Assembly, reconciliation and argument. In the fourteen days before a women's network impact report reaches the board, almost nobody is writing prose. They are pulling data from four systems that disagree, chasing five chapter leads for numbers only those leads hold, deciding which claims survive a lawyer's read, and cutting a forty-slide deck to the six pages a board will actually look at.
The teams that find this fortnight calm are not faster writers. They kept the record as the year ran, so the two weeks are spent reconciling and arguing rather than reconstructing. The teams that find it brutal are rebuilding twelve months of attendance from calendar invites and memory.
What follows is that fortnight, day by day, as it typically runs for a national women's network with several site chapters and a mid five-figure budget.
Day one: freezing the data window and naming the sources
The first decision is the one that saves the most pain later: pick the window and stop moving it. If your fiscal year closes September 30 and the board meets in late October, freeze at September 30 and say so on the first page. Every subsequent number in the deck comes from inside that boundary.
Then write the source list before pulling anything. For each metric, name the system, the owner and the pull date.
| Metric | Source | Owner |
|---|---|---|
| Distinct participants | ERG platform attendance records | Program manager |
| Eligible population | HRIS headcount at freeze date | HR analytics |
| Budget actuals | General ledger cost center | Finance business partner |
| Survey results | Survey tool export | Program manager |
| Chapter activity | Chapter lead submissions | Each chapter lead |
Two hours on this table prevents the worst failure mode of the fortnight, which is a number in the deck that nobody can trace back to a system on the morning of the meeting.
Keep reading: How do I write an ERG charter that HR and legal will actually approve the first time?
Reconciling event attendance against finance spend
Days two and three. This is the reconciliation that catches errors, because the two systems record the same events for different reasons and rarely match on the first pass.
The recurring discrepancies, in rough order of frequency:
- Catering ordered for a headcount, not attendance. You ordered for 60, 38 came. Finance shows the 60.
- Timing. A speaker invoiced in October for a September session. Inside your event window, outside the finance window.
- Spend outside your cost center. A site facilities team absorbed room and AV costs, so the true event cost is higher than your ledger shows.
- Cross-charged sponsorship. A conference sponsorship split with talent acquisition, where each side quietly assumes the other is reporting it.
Work the arithmetic once and it becomes a strong slide. Take an annual budget of 48,000 dollars with 44,600 dollars spent. Across the year you logged 26 events and 612 distinct participants. That is 72.88 dollars per distinct participant and 1,715 dollars per event. Now separate the one national summit at 19,000 dollars: the remaining 25,600 dollars across 25 events is 1,024 dollars each. The summit costs almost twice a normal event per head, and that is a defensible fact rather than a feeling. It is also the number a board member will ask for.
The reconciliation memo
Write a half-page listing every discrepancy over a threshold, say 500 dollars, and how it was resolved. Send it to your finance partner and ask for a reply confirming the figures. That email is your evidence if a number is questioned in the room.
Chasing chapter leads for the numbers only they hold
Days three to seven, running in the background. Chapter leads are volunteers with day jobs, and the request arrives during their quarter close too.
What works: a single form with four or five fields, a stated deadline three days before you actually need it, and a pre-filled draft showing what you already have from the platform so the ask becomes verification rather than data entry. What fails: an open-ended email asking for a summary of the year.
Assume one chapter will not respond. Decide the rule in advance and apply it evenly: a chapter with no submission is reported using platform data only, with a footnote saying local activity may be understated. Never estimate on a chapter's behalf and never leave it out silently.
Turning survey results into statements you can defend
Day six or seven. The temptation here is to lift the most flattering percentage and put it in 40-point type. Resist it, because that is the number that gets challenged.
Every survey statement in the deck should carry three things: the number, the base, and the response rate. Not "84 percent feel more connected" but "84 percent of 211 respondents agreed, from an invited population of 640, a 33 percent response rate."
Adding the denominators feels like weakening your own case. It does the opposite. A board that sees the response rate stated plainly stops interrogating the sample and starts discussing the finding. A board that has to ask for it discounts the whole page.
Also decide what to do with your open text. Never quote a comment that could identify its author, and never quote only the positive ones. Two representative quotes, one appreciative and one critical, do more for your credibility than ten selected to please.
Keep reading: What is the difference between an ERG, a business resource group and a workplace affinity network?
The executive sponsor read and the edits it always produces
Day eight, and reserve a full 45 minutes on her calendar rather than sending a deck and hoping. She has sat in this board room and you have not.
The edits she produces are remarkably consistent:
- Move the ask forward. If next year's budget request sits on slide 34, it moves to slide two.
- Cut the activity narrative. Board members do not want the event list. They want what changed.
- Connect to a business metric the board already tracks. Retention, internal mobility, time to fill, engagement index. Whatever appears in their existing pack.
- Soften a causal claim. Any sentence implying the network caused a retention improvement will be reworded to an association, and she is right to do it.
- Name the risk. Boards trust a report that says what is not working. She will often ask you to add the weakest chapter or the flat participation trend.
Legal and communications review in the same pass
Days nine and ten, and send it to both simultaneously rather than in sequence. Sequential review costs you four extra days and produces contradictory edits you then have to arbitrate.
Legal is checking for identifiable individuals in small cells, causal language about employment outcomes, any eligibility statement that reads as restricting a benefit by protected characteristic, and any claim that would be hard to substantiate. Communications is checking that internal figures match anything already published externally.
Give both reviewers the source table from day one. A reviewer who can see the provenance asks fewer questions and returns the document faster.
See how ERGCircle handles this for employee resource group programs
Cutting forty slides down to the six the board reads
Days eleven and twelve, and it is the hardest part because everything you cut represents someone's year. The discipline: everything else becomes an appendix, nothing is deleted.
A six-page structure that works:
- The ask and the headline result. Budget request and the one number that justifies it.
- Reach. Distinct participants against eligible population, with the trend line.
- Money. Spend by category, cost per participant, prior year comparison.
- What the members said. Two or three defended survey findings with bases shown.
- What is not working. The dormant chapter, the flat quarter, the honest gap.
- Next year. Three commitments with dates and owners.
Build the appendix properly. Board members who care will read it, and the ones who ask a detail question in the room are impressed when you turn to a numbered page rather than promising to follow up.
After the meeting: the follow up that sets next year's budget
Days thirteen and fourteen, and this is where a good report becomes a funded program. Within 48 hours, send a short note to the sponsor and the meeting secretary containing the answers to every question you could not fully answer in the room, plus the appendix page references for the ones you could.
Write down the questions asked, verbatim. They are next year's report specification, delivered free. If two directors asked about frontline participation, that is the metric to instrument in January rather than reconstruct next October.
Then, in the same week, set the four data checkpoints for the coming year: one per quarter, each a two-hour reconciliation of attendance against ledger. Four short sessions replace most of what makes the final fortnight painful.
Making the fortnight boring
The goal is not a faster sprint. It is a report that is mostly finished before the sprint starts, because attendance was captured at the event, budget was reconciled quarterly, survey responses were attached to the sessions that produced them, and every chapter's activity landed in the same record all year.
ERGCircle exists to hold that record: events with attendance, budget against actuals, survey pulses in context, and an annual impact report that assembles from what already happened. The two weeks then go where they belong, into the argument you want to make and the budget you want to win, rather than into finding out what your own program did.
