Key Takeaway: Before you sell shares, stack every cheaper source first: cash already in the business, customers, a grant, a lease, your invoices, future sales. Sell equity only for what's left. In our example that turns 20% of the company into 8%.
By Samer Azar
Dear reader,
Atlassian started on a $10,000 credit card. By 2015, the Australian company behind the planning software Jira had 79 of the Fortune 100 as customers.
For its first 8 years it grew on what its customers paid, and it was profitable before a venture firm came in.
That was a choice, and today it's worth a number. Sell $5 million of shares at a $20 million valuation and you give away 20% of your company. Use the cheaper sources first and you give away 8%.
Atlassian's own filing, 2015.
The cheaper sources are a ladder with 7 steps, and you walk down it before anyone sees a term sheet.
7 steps, cheapest one first
The ladder is every kind of cash that costs you less than your shares. Cheapest first:
Cash already in your business.
Customers' money, paid before you deliver.
Grants.
Asset-backed finance, like a lease on equipment.
Working-capital finance, like an advance on your unpaid invoices.
Revenue-based financing, once your marketing pays for itself.
Equity, at the very top, for whatever is left.
Every dollar you find on steps 1 to 6 is a dollar of shares you don't sell on step 7.
Which step would you reach for first? Reply with the number, 1 to 7.
Your customers can fund step 2
At the end of 2012, Tesla was holding $138.8 million of its customers' money for cars it had not delivered yet. Each reservation started at $5,000 or more, fully refundable.
A bank would have charged interest on that cash. An investor would have wanted shares. Customers wanted their place in the line.
Tesla's annual report says it used those funds "for working capital and other general corporate purposes."
Your version is smaller and works the same way. If your customers trust you enough to wait, ask for a deposit on every order, or a year paid up front. It's the idea from our September issue: get paid before you pay your suppliers.
You still owe them the product, so take deposits only on orders you can deliver.
Same $5M: 20% becomes 8%
Here's the sheet from the video. You need $5 million, and the company is worth $20 million before the round.
The sheet from the video. Illustrative figures.
The share you sell is the equity you sell, divided by the valuation before the round plus that equity.
All equity: $5M ÷ ($20M + $5M) = 20%.
After the ladder: $1.75M ÷ ($20M + $1.75M) = 8%.
Put two prices on every row: what it costs in dollars, and what it costs in percent of the company. Steps 1 to 6 cost you dollars. Step 7 costs you a share you never get back.
Equity gets 1 job only
GIVE EVERY DOLLAR A JOB.
Inventory and unpaid invoices get working-capital finance. A machine or a van gets asset finance. Equity pays for growth that isn't proven yet, so size it on that and nothing else.
In 2012, GitHub took its first outside investment when, in its own words, it "doesn't need money." That's equity at its cheapest: taken last, by a company that could say no.
Three moves for Monday:
Write your ladder, one row per step.
Put a price on each step, in dollars and in percent of the company.
Keep equity for growth that isn't proven yet.
Move every dollar you can down the ladder before you sell a piece of your company.
The rule of this issue
If you'd rather have someone watch your cash for you, apply for a strategy session and we'll walk your ladder with you.
Before you start, decide how much of your own cash goes on the ladder, and arrange your credit line in a good month. Last issue showed why.
Variance comments in 40 seconds
Files from building the AI CFO. Ask why a cost line jumped, and get the reason plus whether it will come back.
You know the month-end question: "why is hosting up 186%?" Someone digs through invoices for an hour to write one line.
My team and I ask the AI CFO instead. On a demo company it answered in about 40 seconds for two cost lines that jumped from February to March.
It reads the entries behind each line and sorts every jump: one-off (an old server cluster left running for about three weeks alongside the new one) or likely to stay (a card-fee reset).
A person reads it before it goes in the pack. The call on one-off or ongoing stays with you.
On a demo company: two lines that jumped, and why.
Steal it:
Inputs: list every cost line that moved more than 20% month on month.
The test: each one gets one sentence of why, and a label: one-off or ongoing.
The human: whoever signs the pack checks every label.
If you run a finance team, forward this to whoever writes the month-end commentary.
Not subscribed yet? Join the CFO Lab
What should we break down next? One tap. It picks what we break down in a coming issue.
What should we break down next?
Your spreadsheet is now Claude Code
Do not get intimidated. Do not put this off for later.
Know who the knocker-upper was? Before alarm clocks existed, there was a person whose job was to walk through town with a long stick, tapping on windows to wake people up for work. That job disappeared overnight when alarm clocks became cheap.
Spreadsheets are the long stick. Claude Code is the alarm clock.
Don't be a knocker-upper.
Reply with your step number and we'll write back with what we'd check before you use it.
Reply and let me know what's on your mind finance / AI / growth-wise. I read every one.
Take care of your cash, and it takes care of almost everything else.
Samer
Next Saturday: another company's own filing, read for the one number you can use on Monday.
If you work in finance, you know a founder about to sell a slice of the company for cash a cheaper step could cover. Forward them this issue.
Not subscribed yet? Join the CFO Lab
Frequently asked
How do you raise money for a startup without selling too much of it?
Walk the round down a ladder of cheaper sources first: cash already in the business, customers' deposits, grants, a lease on equipment, an advance on unpaid invoices, and revenue-based financing. Sell equity only for what's left. In our worked example, the same $5 million costs 8% of the company instead of 20%.
How do you work out the share of the company you sell?
Divide the equity you sell by the valuation before the round plus that equity. Selling $5 million at a $20 million valuation gives away 20%. Find $3.25 million on cheaper steps and sell $1.75 million instead, and you give away 8%. Put that percentage next to the dollar cost of every other step.
What counts as customers' money?
Cash customers pay you before you deliver: a deposit on every order, or a year paid up front. At the end of 2012, Tesla held $138.8 million of reservation payments for cars it had not delivered yet. You still owe customers the product, so take deposits only on orders you can deliver.
What should equity pay for?
Growth that isn't proven yet, and nothing else. Inventory and unpaid invoices can be financed against themselves, and a machine or a van can be leased. Equity is the most expensive step because the share you sell is gone for good, so size the round on the unproven growth and fill the rest from cheaper steps.
Did Atlassian ever take outside investment?
Yes, but late. For its first 8 years it grew on what customers paid, and it was profitable before a venture firm came in. Its 2015 filing to go public says it financed its operations mainly from the cash its operations generated. Both founders were still running the company when it listed.