The Convention Cost Whatever It Cost
A four-day convention is the largest single spend most churches undertake in a year, planned by a committee of volunteers, and reconciled — if at all — six weeks after everyone has gone home.
The convention was a success. Attendance beat last year, the visiting speaker was excellent, and the closing service ran long because nobody wanted to leave. Six weeks later the finance committee is still receiving invoices, three volunteers are owed money they paid from their own pockets, the caterer's figure does not match what the food committee remembers agreeing, and nobody can say whether the event made or lost money — partly because nobody agreed in advance what would count.
Church events fail financially in a very particular way. It is almost never overspending against a decision; it is the absence of a boundary. Costs are incurred by eight people across four committees over three months, some through the church's suppliers and some from personal pockets, and the total only exists as a number after somebody spends a weekend assembling it. By then the decisions the number should have informed — whether to book the larger venue, whether to subsidise attendance — have all been taken.
A four-day convention, 600 attendees
Illustrative, and the cost per attendee is the figure that matters most. It is the only number that makes this year comparable to last year and to the neighbouring church's convention — and it is unobtainable unless the boundary existed while the costs were being incurred. The overrun is also worth noting for what it is not: nobody made a bad decision here. Eight people each spent slightly more than intended on a total nobody was watching.
The boundary is the whole technique
Everything that makes event cost control work follows from creating one container before the first shilling moves, and coding every cost to it regardless of which committee, channel or supplier it came through. In our system that container is a project, and the reason it works for events is the same reason it works for a building fund: it accepts costs of any category, from any channel, over any span of time, and reports them as one total against one budget.
The four channels an event draws on all carry a project. Purchase orders to the caterer and the sound company, standalone expenses for venue hire and reimbursements, materials issued from the church store at unit cost, and logged staff time if you choose to track it. That means the total is genuinely complete rather than the procurement portion of the total, which is the distinction that catches most churches out when they try to use a departmental budget for this.
Where the money actually goes wrong
Four failure modes, in descending order of how much they cost. The first is catering priced per head against an uncontrolled headcount. Six hundred people were expected, seven hundred and forty came, and the difference is a quarter of a million shillings that no committee authorised. The fix is contractual rather than administrative: agree a per-head rate with a banded commitment and a cut-off time for confirming numbers, so the risk of over-catering is shared rather than absorbed entirely by the church.
The second is volunteers spending their own money. It is generous, it is well-intentioned, and it is the single largest source of unrecorded event cost — plus a genuine unfairness, because the volunteer who can afford to wait for reimbursement subsidises the event and the one who cannot simply does not volunteer. The fix is a float per committee with a stated ceiling, drawn in advance, reconciled against receipts. Reimbursement after the fact should be the exception with a documented reason, not the operating model.
The third is the same supplier every year without a comparison. The sound company that has done the convention for six years may well be the right choice, and after six years nobody knows. One RFQ every third year costs a committee an afternoon and either confirms the relationship or reveals that the price drifted. The record of having compared is also what protects the committee member whose cousin owns the company — a real situation in a real church, best handled by a documented comparison and a declared interest rather than by hoping nobody notices.
The fourth is honoraria and travel for visiting speakers, which is small in value and disproportionate in risk because it involves payments to individuals with tax questions attached. Decide the amount before the invitation rather than after the ministry, put it in writing, and pay it through a documented channel with the tax treatment confirmed by somebody qualified. A cash envelope handed over after the final session is the version of this that causes trouble later.
A single-day event under two hundred thousand shillings
Code to the ministry's standing project
Do not create a project for every Sunday special or youth day — you will end up with sixty projects and no comparability. Code the costs to the ministry's standing project and use the note field to identify the event. The total is recoverable and the administrative weight is nil.
An annual convention, crusade or conference
One project per event, per year
Create it as soon as the board approves the event, name it with the year — "Annual Convention 2026" — and give it the approved budget. Every channel codes to it. Next year's project sits beside it, and you have a year-on-year comparison on cost and cost per attendee that nobody had to reconstruct.
A multi-site event with branches contributing and attending
One project, with a transport line per branch
Resist the temptation to give each branch its own project — you will lose the single total, which is the point. One project, with the branch identified in the note on each transport and accommodation cost. The subsidy per branch is then derivable and the event total stays intact.
A building or fundraising event where money comes in as well as out
One project, and accept the income is elsewhere
The project will give you a complete and accurate cost. It will not give you the offering, the ticket income or the pledges, because no giving record exists here at all. Plan the net position as a two-source calculation from the start rather than expecting one report, and decide who owns combining them.
The event finance discipline, in order
- Create the project the week the board approves the event, with the approved budget on it. Before this exists, every cost incurred is a cost you will later have to hunt.
- Appoint one person — not a committee — accountable for the total. Committees plan well and watch numbers badly.
- Get the three largest suppliers under approved purchase orders at least three weeks out, competitively quoted where the value warrants it.
- Set a float per committee with a ceiling, drawn in advance. Make reimbursement the documented exception rather than the default.
- Agree the catering commitment as a band with a confirmation cut-off, and name who is authorised to confirm the final number.
- Decide speaker honoraria and travel before the invitation goes out, in writing, with the tax treatment confirmed by somebody qualified.
- Review the project total at the midpoint of preparation, not at the end. This is the only moment a decision can still change the outcome.
- Close the project within four weeks of the event, chase the outstanding invoices deliberately, and publish the cost per attendee to the board.
- Write down the three things you would do differently, and attach it to next year's project. This is what makes the second convention cheaper than the first.
One point about the midpoint review, since it is the item most often skipped and the one that pays for the rest. At the halfway mark of preparation, the project will show you committed spend — approved orders count as soon as they are approved, not when they are paid — against the approved budget. That is the single moment at which the venue can still be changed, the catering band renegotiated, or the printing order reduced. A review conducted the week before the event is a report; a review at the midpoint is a decision.
What AWRA OpsHub does today
- A project as the container, with one budget amount and an actual cost assembled from four channels — approved purchase orders, standalone expenses, materials issued from your own store at unit cost, and logged staff time — each component shown separately so the total is auditable.
- Consumed percentage, remaining, and an explicit overrun figure, so the midpoint review has a number to work from.
- Governed procurement on the suppliers: requisitions that cannot skip approval, value thresholds, RFQ comparison across caterers or sound companies, and supplier prequalification, all retrievable afterwards.
- Payout runs against the project — draft, approved, posted, landing in the consolidated payments register — for reimbursing volunteers and paying individuals who are neither employees nor registered vendors.
- Tasks with owners, due dates and recurrence inside the project, which is where the supplier deadlines and the annual planning cycle should live.
- Expense approval as an option, after which unapproved and rejected claims are excluded from the event total rather than inflating it.
What it does not do
- No event entity. No event record, no date range, no venue, no attendance figure. Cost per attendee is arithmetic you do outside, because nothing here holds an attendance number.
- No registration, ticketing or attendance capture of any kind, so the denominator of every per-head figure comes from a headcount somebody keeps manually.
- No income side at all. No offering, ticket, pledge or contribution record exists, so the net result of a fundraising event always spans two systems.
- No phased or category-split project budget. The project holds one figure, so budgeting catering separately from venue means separate projects and a manual total.
- No budget profile over time, so there is no burn-rate curve and no early warning — the midpoint review is a number you go and look at, not an alert you receive.
- No hard stop at the budget. The overrun is displayed and never enforced; the next order goes through.
- No withholding-tax handling anywhere. Nothing in the system computes or tracks withholding tax on a payment to an individual or a service provider, so speaker honoraria treatment is entirely outside it.
The division is clean enough to plan around: we will give you a complete, traceable, auditable cost for an event, provided the project exists before the spending starts. We will not tell you how many people came, what came in, or whether the event paid for itself. If the reason you are evaluating software is event management — registration, attendance, ticketing — this is not that, and no configuration makes it that.
Our take
Create the project the week the event is approved and appoint one person accountable for the total — those two steps recover most of the value described here, and both are free. Then fix the two things that actually cost money: a catering commitment banded with a confirmation cut-off, and a float per committee so volunteers stop funding the church from their own pockets. Do the review at the midpoint, when a decision is still possible. Everything else on this page is refinement.
Know what the convention cost before the invoices stop arriving
A project container with four-channel cost, governed procurement on the caterer and sound supplier, payout runs for volunteer reimbursement, and every component traceable. Registration, attendance and the income side are not built.
Explore church finance & operationsFrequently asked questions
Should a convention be expected to break even?
That is a decision to make explicitly rather than discover afterwards. Many churches deliberately subsidise a convention from general funds because its purpose is ministry rather than revenue, and that is a perfectly coherent position — provided the board decided it and the size of the subsidy is known. The failure is not subsidising; it is subsidising by accident and to an unknown extent. Set the intended net position in the board approval, then measure against it.
How do we handle volunteers who insist on paying for things themselves?
Thank them, and then explain why the church would rather they did not. Two arguments usually land: it makes the event's real cost invisible, so next year's budget is built on a false figure; and it quietly excludes volunteers who cannot afford to front money, which makes service a function of means. Offer a float instead. Where somebody genuinely wishes to donate the cost rather than be reimbursed, record it as a donation in kind at its value and treat it as a cost the event incurred and a gift the church received — otherwise the event looks cheaper than it was and the giver goes unacknowledged.
What is a reasonable cost per attendee?
There is no useful external benchmark, because it depends entirely on whether you feed people, for how many days, whether accommodation is included, and what your venue costs. The number is valuable as a comparison against your own previous years and against a plan, not against another church. What it is very good at is exposing scale effects: a convention that costs 3,800 per head at 600 attendees and 2,900 at 900 tells you something real about your fixed costs, and that is a genuinely useful planning insight.
Our biggest problem is invoices arriving weeks late. Can software fix that?
Partly, and the mechanism is the purchase order rather than the invoice. An approved order records the committed amount at the moment of commitment, so the event total is substantially known before any invoice arrives — which is the whole reason the midpoint review works. Late invoices then become a payables problem rather than a reporting problem. What software cannot do is make a supplier invoice promptly; getting the order raised before the service is delivered is the discipline that makes the delay harmless.
How should we pay a visiting speaker from outside Kenya?
Take advice before you commit to anything, because cross-border payments to individuals raise questions about withholding obligations, payment channels and documentation that vary with the arrangement and are not something to improvise. Our system does not compute or track withholding tax at all, so it will not help you get this right or warn you if you have it wrong. Agree the gross and net position in writing before the invitation, so nobody is having a tax conversation on the final evening of a convention.
Can we compare this year's convention against last year's automatically?
You can compare totals and components easily, provided both years were run as projects with a consistent naming convention — the four cost components sit side by side and the difference is readable. What is not automatic is cost per attendee, because no attendance figure exists in the system; keep the headcount yourself, in the same place each year, and the comparison takes minutes. Churches that do this for three years develop a genuinely accurate sense of what an event of a given size costs, which is the most valuable planning asset in this whole article.
Who should be accountable for the total — the finance committee or the event committee?
One named person, appointed by the board, who sits on the event committee and reports to finance. Making the finance committee accountable puts the number with people who are not making the spending decisions; making the event committee collectively accountable means nobody is. The named person's job is not to approve every cost but to watch the total and raise the alarm, which is a role that only works if it belongs to an individual with the standing to be uncomfortable in a committee meeting.
