How to Build a Simple Financial Model (No Banker Needed)

Most business numbers tell you whether you are alive right now. Almost none of them tell you whether you will still be alive in six months. That gap is exactly where founders get blindsided.

A financial model closes the gap. And no, it is not a thing only a banker in a suit can build. It is a spreadsheet that says: “If these few things happen, here is the money that comes out.” If you can multiply and add, you can build one this afternoon.

Why this matters

Looking back tells you if you are alive. Looking forward tells you if you will stay alive.

The founders who run out of cash rarely do it because the business was doomed. They do it because the slow leak never showed up on a spreadsheet until the bank balance hit zero. A plan for the next 12 months, written in numbers, turns “we ran out of money in March” into “we can see March getting tight, so we act in January.”

This guide gives you the whole toolkit: the three words people confuse, two ways to plan revenue, how to model so your numbers can actually be improved, and the monthly habit that catches problems early.

Three words people mix up: budget, forecast, and model

These get tossed around as if they mean the same thing. They do not, and the difference is the whole game.

• Budget - your plan for a period, usually one year. It is a promise to yourself: “We plan to make $40,000 a month and spend $30,000.” You set it once at the start and mostly leave it fixed, so you have a fixed line to measure yourself against.
• Forecast - your best honest guess of where you are actually heading, updated as new facts arrive. The budget says what you planned. The forecast says where you are really going, based on what you now know.
• Financial model - the machine that produces both. You feed it assumptions (price, customers, costs) and it calculates revenue, profit, and cash. Change one assumption and every number downstream updates by itself.

An analogy. Picture a road trip. The budget is the plan you made at home: “We’ll drive 500 miles a day and arrive Friday.” The forecast is your GPS recalculating after you hit traffic: “Now arriving Saturday morning.” The model is the GPS app itself - the engine that turns speed and distance into an arrival time.

Two ways to plan revenue: top-down vs bottoms-up

Before any model, you need a number for revenue - money coming in from sales. There are two directions to reach it, and smart founders use both, then compare.

Top-down: start big, shrink to your slice

You start with the total market and take a small share. TAM stands for Total Addressable Market - all the money everyone could possibly spend on your kind of product.

“The print-shop software market is worth $2 billion a year. If we capture just 0.1% of it, that’s $2,000,000,000 × 0.001 = $2,000,000 a year.”

Easy to say. But where did “0.1%” come from? Usually thin air. Top-down numbers feel impressive and prove almost nothing.

Bottoms-up: start with what you actually do

You build revenue from your real activities: how many customers, at what price, how often. Slower, but far more honest - because every number is something you can actually change.

Two common formulas:

Business type	Revenue formula
Subscription / B2B	Customers × Price per customer
E-commerce / storefront	Traffic × Conversion rate × Average Order Value

Quick definitions: Traffic is visitors to your site. Conversion rate is the percentage of visitors who actually buy. Average Order Value (AOV) is the average dollars per order.

Worked example. Your store gets 100,000 visitors a month. 1% of them buy. Each order is worth $100.

1. Buyers: 100,000 × 1% = 1,000 buyers.
2. Revenue: 1,000 × $100 = $100,000 for the month.

Now look at the levers. Lift conversion from 1% to 1.2% and you earn $120,000 from the same traffic. You did not need more visitors - you needed a slightly better checkout. That insight is invisible in a top-down number.

	Top-down	Bottoms-up
Starts from	Total market size	Your own activities
Good for	Yearly big picture, board slides	Monthly operating, knowing what to fix
Main risk	Fantasy (“just 1% of a huge market”)	Can be too cautious
Editable levers	Few	Many (price, conversion, traffic)

Best practice: build bottoms-up as your real plan, then build a quick top-down as a sanity check. If your bottoms-up plan implies you’ll capture 5% of the entire world market, something is broken. Founders who compare both tend to hit their goals more reliably than those who use only one.

Driver-based modeling: the one habit that makes a model useful

A driver is an input number that drives the result - price, customer count, conversion, churn. An assumption is your guess for that driver before you have real data.

Here is the golden rule:

Never type a final answer into a cell. Type the drivers, and let the spreadsheet calculate the answer.

If you write revenue as “$100,000” by hand, you have learned nothing and you can change nothing. If you write traffic × conversion × AOV, you can move any driver and instantly see the effect ripple through profit and cash.

This is the difference between a model that thinks and a model that just stores a guess.

Three scenarios: base, best, and worst case

The future is unknown, so do not pretend you know it exactly. Build three versions by changing your key drivers:

• Base case - what you genuinely expect. Your honest middle bet.
• Best case - things go well: higher conversion, faster growth.
• Worst case - things go badly: slow sales, higher costs. This one matters most, because it tells you when you would run out of cash.

Example. Base case: 1.0% conversion gives $100,000 a month. Best: 1.4% gives $140,000. Worst: 0.6% gives $60,000. If your costs are $80,000 a month, the worst case loses $20,000 a month - and now you know exactly how much cash buffer to raise before you start, instead of discovering it the hard way.

Because you built drivers, all three scenarios should come from changing just two or three assumption cells. You are not rebuilding the sheet - you are turning a couple of dials.

Budget vs rolling forecast (use both)

An annual budget is set once and frozen, so you can hold yourself accountable to it. A rolling forecast is updated every month and always looks the same distance ahead: as one month finishes, you add a new month at the far end, so you always see roughly 12 months forward.

An analogy. The annual budget is a printed map you bought in January. The rolling forecast is live GPS - it re-routes every time the road changes.

You do not have to pick one. The strongest setup keeps both. The frozen budget gives you accountability (did we hit the plan?). The rolling forecast gives you agility (where are we really heading now?).

Variance analysis: compare plan to actual, then act

Variance simply means the difference between what you planned and what really happened. Variance analysis is the monthly habit of putting plan and actual side by side, asking why they differ, and then doing something about it.

Example.

• Planned revenue: $100,000. Actual revenue: $85,000.
• Variance = $85,000 − $100,000 = −$15,000, a $15,000 shortfall.
• As a percentage: −$15,000 ÷ $100,000 = −15%.

Now the real work - why? Because you built drivers, you can see that traffic was on target and AOV was fine, but conversion fell from 1.0% to 0.85%. The problem is on your website, not in your marketing. That is the power of variance analysis: it points you at the exact thing to fix.

A simple rule keeps you sane: investigate variances bigger than a set threshold, often around 5 to 10 percent. Tiny wiggles are just noise, and chasing them wastes your day.

A simple month-by-month layout

Here is a clean shape. Months across the top, line items down the side. Every revenue figure is calculated from the drivers above it - nothing is typed in by hand except the assumptions.

Line	Jan	Feb	Mar
Traffic (driver)	100,000	110,000	121,000
Conversion (driver)	1.0%	1.0%	1.1%
AOV (driver)	$100	$100	$100
Revenue (calc)	$100,000	$110,000	$133,100
Costs (driver)	$80,000	$82,000	$85,000
Profit (calc)	$20,000	$28,000	$48,100
Cash at end (calc)	$120,000	$148,000	$196,100

The flow runs top to bottom:

  Drivers (you type these)
  traffic  conversion  AOV  costs
        \      |       /     |
         v     v      v      |
        REVENUE = T x CR x AOV
                  |          |
                  v          v
              REVENUE  -  COSTS  =  PROFIT
                                     |
                                     v
            CASH(last month) + PROFIT = CASH(now)

Common misconceptions

“A financial model predicts the future.” No. It is a thinking tool, not a crystal ball. Its job is to show you which levers matter and where the cash gets tight, not to be right to the dollar.

“Hard-coding a number is fine if it’s roughly right.” Typing “Revenue = $50,000, growing 10% every month” hides the truth. A hard-coded number can’t be argued with or improved. What if you can’t actually win those customers? Model the drivers underneath it so you can see the assumption you’re really making.

“More decimal places means more accuracy.” Writing “$1,284,447.18” for next year is fake precision - you don’t know it to the penny. A round “$1.3M” is more honest and easier to trust.

“A flat line that rockets up at month 12 shows ambition.” That is the hockey-stick fantasy, and investors have seen a thousand of them. If your line bends sharply upward, you must point to the exact driver that changes - a new channel, a price rise, a big partner - and defend it. No reason, no credibility.

How to use this

1. Open one spreadsheet and make an Assumptions tab. Put every driver - price, traffic, conversion, churn, salaries - in one clearly labeled place. The rest of the sheet only does math on those cells.
2. Build revenue bottoms-up. Write traffic × conversion × AOV (or customers × price) as a formula. Never type the revenue answer directly.
3. Run a top-down sanity check. If your plan implies an absurd share of the market, fix your assumptions before anyone else has to.
4. Make three scenarios. Copy your base case, then change two or three driver cells for best and worst. Read the worst case carefully - that is your cash buffer.
5. Color-code your cells. Typed-in drivers one color (say blue), calculated cells another. Now you can see at a glance what’s a guess and what’s math.
6. Freeze a budget, then forecast monthly. Lock the annual plan for accountability. Each month, update the rolling forecast so it always looks 12 months ahead.
7. Do variance analysis every month. Put plan next to actual, flag anything off by more than 5 to 10 percent, trace it to the driver that moved, and act.
8. Round, and keep it current. A simple sheet you refresh monthly beats a fancy one nobody has touched since spring.

Conclusion

The single thing to remember: a financial model is not a prediction, it is a steering wheel. Explicit drivers on one tab, three honest scenarios, and a monthly habit of comparing plan to actual - built that way, it catches cash surprises long before they can kill you.

There is a deeper question lurking underneath all of this, though. Your model assumes customers keep paying, but how long do they really stay, and what is each one actually worth over their lifetime? That is where churn, retention, and customer lifetime value come in - and once you can model those, your forecasts stop being hopeful and start being honest.