Today's Flash is out. Inside: £830.69m of "sponsorship" an independent commission traced back to the owner (City is appealing), the price tag on the Fed's quarter point, and a brain that gets brighter in every panel.

One question for the room: how many working days until your Q3 books are locked? Today is day four. "Still open" is a valid answer. It is also a cry for help.

Hi everyone, glad to connect. I’m David, currently serving as CFO for a mid-market manufacturing and distribution company.

My background is mostly focused on traditional corporate finance, specifically capital allocation strategies, supply chain risk mitigation, and maximizing cash flow visibility. I joined the community to step outside my industry silo, see how other sectors are navigating current macro trends, and exchange benchmarks.

Looking forward to diving into the discussions here!

Hi everyone, glad to be here! I’m Elena, currently serving as CFO after spending the bulk of my earlier career as a Corporate Controller. My background is heavily rooted in internal controls, audit readiness, and scaling financial infrastructure. I joined the community to benchmark with fellow finance leaders and talk shop on automation, especially around shortening the month-end close. Looking forward to connecting!

The working file. Direct method, thirteen weeks, one Drivers tab that everything else reads from. Change an assumption in one place and the forecast, the Board Summary and the alerts all move.

What it does that a blank sheet does not: it names the Friday you drop below your buffer, it keeps forecast against actual with a reconciliation control that catches cash movements you cannot explain, and it gives you a Board Summary you can screenshot without tidying it first.

One thing to change before you trust a single number in it. The collection curve ships at 50/30/20 across three weeks. That is a weighted lag of 0.7 weeks, roughly a five day DSO, and it describes a card-on-file business rather than an invoiced one. If you sell on Net-30 or worse and leave that default alone, your early weeks will read far richer than your bank will. Push the weight into the later buckets first. The Assumptions tab flags this as HIGH, and it is the single setting most likely to make the file lie to you.

Second one: the payroll cadence in the example is bi-weekly on odd weeks. Set your real pay dates. Payroll timing is where the near-term surprises live.

Make your own copy

If something in it does not foot, say so here and I will fix the file rather than argue with you.

Trueing up an accrual next month is accounting. Reopening a closed period on day sixteen because the estimate was wrong by six figures is a 5-day close in a costume. I have worn the costume.

The number you publish counts the first. The number you live counts the second. That gap is rarely a discipline problem. It usually comes down to one step held together by a single person remembering something: the intercompany eliminations somebody does by hand, the revenue cutoff that waits on a sales ops export, the bank rec that is fine until a single FX entry is not. You know which one yours is, because it is the one you check twice.

So: what broke this month, and what did you change? The second half is the one people come back for.

Worth saying out loud when you post:

• Your day count, and whether it is the published one or the honest one.

• The step you have automated and would recommend, and the one you automated and regret.

Somebody here has the same broken step. They have either fixed it, or tried and failed in a way that saves you the attempt.

Post yours when you close. Bad months are more useful than good ones.

The working files CFO Bound builds live in this channel.

They are built to be used on a Tuesday, not admired in a screenshot. That means no macros nobody can audit, no tab called DO NOT TOUCH, and every assumption on one sheet where you can see it and argue with it. If a model needs a phone call to explain, it is not finished.

The 13-week cash forecast workbook is the standing example of the standard. Direct method, rolling, with forecast against actual built in, so by week eight you can see the size of your own error. That variance history is the part worth keeping, and the part most 13-week templates skip. Ask in the thread if you do not have the current version.

How to ask for the next one. Say what you are trying to decide, not what artifact you want. "I need a covenant headroom tracker" is a format request. "I have to tell a sponsor how close we are to tripping a leverage covenant, monthly, against a quarterly test date, without rebuilding it by hand every time" is a spec, and it produces a better file.

If you have built something good, post it. A model that survived contact with a real board is worth more here than anything polished for a template gallery.

The questions that get answered in this room have three things in them. The ones that die have none of them.

Give the number. Not "our DSO is creeping up". Try "DSO went from 41 to 58 over two quarters while enterprise went from 30% to 55% of what we invoiced. Mix explains maybe half of that and I cannot yet account for the rest." Nobody can help with an adjective.

Name the assumption you are least sure about. Every position rests on one. If you cannot say it in a sentence, you do not have a view yet, you have an output. Saying it out loud is usually where the answer comes from anyway.

Say what you are actually deciding, and by when. "Do I raise this with the audit committee before the March meeting" is a question. "Thoughts on rev rec?" is a search query.

Anonymize as much as you like, and if you are public or in a process, anonymize more than you like. Round the numbers, move the industry sideways, describe it as a competitor's problem. The shape is what people need, not your cap table.

What you should expect back: a straight answer, and occasionally the answer that you are solving the wrong problem. That second kind is what you are actually paying for.

One rule. If somebody answers you, close the loop. Come back and say what you did and what happened. A room where people report the outcome answers better than one where they do not.

The worst introduction in any finance group is the LinkedIn headline. "Strategic finance leader passionate about driving growth." Nobody knows what to do with that, and nobody replies to it.

Here is the format that actually gets you useful answers. It takes ninety seconds.

1. The chair. Your title, the shape of the company, and the size of the P&L you are responsible for. Not the logo, the shape: SaaS, services, manufacturing, multi-entity, PE-backed, VC-backed, bootstrapped, founder-led.

2. What is on your desk this quarter. One line. A close, a raise, a system migration, a covenant conversation, a first audit, a headcount plan nobody likes.

3. The number you wanted a second opinion on and had nobody to ask. This is the one that matters. Start there.

4. What you can answer for somebody else. Everyone here has scar tissue that is worth real money to another person in this room. Say where yours is. First close under audit, an ERP migration that went badly, carve-out accounting, a 409A that came back high three weeks before a grant.

Optional fifth line: the tool you are quietly furious with. It breaks the ice faster than anything else in this list.

Filled in, it looks like this:

Chair: Controller, $40M ARR B2B SaaS, recapped by a sponsor three months ago.

On my desk: first close under the new owner, next month.

Second opinion wanted: the revenue recognition schedule is three people old and nobody left can explain it. Is the error material enough to restate two prior quarters, or do we fix it forward?

I can answer: anything about surviving a quality of earnings review. I have been through two.

Quietly furious with: the ERP, obviously.

What to skip: the elevator pitch, the company boilerplate, anything you would put in a board bio. You are among people who read financial statements for sport. They will decide whether you are worth listening to from how you argue, not from your title.

One rule that is not optional: nobody pitches and nobody recruits in here. If you sell to finance teams, you are welcome as a member and not as a vendor, and anyone working the room as a lead list does not stay in it.

Every CFO Bound issue lands here automatically, so you can find the one you half-remember without going through your inbox. Posting is off, commenting is not. If an issue is wrong, or right but incomplete, say so underneath it. Corrections from this room get published, with your name on them unless you would rather not.

Every tool in your stack was chosen by somebody who watched a demo and believed it. The review that would have actually helped them gets written eighteen months later, by the person doing the renewal, and it never gets written down anywhere.

This channel is for that review. From people who signed the contract and lived with the thing.

Three rules.

1. No pitching. If you sell it, you do not post about it. You can answer a direct question about your own product, clearly identified as yours, and that is the whole allowance.

2. Disclose the relationship. Reseller, implementation partner, advisor, equity, a discount for a case study, a friend on the exec team. Say it in the post. Nobody minds a relationship, everybody minds finding out about it afterwards.

3. Price is part of the review. "We like it" is not a review. What you pay, what tier, what the renewal did, and what it actually cost to implement including the months of your team's time nobody budgets for.

What is most useful here is not the recommendation. It is the thing you would only say privately: what it does badly, what the sales engineer demoed that does not work that way in production, and whether you would sign it again.

The same rules bind me. When CFO Bound is paid for a placement, that is stated on the placement. And nothing in this channel gets removed for being rude about a tool that pays us. If you have run it and it is bad, say so, and say it here first.

Easiest place to start: the one you would not sign again.