🕯️ Keeping the lights on (while everything else burns)

Eight weeks into a new role, and the AR lead asked for two more credit collectors. It’s a ‘no’.

Good ask, great person, drowning team. But I’ve seen this movie before: two more heads buys a pricier version of the same broken process, and 18 months later you’re stuck having the same conversation with a bigger salary line.

HiBid’s CFO, Chris Stiegal, had a sharper version of the same issue: 30 to 40 people across sales and support, each making five to 10 collection calls a week (none of it their actual job, no record of who’d been chased). He spent over a year evaluating AI calling tools before picking one.

Stuut went live 21 days after first data, covered almost 7,000 accounts, and collected $4.9m on the invoices it managed. Write-offs to an external agency halved in a quarter. Sales and support stopped making calls, and no extra collectors needed.

A good Board shouldn’t be too patient.

I sat in a Board meeting while one independent director, a decorated former CFO, lectured me on the importance of urgency.

It was annoying, and patronizing.

We’d performed miracles cleaning up the finance function. There was still a mountain of work to do, and a whole team to rebuild across many operating locations, but we’d reached the point where the “given what you inherited” caveat was starting to expire.

The function wasn’t remotely ready to run a proper budget process. We simply didn’t have enough clean periods of actuals behind us to provide a meaningful base.

But… he was also right. It’s not like we could put year-end back by three months (although I briefly considered it).

We just had to figure it out.

If your ship develops a serious fault, the logical thing is to take it back to port, put it in dry dock, and repair it before heading back out to sea. But that analogy doesn’t hold running a finance function. The financial calendar just keeps marching on regardless. We had to keep sailing toward the destination while we were still bailing water out of the hull.

For a while, the Board would forgive weak forecast accuracy or poor insight. They understood those things needed solid foundations underneath them. But that patience lasted only until the business suddenly needed an answer.

A big commercial problem would emerge, or a key decision would be needed, or the auditors would rock up for year end.

And nobody cared that the foundations weren’t ready… and nor should they.

We just had to make finance work anyway.

In week one, we started the series with how to spot a broken finance function before you take the CFO job… so you don’t end up eating an unpleasant sandwich you didn’t ask for.

Last week, we focused on how to uncover the whole truth in a newly inherited finance function.

And this week, now that you’re standing in the engine room, head in hands, staring at the broken machinery scattered across the floor, we’re focusing on how you put it back together.

But before we do that, we have some unfinished business from last week…

Last week, we finished our balance sheet colonoscopy with two key registers:

That first list is now a clogged toilet with your name on it.

The technically pure answer is simple: once you know an accounting error exists and needs correcting, correct it as quickly as possible.

But in a real shitshow, discovery doesn’t happen neatly. It can take months. If every issue gets pushed through the accounts as soon as it appears, you end up with little firecrackers of unknown sizes, timing and location going off everywhere…

Pretty quickly, nobody knows where the underlying business ends and the legacy cleanup begins.

The better approach is to collect all those little firecrackers and turn them into one bigger, but controlled, explosion.

As a man of a certain age, I get far more enjoyment than I should from watching a stick of dynamite placed under a cooking pot.

In practice, that means ring-fencing the legacy issues and putting them under your direct control as CFO.

That does a few useful things:

In practice, and in especially difficult situations, this might need two or three windows of clean up, but the same ideas apply.

But meanwhile, you have an even bigger problem…

So far, I’ve presented this as a nice sequential process. Pick the right role. Find the bottom. Clean up the mess. Fix the root causes.

Anyone who has actually fixed a shitshow knows that isn’t how it works.

You can’t press pause on time while you focus on the basics.

Month-end keeps coming around, bringing another Board meeting with it. Year-end might land halfway through your cleanup, with the auditors arriving at a spectacularly inconvenient moment. There’s a budget to set, etc…

And that’s before you even think about the business.

Maybe a major customer is at risk of churning and contract profitability suddenly becomes urgent. Meanwhile, you know the analysis is at least partly bullshit. If the P&L wasn’t right in aggregate, it’s not getting more reliable the finer you slice.

So perhaps the hardest question in this whole endeavor is:

How do you deliver a minimum viable finance function while you’re still fixing it… all at the same time?

The capacity problem makes this even harder.

Chances are, you inherited a slow close too. And the cleanup only makes it slower.
There are roughly 22 working days in a financial period. Maybe ten of yours are already disappearing into close. During the worst of the cleanup, perhaps that stretches to fifteen.

That leaves you seven days to do everything that isn’t closing the books.

Meanwhile, your mate down the road with her supersonic transformed finance function closes in three days and has nineteen of the 22 to forecast, analyze, support the business, and contemplate her next AI agent.

You have seven. And somehow, in those seven days, while fixing the finance function is consuming all your time, you still need to support the business.

So, how do you keep the ship moving when half the crew is still below deck fixing the engine?

Well… you cannot polish a turd, BUT … you can roll it in glitter.

So let’s talk about how…

We’ll use my old favorite model to do that, breaking the finance function into 5 parts: inwards, backwards, forwards, upwards, outwards.

Naturally, the goal is to build a finance function where the things on the left work largely on automatic, freeing the team to spend more time adding value on the right.

But I’m afraid you have to earn that.

For now, the broader finance team should be heavily focused inward and backward: getting the basics right, rebuilding accuracy, and creating a foundation you can trust.

Only once that foundation is stable should you start pushing more forecasting, decision support, and stakeholder management back out into the wider team.

That creates an obvious problem in the meantime.

The business still needs a forecast. It still needs help making decisions. And the Board, auditors, and other stakeholders still need managing.

That gives us a simple Repair Mode operating model:

For a period, the further down you move this table, the tighter the circle of people you trust to do the work. (Next week we will focus more on who those people are, how you find them, and how this evolves as you move out of repair mode).

Let’s get into each in more detail:

Here we are talking about the core financial cycles. How you bill and collect cash. How you spend money. How you move inventory.

It is reasonable to focus a large part of the team ‘inward’ first. The transaction volumes are massive, and the processes clearly aren’t working properly. If they were, we probably wouldn’t be having this conversation.

Your balance sheet colonoscopy should already have told you which processes are producing the most … ahem … mess.

Now you need to go upstream and fix the source.

Beware diagnosing a software problem before you understand the process problem. There may well be a tooling issue buried in there, but the immediate fixes often come back to boring basics: ownership, process discipline, controls, handoffs, and people doing what they are supposed to do.

You might also be tempted to centralize everything. That might make sense from a reporting-line perspective, but be careful. Finance processes rarely break entirely inside finance.

A broken billing process might start with poor contract setup in Sales. Inventory accounting might be downstream from weak warehouse discipline. AP problems may begin with terrible purchasing behavior in the business.

If you centralize execution too aggressively, you risk destroying the connective tissue with the operating teams whose behavior actually needs to change.

But you MUST centralize governance.

That means clear local ownership of each process, with somebody centrally accountable for challenging it. And measuring the exception points relentlessly: billing lag and accuracy, credit terms, overdue debt, parked invoices, disputes, inventory adjustments and write-offs.

In repair mode, I care more about where the exceptions are hiding than whether the average looks healthy. Then somebody needs to jump on those exceptions like their life depends on it.

In my case, that would usually be the fix-it Controller we talked about last week, with a weekly cadence at minimum, and much faster where the risk demands it.

Because there is a process culture to change here, and that does not happen through one clever workshop, an aha moment, or a couple of people changes.

It happens through miserable relentless measurement, governance, escalation, and corrective action until the process produces predictable exceptions that are spotted, owned, and resolved without senior intervention.

Until you have Inward working properly, broken processes will keep feeding garbage into your historical reporting. Previously, that mess was making its way into the reported accounts. Hence the balance sheet colonoscopy.

That has to stop.

Even if billing is still producing discrepancies, or rogue spend means commitments are not being captured cleanly at source, you need to produce reliable monthly accounts.

First in aggregate. Then progressively at location, cost-center, and account level. Eventually, the underlying transactions should support the whole thing.

That may require manual work during close. Accruals. Reclasses. Corrections. Whatever is necessary to compensate for the processes you haven’t fixed yet.

The temptation will be to tolerate those inaccuracies during the year, and clean them up quarterly or at year-end.

Resist it.

The pain of producing accurate accounts every month is exactly what forces you upstream to fix the processes creating the errors. The danger is letting that manual cleanup become permanent.

Ultimately, you need self-cleaning finance teams: identifying and correcting their own errors before they travel further downstream and become somebody else’s problem. But that is not a switch you flick.

The value of your reporting rises quickly too as the fidelity improves at location, cost-center, and account level. You can start producing genuinely useful analysis on margin by product, customer and channel, labor productivity, purchasing efficiency, overhead intensity, and working-capital performance.

The KPIs become reliable. You can slice them without discovering that the answer changes depending on where you look.

Yes, it is still backward-looking. But once you can reliably explain what happened, figuring out how to move that into a forward-looking function gets easier.

A shitshow finance function will usually be doing one of two things on forecasting:

The second creates a particularly nasty cycle. Nobody trusts the numbers, so they ask for more detail, more analysis, and more forecasts. That keeps everyone busier, reduces focus on the fundamentals, and usually makes everything a lot worse.

You have to break that cycle.

But you also can’t run a business entirely through the rearview mirror: CFO 101.

The answer, as much as possible, is to move forecasting up and in: simpler, more top-down, and controlled by a small number of senior people with full context.

That might sound counterintuitive. But until you trust the granular actuals, what value is there in producing an incredibly granular budget against them?

False precision.

As a rough rule, I think the fidelity of your forecast should lag the fidelity of your actuals by about one financial year.

You need enough clean history to form a credible base, and enough control around the underlying processes to understand how the drivers translate into the financials at thinner slices of a growing number of dimensions.

So go back to the economic engine of the business. Identify the ten or twelve variables that really drive revenue, margin, operating costs, working capital, and cash. Understand those deeply, and build a simple budget and forecasting cadence that is robust at that level.

Top-down forecasting in this environment is faster, more controllable, and enough to help spot and steer shifting priorities.

That will mean resetting expectations with the Board. The forecast may temporarily have less depth underneath it. You might not be able to drill into every variance or produce pages of detailed assumptions.

But the assumptions you do make, you should understand intimately, and be able to bridge them well.

Then, as you move from lo-fi to hi-fi finance, you build the detail back in. Deeper P&L breakdowns, operating locations, channel margins, etc.

What matters initially is having a credible view of where the business is heading, the major sensitivities, and where the biggest risks sit.

As confidence in the actuals improves across functions and locations, you can progressively add detail and eventually reintroduce bottom-up forecasting.

Complexity gets earned back.

This is not the time to be rolling out a business partnering strategy across every touchpoint.

Even good finance people add limited value when they are working with weak information, and some of the judgment around you may be flawed too. You are still learning who in the team you can trust. (We’ll come back to that next week).

So focus scarce finance capacity on the decisions that matter most. The ones that can materially move the P&L, cashflow or balance sheet, or create consequences that are expensive and difficult to reverse.

Pricing decisions. Material inflation. Major margin-conversion problems. Cost initiatives. Large investments. Significant customer or product decisions.

In practice, that probably means focusing support around the C-suite, and perhaps one level below it to begin with. And again, keep that support centralized and close to you.

That reduces the risk of decisions made from poor context, but it also gives you the chance to set the standard for what good business partnering actually looks like from the Boardroom down. Rigorous analysis. Commercial context. Clear choices. A point of view.

Then, as the information improves and confidence in the team builds, you can steadily extend that support deeper into the business.

I’ll keep this one brief.

Whether it’s the Board, investors, banks, auditors or other stakeholders, during a shitshow repair job you are carrying a huge surface area of uncertainty while trying to rebuild confidence in the numbers.

This is basically an un-delegatable job.

There will be so many moving parts that, at times, you may be the only person with the full context. You are trying to rebuild trust through all of that uncertainty.

The last thing you need is the lens or voice becoming another moving part.

So keep the message consistent. Be transparent, but always with context: what you know, what you don’t, what has changed, and what you are doing about it.

You might feel like you are being a control freak… who cares… some things are too impotant.

It is also how you keep buying the air cover your team needs to do the repair work.

It’s important to take this post in context.

This is not how you want to run a finance function in normal circumstances. And it is definitely not best-in-class.

But it is how you put finance onto a war footing and keep the ship moving while you clean up the mess underneath it.

Of course, you can’t stay in that state forever.

At some point, you have to figure out how to start lifting out of repair mode and build a finance function you can actually be proud of.

Knowing when to make that transition, where to start giving ownership back, and how to avoid losing control as you do it is the tricky bit.

That’s where we’ll go next week.

:::::::::::::::::::::::::::::::::::::::::::::::::::
:: Thank you to our sponsor ::
:: Stuut ::
:::::::::::::::::::::::::::::::::::::::::::::::::::

If you enjoyed today’s content, don’t forget to subscribe.

Disclaimer: I am not your accountant, tax advisor, lawyer, CFO, director, or friend. Well, maybe I’m your friend, but I am not any of those other things. Everything I publish represents my opinions only, not advice. Running the finances for a company is serious business, and you should take the proper advice you need.