The good news for people who want to become a CFO or VP of finance: You don’t need an accounting degree anymore. In fact:

There are fewer CFOs today who began their careers at the Big Four accounting firms such as Ernst & Young, KPMG, Deloitte and PricewaterhouseCoopers, for example, while the number of less-tenured or experienced CFOs who hold CPA certifications has shrunk as well.

In 2012, 55% of S&P 2000 CFOs had CPA designations, the study found.

By 2022, that figure was down to 43%.

— CFO Dive

The bad news: You don’t have an accounting degree.

This hit me like a Mack Truck my first days in office. (I say that as if I was elected to a gubernatorial position.)

When people need to get paid on time, and travel needs to get booked, you can no longer be a “big picture guy.”

I had never needed to physically issue a credit card, run payroll, or call a customer who was past due (scary!).

And now I had an entire executive staff asking for their own cards, 100 people to pay on Monday, and our cash collections were more than 2.5x our Days Sales Outstanding goal.

I spoke to three new CFOs/VPs of finance who recently stepped into their seats. As much as I try to bury my imposter syndrome, these convos reminded me of all the burning questions I had on day one.

What’s the difference between a controller and an accounting manager?

Before we get into the differences: Meet your new best friend!

Both have similar skill sets — their core is operational accounting.

Here’s the first line from the actual job rec I wrote:

Reporting to the CFO, the Controller will be responsible for broad ownership of financial consolidation, reporting, systems, and operations, including closing cycle, accounting tech stack management, payroll, tax compliance, internal controls, audits, and reconciliations. You will meaningfully contribute to management reporting, vendor contract management, and treasury.

Let’s get a bit more specific:

• Financial consolidation and reporting: Close the books on a monthly and quarterly basis, and produce financial statements that we can give to our investors.

  • The results you produce will also be used by our FP&A team to run the budget to actuals in the operating model.

  • This will require you to come up with a solid general ledger of accounts to capture how we do business (the expense pick list).

• Financial operations: Control how all money enters and leaves the building.

  • This involves calling up our customers when they don’t pay us, making sure we pay our own vendors on time (but at the last possible moment, lol), and ensuring payroll goes out as planned. Cash is king. Cash is oxygen.

• Financial processes: Helping us streamline how we buy and sell things so they can be recorded in a timely, and accurate fashion.

  • For example, are we collecting money from customers using Stripe or ACH? What’s the trade-off in how fast we get paid vs. the cost?

  • Are we allowing employees to book corporate travel on their own credit cards anywhere on the web vs. through a designated travel portal backed by a corporate credit card?

  • Who should we allow to approve spend at the company? And how much can they approve?

When it comes to the differences in title, controller assumes a “head of accounting” place in the finance hierarchy and implies you’ll most likely build out a team under this person (they’ll be a manager of both people and processes).

Accounting manager is generally considered a more junior title to controller. It leaves more optionality to bring in someone else over this person later on, or to promote them, if they do really well, to either controller or head of accounting.

Note that as you expand internationally it’s common for companies to have a “controller” in each geography (e.g., North American Controller, European Controller, APAC Controller). That person essentially “owns” and “controls” the books and the larger accounting team in that jurisdiction.

If you’re not an accountant, I strongly recommend going heavier here.

Just make sure they are a “fingers on the keyboard” leader, and not an “iPad leader.” You need a player-coach — not someone who is going to expect to get everything forwarded to them for review on their tablet.

I’ve seen this hire go wrong — you sign a super expensive person who managed 12 people at [big brand name tech co]. They’ve long been out of the weeds and will not do well in a scrappy environment that depends on them rolling up their sleeves and keying in rev rec.

I think the best candidate is someone who’s a hungry accounting manager looking to step up into the controller role, coming from one of those brand-name companies. They’ll be eager to prove their expertise and build something on their own. And their core technical skills are freshly tuned. You essentially get the best of both worlds in this sense — someone who’s learned what “good” looks like at a larger, reputable company, but also isn’t expecting to hire a massive team and approve things from afar.

I’m walking into a company with a fractional CFO and outsourced bookkeeping. When do I bring this in-house? What do I do first?

My controversial advice: overpay for the transition period.

It’s like when you take your first kid home from the hospital. You know once you step out of those doors, you’re on your own … there’s no more nurse or doctor to help when the thing starts crying.

Day one without the fractional firm, even if they were doing an exceptionally terrible job, is still terrifying. Especially if they left you with a financial model built of string cheese.

Here’s what I’d say about rolling off a firm:

• Break the transition into two parts

  • Accounting

  • Finance

• Be clear with the firm that you are going to bring the function in-house over a number of months — not immediately.

  • I’d suggest using a quarterly milestone as the target transition date.

  • And then ask to pay them [half] of what you paid them before on a monthly basis to have them on call for questions for three more months.

  • You’ll find they are shockingly cool with this. Like Hinge, they’re by design to be deleted. It means they helped guide the company to a point of success where it garners the attention of full-time workers.

• Think about if there’s anyone at the fractional firm you’d like to hire.

  • It’s common to hire good people, who already know the business, from the fractional firm.

  • Just check the contract first to make sure you won’t get sued for poaching.

  • Since it’s common practice, you might pay a fixed finder's fee. For example, I’ve seen terms where you can hire the FP&A analyst on your account if you pay the firm $10,000.

• For the accounting portion of this transition, don’t do anything until you actually make your controller/accounting manager hire, and they are at their desk.

  • Remember — you don’t know all the right questions to ask

  • You want your accountant to get on the phone and ask the “unknowns.”

• For the finance part of the transition, take the firm’s existing model and come up with as many questions as possible.

  • This is your first real view into the business’ revenue and cost drivers; even if it’s mathematically wrong, you can get a grip on the levers.

  • Get on the phone with the firm and have them explain all the answers.

• Build your own operating model

  • Then take the existing firm through it to see if they agree/disagree.

• Start the transition period as you begin to measure the company’s performance against your own model.

  • Or, if you continue to outsource the controller and accounting functions as you sort out the transition, platforms like Brex offer flexible access roles that make it easy for accountants, advisors, bookkeepers, and their clients to work together seamlessly.

What’s the best practice for translating budget into spend controls (ideally in a seamless way)?

At the highest level, establish a threshold for spend to bubble up to you via automated emails. For the stage my company was at, and since I was still learning the spend, I set up notifications for any spend over $1,000 (so I’d get emailed that it was spent, but I wouldn’t need to green light it) and approvals for anything over $10,000 (so I’d have to hit approve on any outgoing non-payroll ACH transfer over this amount).

About 95% of these come through without any sort of interference or questions on my end. I’m just collecting mental data points. And it’s cool to be able to hit approve within my email and not have to log into a portal.

For the other 5%, I forward to the person and say something along the lines of:

“Yo, this real?”

Getting all non-people spend into one platform for visibility was a huge goal for me. Other than hiring a strong controller, my goal was to de-risk my lack of payments know-how by forcing everything into one pane of glass.

Any expense outside of payroll would have to pass through my portal.

Everyone’s asking me for a credit card on day one. I don’t know who should get one.

Hold your horses. First, figure out who has one already. And determine which banks they are linked to.

When I started I learned that one founder had a Capital One Card, the other founder had a Bank of America card, the marketing guy thinks he had a Capital One Card at one point but uses his personal card and gets reimbursed, and that random person over there (who, by the way, is like an entry-level employee) had a Capital One Card (with the highest limit!).

Once I set up Brex, I printed out the spends on each of the other bank cards and started to migrate over the recurring spends to the finance department’s (non-individual-owned) virtual cards. This way nothing would go unpaid during the transition period.

Then I killed off all the legacy cards (note: this feels really good, like when you unintentionally glide scissors through paper).

Next, I sent physical cards to whomever reported directly to the CEO. This was an easy cut-off that I could substantiate to other people.

I made one or two additions to that rule for the IT person and the marketing events person, who both had high spends.

This was all part of the corporate credit card policy I laid out. (More on that in a future post.)

Read: Here's how to choose the best corporate card for your business

Do you have a different policy around giving people virtual credit cards vs. physical credit cards?

Yes, indeed. Virtual credit cards are way easier to turn on and off, and you don’t have to worry about them getting lost (if I had a dollar for every time an exec left their credit card in the foldy thing the check comes in …).

What I like most about virtual credit cards is I can spin one up for a specific purpose — like Google Ad spend — and give it to the right person in marketing (love me a vendor card). I may not have given this individual (say, a demand gen manager) their own physical card, but since I can configure it to be used with a specific vendor for a specific amount (which I can cap based on the budget), I feel more comfortable eliminating this friction and enabling the team to go faster.

Maybe I’m weird, but I like to create virtual cards for specific types of spend — like rent. We have all our WeWork (RIP) and Regus spends on a specific virtual card so it’s easier to monitor. Probably a me thing, but it’s a cool option to have all of it in one place.

Get a card for every occasion from Brex — T&E, vendor, purchase, stipend, lodged, you name it.

How long should it take to close our books each month (and quarter)?

Ideally five business days. That’s the gold standard.

I polled three of my fellow first-time CFO compatriots on how long it was taking to close the books pre-takeover. Their fractional CFO/bookkeeping firms were averaging 13 business days — or by the 17th or 18th calendar day of the month. Yeesh!

So when you come in, there’s a good chance the bar will be pretty low.

My goal was to creep the close down by three days each quarter until we got to five. And we blew past that timeline.

Something important to call out for first-time finance leaders — you still know a lot about the company's performance on day one — like cash and new bookings. So there’s a lot of signal you get out of the biz before the books are closed. Don’t just sit there and stare at your controller.

How long should it take us to collect cash from when we close a deal?

Each existing customer contract will have net payment terms — how many days they have to issue payment from the invoice’s issue date. So n30 means within 30 days of receiving the invoice. What you’ll probably pick up on is you have a ton of contracts with inconsistent terms. It’s like a tangled web of spaghetti that you’ll have to solve for upon renewal.

I’ve worked at companies where standard terms were n30, but we had a few big dog customers whom we gave n90. It wasn’t ideal, but it was the cost of doing business with them.

CJ’s random fact of the day: Walmart is notorious for paying their suppliers like n365. Imagine that.

My advice is to figure out what the industry standard is, and then build that into your templated deal forms going forward. Tell reps they can’t mess with payment terms unless they come to you.

This is crucial for cash flow forecasting and the assumptions your FP&A person makes in the operating plan. When you get this wrong, you intimately learn the nuances of the cash conversion cycle, and it’s one of those things you can’t unsee ever again.

Wait, I have to negotiate these software deals?

Your head of engineering knows Kotlin. He does not know how to negotiate multi-year contracts based on payment terms. He has never read anything by Herbie Cohen, the world’s greatest negotiator.

It’s shocking how many people work for software companies but don’t understand that the price another company quotes you is not the final price you have to pay.

If you aren’t volleying back and forth with the vendor on price minimally 2x, I guarantee you are leaving money on the table.

Will this take a lot of time? Absolutely. And you have to attend sales pitches for tools that you don’t understand (still not sure what Chili Piper does TBH).

But that’s actually a feature, not a bug. I’ve learned more about the product team by sitting in on a Full Story demo and helping negotiate, than what I would have learned from behind a spreadsheet looking at a line item expense.

Plus, you have the best understanding of the organization’s willingness to pay. Some reasons why you should be handling (most) vendor negotiations:

1. Optimizing payment terms: Did you raise money recently? Will that allow you to pay for contracts annually, upfront, and get 5% more off? On the flip side, is there a month where you know burn isn’t going to look good and you’d like it to appear in the following quarter? You can negotiate in those payment terms.

2. Integration bandwidth: Is the head of HR trying to negotiate a new HRIS at the same time the head of sales is looking at a new CRM? You know there’s no way the org has enough mindshare to implement both at once.

3. Vendor consolidation: Are the head of sales and the head of marketing looking for capabilities that could be consolidated on a single vendor (like HubSpot) for better discounting and to reduce vendor sprawl?

4. Headcount visibility: You know how many heads each department is forecasted to grow to over the next three years (assuming you’ve built a longer-term model). This informs the SaaS license ramp in multi-year contracts, which is an asymmetrical information advantage you have on the tech salesperson you’re negotiating with.

5. Usage-based forecasting: The cost of your developers attempting to forecast AWS spend could be catastrophic. It could sink your gross margin.

6. Get better: And most importantly — by doing more negotiations you become the best negotiator at the company. There’s a snowball effect.

Delegating your vendor negotiations to department leads is pretty much the same thing as delegating your cash management to them. They wouldn’t let you code the product (and they shouldn’t), so don’t let them sign up for a three-year Snowflake deal that they forecasted on a sticky note and signed mid-term.

To succeed as a CFO or VP of finance who doesn’t have an accounting background, you need to build a system you can trust. For me, the foundation was built upon people and payments.

This post originally appeared on mostlymetrics.com.