Set a runway target, then divide your cash by those months to find the net burn you can carry. Compare that target with current burn to size the required saving. The example below tests cost cuts by their effect on revenue and the date each saving actually starts.
Start from the months you need
Start from the runway you need: pick the number of months, then divide your cash by it to get the highest net burn you can run. The example company has €540,000 in the bank and wants 12 months, so its net burn can be at most €45,000, against €60,000 today.
The example is an eleven-person software company with €540,000 in the bank. It spends €98,000 a month and collects €38,000 from customers, so its net burn is €60,000 and its runway is 9 months. A senior salesperson at €9,000 a month has been approved and is due to start next month. With that hire the burn becomes €69,000 and runway drops to 7.8 months.
The founders want 12 months: enough to reach their next milestone and still have time to raise. Twelve months on €540,000 means a net burn of €45,000 at most.
target net burn = cash / months needed = 540,000 / 12 = 45,000 gap = 60,000 - 45,000 = 15,000 a month (24,000 if the hire goes ahead)
Now the question is specific: find €15,000 a month, and do not add €9,000.
Sort the ledger by what stops
Sort your spending by what stops when you cut a line, not by its size. List every line of monthly spending with two extra columns: what it pays for, and how soon a cut would take effect. The size of a line says little about whether to cut it.
| Line | Per month | What it pays for | A cut lands |
|---|---|---|---|
| Team (11 people, with employer costs) | €74,000 | Product, sales, support | After notice periods |
| Paid ads | €7,000 | €4,000 in campaigns that bring customers within a quarter, €3,000 in brand | Immediately |
| Contractors | €6,000 | Two part-time specialists | After 30 days' notice |
| Cloud | €4,200 | Production, plus test environments nobody has used since spring | Immediately |
| Software | €3,800 | Tools, including unused seats and two doing the same job | At each renewal |
| Office | €3,000 | A space sized for 15 people | When the lease ends in four months |
The team is three quarters of the spending. Paul Graham puts it plainly in The Fatal Pinch: "In most startups, expenses = people." That is why the order below matters. Everything before the last step is about avoiding cuts to people you need.
In what order should you cut costs?
Cut spending in this order: commitments that have not started, spending nobody would miss, spending that pays back slowly, and people last. In the example, the four steps take net burn from €69,000 to €43,800 and runway from 7.8 to 12.3 months on paper.
| Step | Monthly saving | Net burn after | Runway |
|---|---|---|---|
| Today, with the approved hire going ahead | €69,000 | 7.8 months | |
| 1. Do not start the new hire | €9,000 | €60,000 | 9.0 months |
| 2. Unused software and idle cloud | €2,200 | €57,800 | 9.3 months |
| 3. Brand ads, one contractor, a smaller office | €7,500 | €50,300 | 10.7 months |
| 4. One role | €6,500 | €43,800 | 12.3 months |
Step 1: stop what has not started
Review hires, office moves and contracts that have not started. Check signed commitments for notice periods or cancellation costs before counting a saving. They also do not show up in today's burn, so founders forget to count them. Start here.
Step 2: spending nobody would miss
Seats for people who have left, two tools doing one job, and test environments running for demos that ended. This company finds €1,200 in software and €1,000 in cloud. It is a small share of the gap, but nobody's work changes, so it goes first.
Step 3: spending that pays back slowly
Keep the ads that bring customers within a quarter and cut the €3,000 brand campaign, which may pay back but not within the runway you have. End one of the two contractor agreements, and move to a smaller office when the lease ends. Together that is €7,500 a month, and runway reaches 10.7 months.
Step 4: people
The last €5,300 of the gap is where the choice gets hard. There are three ways to close it. One role goes, at about €6,500 a month with employer costs, and runway reaches 12.3 months. Or revenue rises by €5,300 a month, which is 14% more from the same customers; raising prices is the quickest way if the product can take it. Or the founders accept 10.7 months and plan the raise earlier.
The Fatal Pinch also covers pay cuts across the team. Graham calls them a weak fix that only works when the problem is small, and when it is not, the people taking the cut can see the problem has only been postponed. If a role has to go, the People page view of cost by department is where to look. Make the decision once and make it big enough. Cutting again three months later means the team goes through it twice.
Cuts land later than you decide them
Cuts land later than you decide them, because the table assumes every saving starts today. In this company the cloud, software and ad cuts take effect in month 1, the contractor after a month's notice, the role after two months' notice with one month of severance paid, and the office after four months.
Month by month, that plan runs out of cash during month 12, at about 11.7 months instead of the 12.3 on paper. The difference is small here and still enough to miss the target. The fix is to start the slow cuts first and give notice this week, or cut a little deeper to leave room. Put the dates in your run-out calendar and check the result against the target.
What should you not cut?
Leave off the list anything that brings in revenue within the runway you have, anything customers touch every day (support response times, uptime and the people who keep the largest accounts happy), and your own pay, if cutting it would hide the real cost of the company.
Revenue comes first. The €4,000 of ads that produce customers this quarter stays, because cutting it lowers burn and revenue together.
Anything customers touch every day. Support response times, uptime and the people who keep the largest accounts happy protect the €38,000 that comes in each month. Losing a customer to save a small amount of spending makes the burn worse.
Your own pay, if cutting it would hide the real cost of the company. If founders drop their pay to zero to make the numbers work, the burn looks better than it is, and it jumps back when they start paying themselves again.
How do you check that a cut happened?
Check that each cut happened by comparing next month's bank statement line by line against the plan. A cancelled tool that keeps billing and a contractor who is still invoicing both leave the burn unchanged, and if a cut is not visible there, it has not happened yet.
In Plainhub, each cut can go in as a plan with a minus sign: a cancelled tool is a plan of -€1,200 a month. Give it the date the saving actually starts, and once committed it lowers your burn from that date on. To test a spending level before you commit to it, the burn rate calculator shows net burn and runway for any total, and default alive or default dead tells you whether the smaller company reaches profitability on the cash it has.