if you are not, because your job is to hold the date and ask people, and the moment you become the main source of talks the series has a fixed lifespan. Several of these groups were curated by people who spoke rarely or not at all.
What is your problem is narrow and unglamorous: the date exists, the room is booked, two people have been asked, and somebody puts it in public. That list is short enough to sustain for years, which is precisely why it is the list.
One Last Thing
Publish the date somewhere it will still be readable in five years, and publish who spoke afterwards. It takes minutes, it feels like housekeeping, and it is the only reason any of the detail on this site exists to be read at all.
Everything else here is inference from a record that three volunteers happened to leave behind. Leave one yourself.