Home/Blog/Runway & Burn/How to Build a Runway Model You Will Kee…
Runway & Burn

How to Build a Runway Model You Will Keep Updating

How to build a runway model: four inputs, five or six categories, connected accounts, lumps on the calendar, a fixed monthly refresh, and then stop building.

Cover of the FIKR Space guide: How to Build a Runway Model You Will Keep Updating

To build a runway model, keep to four inputs, split spending into five or six categories you can act on, connect your accounts so the model updates itself, put lumpy payments in the month they land, and refresh it on a fixed day each month. Then stop building and start using it.

There is no sophisticated version of this that a founder needs. Complexity in a runway model is almost always a substitute for maintaining a simple one. An elaborate model that is three months stale is strictly worse than a crude one that is current, because a stale model still displays a number, and a number looks trustworthy whether or not it is.

How to build a runway model: the four inputs

The whole structure fits on one screen. Opening cash. Money in, by month. Money out, by month. And the balance that results, projected forward until it crosses zero.

That is all. Everything else in this post is about making those four inputs honest, because the arithmetic was never the hard part. If you can say where each input comes from and how it will be refreshed, you have a model. If you cannot, adding more tabs will not help.

InputWhat it isWhat keeps it honest
Opening cashThe balance in the accounts todayNothing else is added to it
Money in, by monthWhat you expect to receive, month by monthSplit into contracted, probable and hoped
Money out, by monthWhat you expect to pay, month by monthCategorised, with lumps on the calendar
Projected balanceThe result carried forward to zeroRecalculated on a fixed day

Step one: categorise the outflow so it can be steered

A single expenses figure gives you the clock but nothing to act on. When you need to buy back time, you want to see at once where the money is and which levers are real.

So divide the outflow into a handful of groups you can make decisions about. The playbook suggests people, infrastructure and hosting, software and tools, marketing and acquisition, professional services, and a bucket for genuinely one-off items. Five or six is the right number. Twenty is bookkeeping, and you will stop maintaining it.

The aim is not accounting accuracy. It is that, on the day you need to lengthen your runway, the model already shows you which category to open first. The order in which to pull those levers is covered in how to extend runway.

Free guide, 31 pages: The Runway Playbook. Send me the guide.

Step two: connect the accounts instead of retyping them

The most common reason a model goes stale is that updating it is a manual chore somebody has to remember. Anything that pulls transactions in automatically removes the main reason the number drifts out of date. It also removes a whole class of transcription errors that quietly corrupt a hand-kept sheet.

If you take only one operational step from the whole playbook, make it this one. The model that updates itself is the model you will still trust in six months. A beautiful structure that depends on someone's memory will be abandoned the first busy week.

Step three: put the lumpy items on the calendar

Averages smooth things that reality does not. Annual and quarterly payments, such as insurance, audit fees, software renewals and tax, do not spread themselves evenly across the year, even though an average treats them as if they do. The same is true of payroll timing, sales tax or VAT cycles and any large one-off purchase.

Put each known lump in the month it will actually land. A company whose average monthly burn looks comfortable can still fail a particular month, and the model should show you that month well before it arrives.

Lumpy itemWhy an average hides it
Annual insuranceOne heavy month, then none for the rest of the year
Quarterly tax billThree quiet months followed by a large one
Software renewalsOften clustered, and easy to forget
Payroll timing quirksA payment can fall on either side of a month end
A large one-off purchaseDistorts any single month it falls into

Step four: split revenue into three buckets

Money in should not be a single hopeful line. Separate what is contracted, meaning signed and with a date, from what is probable, such as a customer renewing on their normal pattern, from what is merely hoped for, which is the pipeline. Steer by the first two. Keep the third visible as an upside line, so that you can see what it would do without letting it fund a decision. The reasoning behind that split has its own post in how to forecast revenue for runway.

Step five: choose a fixed refresh day

Pick one date each month and keep it. Not whenever someone feels nervous. The playbook suggests about half an hour: pull the actuals, compare them with what the model predicted, update the forward assumptions, and write down the new number and why it moved.

That last part turns bookkeeping into an early warning system. The movement is the signal, and you only see movement if you have a record of what you said last time. What to ask on that day is the subject of how to review runway monthly.

Run two versions of the answer

Once the inputs are in place, the model should produce two figures. One assumes the current pattern continues. The other includes the hires and spending you plan to add. Both come from the same structure, with the second simply carrying more outflow.

Having both in the model, rather than recalculating the second one occasionally, is what makes the difference between current burn and planned burn visible. The calculation itself is a short projection from those inputs.

FIKR Money, financial clarity: stop guessing your runway. See FIKR Money.

Then stop building

This is the instruction founders find hardest. Do not build scenario trees, unit economics dashboards or a long-range forecast until the simple version has been running for a full quarter and has actually changed a decision.

Sophistication here is a well-known form of procrastination. It is seductive precisely because it looks like rigour. Every hour spent adding branches to a model is an hour not spent asking why the number moved, and the second activity is where the value sits.

A useful test is to ask what decision the next addition would change. If you cannot name one, leave it out.

Who should own the model

A model needs one named person responsible for keeping it current. That person need not be an accountant. In a small company it is often the founder, and that is fine, provided the monthly refresh is in their diary rather than in their good intentions. What matters is that the answer to the question who updates this is a name, not a shrug.

It also helps if a second person can read it. A model that only its builder understands is another thing carried around in someone's head, which is precisely the habit the whole method is trying to end. The one-sentence definition of runway helps here, because it tells a newcomer what the model is and is not counting.

Signs that the model has gone stale

A stale model does not announce itself. It keeps displaying a number. These are the quiet signals that it is time to repair it rather than trust it.

  • Nobody can say when the actuals were last compared with the forecast.
  • The opening cash does not match what you see in the accounts today.
  • A large payment has gone out that the model never showed.
  • The categories have been replaced by a single catch-all line.
  • The team has stopped quoting the figure and started quoting a feeling.

Any one of these is a reason to spend the half hour. Left alone, they combine into a number that is distrusted in private and used in public, which is the worst of both outcomes.

How the model supports the decisions that follow

The point of all this upkeep is what it lets you do afterwards. A live model makes spending visible as a rate, and once you can see a rate you can steer it. It also gives you the raw material for the decisions in the later parts of the playbook: setting trigger points in advance, which is covered in how to set runway alert thresholds, and pricing a role before you approve it, which is the subject of how to price a hire in months of runway.

If the model cannot feed those decisions quickly, it is either too complicated or too out of date. Both are fixable, and the fix is nearly always to remove something rather than add it.

Free guide, 31 pages: The Runway Playbook. Send me the guide.

A checklist for the build afternoon

  • Opening cash is what is actually in the accounts today.
  • Money expected but not yet received sits on a separate line with dates.
  • Revenue is split into contracted, probable and hoped, and the base case uses only the first two.
  • Outflow is divided into five or six categories you can act on.
  • Annual and quarterly lumps are on the calendar in the right month.
  • Payroll timing and tax cycles are reflected rather than averaged away.
  • The accounts are connected, so the model updates without retyping.
  • A refresh date is in everyone's diary.
FIKR Money, financial clarity: stop guessing your runway. See FIKR Money.

Frequently asked questions

Do I need special software to build a runway model?

No. The structure is four inputs and fits on one screen, and the method is achievable with a spreadsheet and a disciplined half hour each month. What matters is that the model stays current.

How many expense categories should I use?

Five or six that you can make decisions about. More than that becomes bookkeeping and tends not to be maintained.

When should I add scenarios or a longer forecast?

Only after the simple model has run for a full quarter and has changed a decision. Before that, extra sophistication is mostly a way of delaying the useful work.

What is the single most useful habit?

Connecting the accounts so the model updates itself. The most common cause of a stale model is a manual update that someone has to remember.

The full method, with the checklists, is in The Runway Playbook, a free 31-page playbook.

LG

Luis Goncalves

// Founder & CEO at FIKR Space

Three-time founder. Built and exited Evolution4All before this. Now building FIKR Space — the operating infrastructure underneath every innovation ecosystem (startups, accelerators, governments, investors). Lisbon-based, works global.

Claim your founding member spot.

Every founding member is hand-reviewed. Tell us who you are and how you want to co-create — we get back inside 48 hours.

Role
Job Title
Country
Cookies

We use cookies.

Essential ones keep the site working. The rest only run if you say so. No dark patterns, no pre-ticked boxes.Privacy policy

Co-create with us. It's about building it together.Co-Create With Us