I've worked with companies at both ends - early-stage burning cash and watching runway shrink, and profitable businesses with money tied up in stock and receivables. Different problems, same tool.
That's why cash flow forecasting matters. But it's not just about knowing when you'll hit zero. Done well, it can be one of the more useful tools in your kit.
It's an alignment tool
The underrated benefit of a cash flow forecast is that it gets all your key decision-makers around the same spreadsheet.
Everyone has assumptions about the future. Revenue growth, hiring plans, project costs, sales pipeline. A forecast makes those assumptions explicit. You write them down, you assign numbers, and suddenly you find out that not everyone is thinking the same thing.
Example: sales forecasts closing a big project next month. The revenue looks great in the model. But when you sit down with the ops team, they flag that they don't have the right skills in-house to deliver it - or they do, but those people are already allocated. The forecast surfaced the gap. Now you can figure out whether to hire, subcontract, or push the timeline before you've signed a contract you can't deliver.
The feedback loop
One of the most powerful parts of forecasting is comparing what you thought would happen to what actually happened.
Nothing is more real than what's gone through the business and hit the bank.
If someone consistently says they'll spend less than they actually do, you now have data. Next time, you know to push them harder on their assumptions. The feedback loop sharpens your forecasting over time and builds accountability across the team.
This only works if you actually do the comparison. Make it the first thing you do when rolling the model forward each week.
Keep it simple
One of the biggest mistakes I see is overcomplicating the model from day one. Detailed breakdowns, multiple scenarios, formulas linking to formulas. It looks impressive but it's fragile and slow to update.
Start simple. The main drivers of cash in and cash out. Revenue. Key expenses. Timing. That's it.
At the start, you won't know exactly how the business will behave. That's fine. Build something simple. Adapt it as you learn more about reality. When you go through a rough patch and add complexity to manage it, strip that complexity back out once you're through.
→ The goal is the minimum model that gives you the information you need. Nothing more.
The big variable: sales timing
Expenses are relatively predictable. You've got baseline costs - rent, salaries, subscriptions - that don't move much month to month. Some costs scale with revenue, but the foundation is stable.
Revenue timing is where things can start getting wonky.
When will the deal close? When will you invoice? When will they actually pay?
Each of those steps can slip. A contract in negotiation might take twice as long as you expect. A client might pay on 30-day terms, or 60, or 90 if they're a large overseas company.
Sales timing almost always pushes out. Plan for it. Run scenarios where it takes 2x or 3x longer than you expect. Make sure you've got enough cash to survive that.
→ This is the scenario that catches most businesses off guard.
Build in checks
A good model evolves with your business - new revenue lines, shifting costs, updated assumptions. But every change is a chance for something to break. You need a way to spot errors before they throw off your numbers.
One of the worst feelings is realising your cash is off because something didn't get picked up properly. You thought you had three months of runway and it's actually two.
Build checks into your model. Totals that should reconcile. Sanity checks on key numbers. And don't accept errors - if a check is flagging, fix it before you move on.
If you can, get someone else to peer review. A second pair of eyes catches things you'll miss.
Getting started
You don't need fancy software. A spreadsheet works fine.
List your cash inflows - when you expect to get paid and how much. List your cash outflows - salaries, rent, subscriptions, one-off costs. Map it out week by week or month by month. Compare to actual regularly.
Start simple. Improve as you go.
If you've got questions, happy to chat.