Below is Kevin Standlee's article about convention organisation structure in the US, after which I will make some comments about how Intersection differed from that, followed by two more theoretical articles on convention structuring.


By Kevin Standlee

ConFrancisco had the following levels in outline, but see my comments later on:

Board of Directors, SFSFC, Inc. (the corporate parent)
Executive Committee (Chairman, Deputy Chairs, Division Chiefs, and
Executive Staff including Secretary, Treasurer, Hotel Liaison
and Convention Centre Liaison)
Committee (Department Heads and deputies and "senior staff" positions
with significant pre-con involvement)
Volunteers -

An important thing to note was that SFSFC's Board of Directors was not, for the most part, involved in any but the broadest management of ConFrancisco. The Board was the eleven directors of the corporation, San Francisco Science Fiction Conventions, Inc,, and included not only people who were senior convention management like Chairman, but also a couple of department heads and even one person who wasn't on the committee at all.

The Board delegated nearly every aspect of running ConFrancisco to the ConFrancisco Operating Committee, reserving only the right of' selection and removal of the Chairman (a task that alas had to be done twice due to the tragic circumstances), approval of a high-level budget, and review of all major contracts. The Board also generally followed a policy of giving the CF Operating Committee as long a leash as possible, which made some people view it as only a rubber-stamp body. It only met regularly twice a year, with an occasional special meeting. The Board invoked a special meeting AT the convention in order to raise the spending cap on the budget because we were substantially further ahead on money than anticipated, and also in case some corporate level emergency arose. The meeting met, untied the Treasury, then recessed "at the call of the chair" to meet again if necessary (it wasn't).

Anyway, the point here is that the Board level doesn't really count for the management; for that you have to start with the Chair and the Executive Committee. The ExecCom was mostly Division Chief's, but also included Division-level responsibilities like Treasurer and Secretary (I was double-hatted as WSFS Division chief as wel1 as Secretary). Below that were the Department heads and "committee-level staff" numbering. around 150 that made up the Committee. Below that were straight "staff" positions, followed by Volunteers (gophers). Staff normally were recruited pre-con while volunteers were on-the-day, but as Ben points out, field promotions were common.

Deputies are in an odd situation. A deputy division chief was generally "committee" except when acting as that division chief, in which case s/he became "executive committee." OTOH, a deputy department head was "committee," not "staff."

I gave a field promotion in my own division when the site selection department head asked to be relieved of' duty and I promoted her deputy to full department head.

Despite my best compulsive efforts, the structure was not totally internally consistent.

Rather than a hierarchy, I tended to visualise it as a series of' concentric circles, with each larger circle containing all of the smaller ones. That is, the Chair was part of the Executive Committee, which was part of' the Committee, which was part of the Staff, who were all Volunteers.


By Fiona Anderson

In some ways this was like Kevin's write-up, in other ways it was far different. For one thing, I don't think any of us Brits are nearly as formal about structures and who's at what level, due to our constantly swapping roles each con, where one con you'll be in charge of X and the next con X will be in charge of you.

There were 2 Boards of Intersection: one was the Board of Directors of the Company "Worldcon Scotland Ltd", who were directly and personally financially liable if the con went belly-up. Then there was the Board who were in charge of running Intersection as a convention.

However, not being on the Board of the Company wasn't an escape from financial liability - as we kept having the concept of "Shadow Directors", whereby anyone who could be deemed to be acting as a Director of the Company could be held to be a Director, whether or not they had that title. Hence members of the Board of the Convention counted as Shadow Directors. Everyone who was approached to be a potential Board member therefore had this explained to them in detail, so everyone knew the potential risks involved to them personally in the case of financial disaster.

Thus the pre-con organisational structure was:

Executive, certain picked individuals (6 persons)
Board (ie the Division Heads)
Committee (ie the Area Heads)
Staff (anyone working in an Area)
Gophers (anyone working to the direction of the Gopher Hole).

Virtually all decisions in a Division were left up to the discretion of a Division Head, with the Board only taking policy decisions on things that affected everyone, or which cut across several Divisions, or issues raised by Board members for discussion.

The Executive met in between Board Meetings, and had the authority to take decisions on any matter that was too urgent to wait for the next Board Meeting.

The Co-Chairs obviously took any decisions that were too urgent even to wait for an Exec Meeting, though they would usually consult with at least one other person before doing so. The at-con structure differed in that we switched over to using the DCM structure previously discussed in the Ops instructions, losing the Exec and Board, with all problems being referred up to the appropriate DCM as the occasion arose, so that it looked like:

Ops, Program, Finance, Site, Co-Chairs
All other departments

This worked reasonably well on the day, with the various innovations we introduced by having 5 people on shift concurrently for each DCM position, instead of only one at a time which was the Eastercon norm. With the number of decisions that did get referred up, it was essential to have an expert in each field available, as an off-the-cuff solution could have cost the convention thousands if the person wasn't fully aware of what Intersection had committed to - and the scope and complexity of such knowledge can't be imagined from trying to scale-up from the Eastercon model, it just doesn't compare.


By Eamonn Patton

(Fiona: when I first read Eamonn's article, which deals with computing systems, I had this brilliant idea that you could apply his principles to conrunning systems instead - well it sounded plausible to me… however having said this publicly, I of course had someone immediately point out the many and obvious flaws in my theory - so I have included Mike Scott's article directly after this one J And Eamonn shouldn't be held responsible for my brainwaves either!!)

I believe we (ie European fandom) should be trying to establish a common, scaleable, and supportable infrastructure for future conventions.

Taking each of the adjectives in turn, by "common" I mean using the same systems for many different conventions. Some of the advantages of such an approach are: that we build a pool of people who are trained and experienced in the systems and that we produce systems that are able to talk to each other. Eg the Membership system should be capable of feeding the Finance system without any re-keying of data. Common systems also help conventions avoid re-inventing the wheel. It will probably (for the foreseeable future at least) be necessary for such systems to be available on multiple computing platforms. Persuading all of the conrunners to abandon their Tower of Babel is left as an exercise for the alert reader.

But seriously, the only reason people will want to abandon the current situation is because they see clear and immediate advantage to their own convention (rather than conventions as a whole). "Scaleable" is the first part of the puzzle. Other than very small conventions (shall we say<150 attendees) any systems should be capable of adapting to the size of the convention. This means an ability to switch parts of the system on and off as the target size of the convention varies. For example it is likely that only the very largest Eastercons or Worldcons will need the ability to reclaim VAT. So a Finance system should be capable of handling VAT. But a small convention should not be forced by the same system to collect useless (for it) VAT information. Should VAT be addressed on an EU-wide basis (sound of accountants' hair being torn out)? The system should adapt to the needs of the convention. Another example drawn from Finance (the Malcolm Reid influence!) is the need for the systems (and Membership is also affected by this) to be capable of supporting multiple currencies, hard currencies, soft currencies, squidgy currencies, etc. There again multiple currency support should be invisible to a 250 person con in Stockport.

And finally "supportable". Possibly I mean supported. There are examples of well-supported Public Domain software (eg GNU). If we have a supported system, then the system can evolve over time to meet new requirements, adapt to changing hardware and software standards. This probably is the key to the whole business - by making the systems supportable and supported we can concentrate our effort on making them do what we want. We might want some sort of controlling organisation to guide developments (have to be careful though - there is certainly considerable potential for ideological wars). We certainly need to move away from personal systems that are understood and supported by only the originator.

The sum of the above should be that, by the time people have forgotten about the effort involved in a Worldcon and decide that they want to do another, we should have reasonably well-proven and capable systems in place with a body of people capable of using them and supporting them. Anyone wanting to do some aspect of conrunning simply picks up the relevant system and there, along with that system is a manual which explains what is necessary and how you can use that system to do what you want ie experience has been captured and made widely available.


By Mike Scott

I disagree with almost everything said in this article and particularly with Fiona's suggestion that it could be applied to general conrunning systems as well as to computer systems. It is futile in the extreme to attempt to devise a structure that will work for anything from a 250 to a 10000 persons con. There are a number of size points where a con becomes qualitatively different, and entirely new structures are needed. I would place the levels as:

Under 150
One person can comfortably run one of these, and having more people involved is for additional creative input or for the social aspects of conrunning

Can be run by a small committee with no formal structures. Beccons 1 and 3 were in this class, and 5 was just a bit bigger and showed the strain a little. Two levels of heirarchy : committee and gophers

Novacon size, more or less. Starts to need some kind of structure to keep things together, but the sort of thing you can jot on the back of an envelope should be fine. Three levels of heirarchy, as you start to get a few responsible people intermediate between committee and gopher.

Eastercon size. Formal committee and staff structures now become necessary, but the committee can still be involved in whatever fine details take their fancy. However, a good deal of the on-the-day running has to be devolved to the staff. Four levels of heirarchy: committee, staff, shift managers, and gophers.

In the past twenty years, Britain has only ever had one con in this size range, Seacon 84. They tried to run it with an Eastercon structure and it didn't quite hold together. It is interesting to not that the BEC is likely to be this size and we have virtually *no* experience of the kind of structures needed - one might look to the US for examples, such as Orycon or Westercon. By interpolation, one might expect to need a four or five level heirarchy.

Non - NA Worldcon size, with the UK being around the top of the range and continental Europe and Australia being around the bottom. Five level heirarchy: main board, division head, area head, shift manager, and gopher.

North American Worldcon. I've never worked on one other than as a gopher, but I believe that they still manage with a five level heirarchy.

Anyway my point is that it is fatuous to expect to use the same systems for cons at the top and bottom of this list. One needs different systems for each level, both in IT and in managerial terms. There are many many shortcuts and time-saving methods that work fine for a small con but can cause disaster for a large one. To attempt to run a con using a system that allows these shortcuts is to court disaster.


By Fiona Anderson

Actually having re-read both Mike and Eamonn's articles over a year later now, I think they are both right in some ways.

Undoubtedly Mike is correct, from a management point of view - in that if you've only ever run 150 person cons you're going to have to drastically change and invent new systems to be able to even consider approaching running an Eastercon, and reading his article certainly demonstrated to me that I had allowed the beauty of the idea to seduce me from noticing the impracticality of it….foot in mouth time!J

But Eamonn's comments struck home quite deeply, as it was only too obvious that very few of the systems Intersection used were designed either to do the job properly or to interact with each other, and many had to be changed to a greater or lesser extent to make them work.

The case of Finance and Memberships is particularly acute - the amount of time taken on turnaround between a person writing to the Office, their details and money being passed to Memberships, Memberships entering them on the database, their financial details and money being passed to Finance was not small, and yet even then there was no real certainty that the correct details had passed through all the stages unchanged, for whatever reason. Consistent, communicating systems would have made everything very much smoother.

However the real problem here is that we don't collectively have enough regular conrunning experience at the larger sizes yet. The US fans manage to run their 6000+ Worldcons with less hassle than we have with our <5000 ones, precisely because they not only have Worldcons or Nasfics regularly, they also have a number of the 2500 size regionals as well. This means that it has been possible for them to have fans with specialised interests at those higher sized cons, for instance there appear to be what could almost be called "guilds" for Art Show organisers, or Techies or whatever, with the result that you as the concom don't have to worry so much about setting up the individual bits, which can more or less run themselves.

Contrast that to the horrendous effort that goes into devising systems for each of the bits of a Worldcon here - Caroline's Dealer's Room article is a perfect example of how many new systems she had to invent in order to make it happen effectively.

What we really need is to have people interested in looking at each bit individually of a con, and to have those people tell us their conclusions as to scaling up/down, or what other changes are needed for each size of con - which is the point Mike was making.

However, we shouldn't forget that at the same time, systems that aren't integrated can be more hassle than they're worth - Dave Clement's article on the Science Programme's interaction with the rest of Programme Division illustrates that point perfectly, I think - which is one of the points Eamonn was making.

Hopefully, people will take AFN as an attempt to look realistically at what systems worked, or if they didn't why not, and use that knowledge as a springboard for designing their own con systems, macro and micro, adapting them for whatever their own circumstances are….