Most entrepreneurs sit down to financial projections for the first time because someone asks for them — a bank attaching them to a loan application, the HZZ (Croatian Employment Service) or HAMAG attaching them to a grant, or an investor before a meeting. As a result, they are often treated as a formality: a table to be filled in, submitted and forgotten.
That's a shame, because well-prepared projections do something far more useful than satisfying paperwork. They are the first serious test of whether your business idea makes economic sense — before you pour years of your life and a fair amount of money into it.
And that is the central point of this whole text: financial projections are not about filling in tables or trying to guess the future. They exist to show — first to you, and only then to the bank and the investor — that your idea has sound economic logic, realistic assumptions and a model that can hold up.
The logic behind the numbers always matters more than the numbers themselves. A table will accept any figure you type into it without complaint. The market, the bank and your bank balance will not.
What financial projections actually represent
Put simply, financial projections are the numerical translation of your business model. Everything you described in words in the business plan — who you sell to, at what price, with which costs, at what pace — here takes the form of concrete amounts over a period of usually three to five years.
This isn't one single table, but several connected parts that look at the same business from different angles. It helps to understand what each is for, without getting lost in accounting detail.
The revenue projection shows how much you plan to sell and at what price. It is the starting point for everything — every other figure hangs on this one.
The cost projection shows what running the business costs: both what is directly tied to the product or service, and the fixed part that you pay regardless of how much you sell.
The profit and loss statement (P&L) combines revenue and costs and answers the question everyone asks first: is the business making money or not. There's one thing many people overlook here — the P&L records revenue and costs at the moment they arise, not when the money actually lands in or leaves the account. That's why profit on paper and cash in the till are not the same thing, more on which later.
The cash flow projection tracks actual inflows and outflows of money — when your customers really pay you, when you pay suppliers, loan instalments, VAT. This is the part that shows whether you'll even be able to pay next month's bills, regardless of what the P&L says about profit.
The balance sheet is a snapshot of the position on a given date: what the firm owns, what it owes and how much the owner's equity is worth. Banks examine it carefully for larger investments and loans.
An experienced reader of projections — a credit analyst or an investor — rarely looks first at the profit figure in year five. They open the assumptions first: where does that level of sales come from, why exactly that price, how did you arrive at those costs. If the assumptions hold water, the rest of the model makes sense. If they don't, even the most beautifully formatted table won't help.
How to estimate revenue realistically
Revenue is where most business plans fall apart. Not because the figures are added up wrongly, but because they are set unrealistically.
A classic example is the "if we just take 1% of the market…" line of thinking. It sounds modest, but it is actually the weakest possible assumption, because it doesn't explain how you reach that 1%. The market is not a pool from which everyone can simply scoop out their percentage. Reaching each customer takes concrete work — an ad, a quote, a call, a referral — and each of those channels has its own capacity and its own cost.
That's why serious revenue projections are built "bottom up". You don't start from the size of the market, but from what you can actually do: how many enquiries you can generate, how many of them convert into customers, how much time, people and capacity you have to serve those customers.
A very practical way of thinking, especially for services, retail and an online shop, is to break revenue down into three elements:
number of customers × average purchase value × purchase frequency.
A hair salon, for example, doesn't plan revenue by "taking a share of the city's market". It plans it through the number of chairs, the number of working hours, the average duration and price of a service, and the realistic occupancy of appointments. An online shop starts from the number of visits, the percentage who buy and the average basket value. B2B sales start from the number of quotes sent, the percentage accepted and the average contract value. In every case, revenue comes out of concrete, verifiable figures, not out of wishful thinking.
And there are three things entrepreneurs regularly forget to build in:
Capacity. Revenue cannot grow indefinitely if you physically have no way to deliver. Two hairdressers cannot produce the revenue of five. A machine has a maximum number of hours. If a projection shows rising sales without additional people, shifts or equipment, the first thing any analyst will ask is — who is going to do it.
Seasonality. Very few businesses sell evenly across twelve months. Retail and e-commerce make a large share of annual turnover in the final quarter, hospitality and tourism live off the season, B2B services slow down in summer. A projection with the same revenue every month almost always reveals that growth hasn't been seriously thought through.
Ramp-up. A new business doesn't launch at full throttle in the first month. It takes time for the market to discover you, for referrals to start, for the sales funnel to fill. It is realistic that the first few months, and sometimes the first year, will be noticeably weaker than a "normal" year of operating. A projection that shows stable, high revenue from the first month says more about the owner's optimism than about the market.
If you remember only one thing from this whole section, let it be this: every figure in a revenue projection must have an answer to the question "where did you get that". If the answer exists and holds up, the projection is convincing. If the answer is "that's how it feels to me", it isn't a projection but hope with decimals.
How to estimate costs
If entrepreneurs overestimate revenue, they almost always underestimate costs. This is perhaps the single most common mistake in all of business planning.
The reason is psychologically understandable. When you build a business in your head, you naturally focus on what excites you — the product, the customers, the revenue. Costs are the boring part, and it's easy to unconsciously leave out the one that isn't in front of you at that moment. And it is precisely the sum of those "small", omitted costs that most often sinks profitability on paper.
It helps to divide costs into three groups, because that division later also enables you to calculate the break-even point. Direct costs are tied to each product or service sold — materials, the purchase value of goods, commissions, delivery. Fixed costs are paid regardless of sales — rent, base salaries, accounting, insurance. Variable costs grow with volume — additional materials, transaction fees, part of the logistics.
More important than the classification itself is that you don't skip any category. The list that is most often cut too short in practice looks roughly like this: salaries and — crucially — the contributions and taxes on them, not just the net amount; rent, the building reserve fund and all utilities; internet, phone, domain, hosting; marketing in a realistic, not a token, amount; logistics and fuel; software, licences and subscriptions; accounting and legal advice; maintenance and servicing of equipment; insurance; financing costs, meaning the interest and fees on loans; and taxes, including VAT and income tax or corporate profit tax.
A special warning applies to salaries. Owners often enter the net amount into the projection and think they're done. The actual cost of labour is significantly higher than the net, because contributions and taxes are added on top, and their level also depends on the legal form of the business. This is one of the reasons why the choice between a sole trader and a limited liability company is not merely an administrative formality but directly affects the cost side of your projections.
And one more habit that separates a serious plan from an optimistic one: build in a contingency reserve, usually five to ten percent of total costs, especially in the first year. Unforeseen costs are not the exception, they are the rule — you just never know in advance exactly which they will be. A conservative approach to costs doesn't mean pessimism. It means you'd rather pleasantly surprise both yourself and the bank than wonder six months in where the money went.
Why cash flow is often more important than profit
This is the part I would single out, if I had to choose, as the most important in the whole article. The reason is simple: firms rarely fail because they aren't profitable on paper. They fail because one day they have nothing left to pay what they must.
Profit and cash flow are two different things, even though at first glance they seem identical. Profit is the difference between revenue and costs accounted for at the moment they arise. Cash flow is the difference between the money that has actually come into and gone out of the account. Between those two moments — when you "earn" something and when the money truly lands — a great deal of time can pass, and it is precisely in that gap that perfectly healthy businesses go under.
Here's how a profitable firm runs out of money. Imagine that in one month you sell goods for €50,000 that you bought for €30,000. The P&L shows €20,000 in earnings and everything looks great. But the customer is B2B and pays in 60 days, while you paid the supplier immediately. On paper you earned €20,000; in the account you're missing the €30,000 you've already paid out, while the €50,000 arrives only in two months' time. If in the meantime salaries, rent and VAT have to be paid — and they do — the profitable firm is in a liquidity crunch.
A few typical things that "eat" cash even though they don't reduce profit: slow collection alongside fast payment to suppliers; inventory bought in advance that now just sits there, tying up capital; loan instalments, since only the interest passes through the P&L as a cost while the principal comes out of cash; equipment investments paid up front; and VAT, which you charge on issued invoices and pay to the state often before you've collected it yourself.
There is also what's called the growth paradox. The faster you grow, the more money you tie up in inventory and receivables before you collect it. Turnover rises, everything looks like success, yet there is less and less cash. Many owners experience their greatest liquidity stress at the very moment of fastest growth, without understanding why.
That's why a cash flow projection must track money month by month, with realistic collection and payment terms. Not when you'll invoice, but when the money will actually land. A business isn't paid with profit from a table. It's paid with cash in the account — and those are two columns that are far too often confused.
What banks really look at
When a bank reads your projections, it isn't looking for perfect numbers. It's looking for a credible model and one concrete answer: will you be able to repay the loan regularly.
Everything else revolves around that. What the bank cares about most is repayment capacity — does your business generate enough stable cash, after all costs and taxes, to cover the loan instalment with a buffer. For this they often use a ratio comparing your operating cash flow with your annual loan obligations. As a rule, they want to see that there is noticeably more cash than the instalment requires, not just barely enough to cover it. A tight ratio means the first weaker month puts repayment in question.
The second thing the bank looks at carefully is your own investment. How much of your own money you're putting into the project. That isn't just a financing question, but a signal — an entrepreneur who invests their own capital has a real reason for the business to succeed, and behaves differently from one who risks only other people's money.
The third is the conservatism and realism of the assumptions. Banks have seen thousands of projections and recognise overly optimistic figures in a second. A sudden jump in revenue without additional people, margins higher than across the whole industry, perfect collection without a single delay — all of that reduces trust, it doesn't increase it. The plan that often does best is one that, alongside the base scenario, also shows a worse one: "what if sales are 20% below plan, and how do we still repay the loan then". That doesn't reveal weakness, but seriousness.
In the end, the bank doesn't expect you to predict the future. It expects you to show how you think when the future doesn't go to plan.
The bank, then, doesn't buy optimism. It buys credibility. A more modest projection that holds water is always worth more than an ambitious one that falls apart at the first question.
What investors care about
An investor looks at the same document from a completely different angle than a bank. The bank cares whether you'll repay what you borrowed. The investor cares whether what they're investing in can grow enough to return their stake several times over.
For this reason, what an investor cares about first is growth potential and scalability — can the business grow without costs growing at the same pace. A business that needs twice the people and space to double its revenue is hard to scale. A business that can serve a much larger number of customers with the same team and the same base — software, a platform, a specialised service — is more attractive precisely for that reason. Margins matter here too: a high gross margin is often a sign that the model has room to grow.
The second layer is unit economics, the economics of a single customer. The investor wants to see that you understand two fundamental figures: how much it costs you to acquire a customer and how much that customer is worth over the entire relationship with you. If you acquire a customer for more than they bring you, growth only accelerates the loss. As a rough guide, it's often said that a customer's value should be at least three times the cost of acquiring them.
A high customer acquisition cost is not in itself a bad sign — the problem only arises when it isn't matched by a sufficiently large long-term value from that customer. The entrepreneur has to understand that difference before the investor does.
The third concept is burn rate and runway, especially for startups that aren't yet profitable. Burn rate is how much money you "burn through" each month until you start earning. Runway is how many months you can hold out on the money you have. The investor wants to know that the investment covers a long enough period — usually we're talking a year and a half to two — for you to reach the next stage or a new round.
What's worth stressing here: an investor is rarely impressed by the exact profit figure for year five. Everyone knows nobody can guess that number. What convinces them is the way of thinking behind the numbers — whether it's clear that you understand your market, your costs and the economics of a single customer. The investor doesn't invest in a table. They invest in an entrepreneur who knows what that table means.
What if revenue is lower than planned?
This is a question almost nobody likes to ask while preparing projections, and it is precisely the one that separates a serious plan from a nice wish. Because revenue will at some point be lower than planned. The question isn't whether it will happen, but what happens to your business when it does.
That's why good projections don't stop at one, most favourable outcome. They at least briefly run through a worse scenario: what if sales are 20 or 30 percent lower, if collection drags out, if the season underperforms. This isn't pessimism, but a matter of knowing in advance where your limit is. How much revenue you can afford to lose before you run out of money for salaries and the loan instalment — and what you would cut first when it comes to that.
That exercise is called sensitivity testing, and in practice it's simpler than it sounds. You don't need a new model. You take the existing one, pull two or three key assumptions downward and see what happens to the cash in the account. If the model holds even then, you have a business that can withstand a bad month. If it falls apart at the first 15 percent drop in revenue, it's better to find that out in a table than in reality.
And that's where the real value of conservative assumptions lies. A plan set a little below what you hope for leaves you room for error. A plan set on the best possible outcome has nowhere to go but down — and every deviation, even the smallest, immediately becomes a problem. The most dangerous projection isn't the one that's too cautious. The most dangerous is the one its author is unreservedly convinced of, because such a plan leaves no cushion for the moment when reality deviates.
The most common mistakes
Most mistakes in financial projections repeat from plan to plan, regardless of the line of business. If you recognise them before you submit the document, you're already ahead of most.
Overly optimistic revenue tops the list — aggressive growth with no explanation of how you reach it, regularly accompanied by that famous "1% of the market". Underestimated costs come right after, most often through forgotten contributions, taxes, marketing at a realistic amount and maintenance. Ignoring cash flow is the third big trap: a plan that looks only at profit, not at when money actually comes in and goes out.
There are also mistakes that betray carelessness. Flat sales across the whole year, with no seasonality and no slower start. Unrealistic margins, higher than across the entire industry, without a word of explanation as to why you of all people would be the exception. A mismatch between the text and the tables — in the description you write that you're focused on B2B, but the tables draw revenue from B2C. That's a small thing that immediately tells an experienced reader the plan and the model weren't prepared together.
And two that are almost the hallmark of an unprepared plan: a ready-made Excel template downloaded from the internet, filled with someone else's numbers without a single adjustment to your business — the numbers are there, but there's no story behind them; and a single growth line with no scenarios, the same rate every year, as if the market didn't exist. Both banks and investors increasingly want to see at least a base and a worse scenario.
The common denominator of almost all these mistakes is the same: the projection was made to impress, not to be true.
How to make projections that come across as convincing
Now the reverse — what separates projections that are believed from those that raise suspicion.
First and most importantly, be conservative. It's better to slightly underestimate revenue and slightly overestimate costs than the other way round. A projection you later exceed builds trust. A projection you fail to reach destroys it, both with the bank and with yourself.
Second, be transparent about your assumptions. Beside every important figure, briefly note where it comes from — from experience, market data, comparison with competitors or a trial period. Assumptions are the heart of the model; the numbers are merely their consequence. A reader who sees clear assumptions trusts the whole document, even when the results are modest.
Third, connect the projections with the rest of the plan. The revenue structure must match the channels and customers you described in the marketing section. Costs must follow what you said about people, equipment and location. Revenue must be covered by capacity. When all of that "fits" together, the document reveals that behind it stands a person who understands their own business.
One small thing that few people do helps here too: alongside the tables, add short comments explaining the larger jumps. "2027 — opening a second location." "2028 — entry into a new market." This connects a figure to a business decision for the reader and shows that growth wasn't entered at random but planned.
Good financial projections don't try to dazzle with large revenue. They show that the entrepreneur understands their own business — how it earns, what it costs, when money comes in and when it goes out. That is what creates trust, both with the bank and the investor, and frankly with the owner too.
Frequently asked questions
Do I need projections if I'm just starting a sole trader business and have no historical data? Yes, precisely because you have no history. When there are no past figures, a projection is the only way to check in advance whether the idea makes sense. It starts from assumptions based on capacity, prices and comparable businesses — and it's exactly that exercise that often reveals a problem before you spend your first euro.
How many years ahead should projections be made? For a bank, three years is standard, sometimes with the first year broken down monthly. For investors, five years is more common, because they're interested in long-term growth potential. The first year is always the most detailed; the further out the period, the rougher the figures and the more they speak of direction than of precision.
What's the difference between profit and cash flow, put simply? Profit is how much you've earned on paper. Cash flow is how much money you actually have in the account. They differ because you recognise revenue when you issue the invoice, while the money arrives only when the customer pays — which can be much later. A firm can be profitable and run out of money at the same time.
How do I set a realistic growth rate in the first years? By tying it to something concrete: additional people, a larger marketing budget, a new location, a new channel. Growth that rests on nothing but a percentage isn't serious. Plan the first few months modestly because of the ramp-up period, and justify every later jump with a concrete reason.
How do I include seasonality in the projection? By planning revenue monthly, rather than dividing the annual amount by twelve. Look at how your line of business breathes through the year — which months are strong, which are weak — and mirror that in the table. Cash flow is the most sensitive to the season, because in weak months costs often stay while revenue falls.
What if the projections don't come true? That's expected, and nobody asks you to guess to the euro. Projections aren't a promise, but a model. The value lies in comparing the plan with reality and understanding why a difference exists — that helps you make better decisions. That's exactly why there are multiple scenarios.
Can I prepare the projections myself in Excel, or do I need professional help? You can put them together yourself, and that's actually useful, because it forces you to think about every number. Excel is a perfectly sufficient tool. Professional help is worth it when you're going before a bank or investor, when the model is more complex, or when you want to be sure the assumptions and tables hold water and everything is mutually consistent.
How much of my own money does a bank expect me to invest? It depends on the type of financing and the bank, but the rule is simple: the more of your own capital you put in, the calmer the bank is. Your own investment is a signal to the bank that you too are bearing the risk. A project financed entirely with other people's money has a harder time getting through, no matter how good it looks on paper.
What is break-even and how do I quickly calculate it? Break-even is the level of sales at which revenue covers all costs — neither a loss nor a profit. A rough formula: divide fixed costs by the difference between the selling price and the variable cost per unit. The result tells you how much you have to sell just to operate "at zero". It's one of the most useful figures for checking how serious an idea is.
How do I show scenarios without the plan becoming overcomplicated? Three are enough: a base (realistic), a conservative (worse) and an optimistic one. You don't have to build three completely separate models — change a few key assumptions, say lower revenue by 20% or extend the collection period, and show how the result changes. That shows you've thought about risk too, not just the prettiest outcome.
Conclusion
Financial projections don't primarily serve the bank or the investor. They serve you.
That's easy to forget when you're preparing them under the pressure of a deadline and someone else's request, but it's the truth. The bank and the investor see the projections once. You live every day with the consequences of the decisions that flow from them. A well-prepared model reveals problems while they're still on paper and free — over-optimistic revenue, an underestimated cost, a month in which the money runs out — rather than later, when they become real and expensive.
That's why the logic behind the numbers always outweighs the numbers themselves. The goal isn't an impressive profit figure for year five. The goal is to prove, first to yourself, that the business makes economic sense: that prices cover costs, that capacity can sustain the planned sales, that money arrives before it leaves. A projection that shows this honestly is worth more than any that merely looks good.
If you're only just considering starting a business, part of that logic begins even before the projections — already at the decision of sole trader or limited liability company, because the legal form affects costs, taxes and the very way you'll build the numbers in the first place.
And when it comes to actually preparing the business plan and financial projections — whether for a bank, a grant or your own check — it's worth having someone who has already built such models and knows where they most often break. That is part of what I do with entrepreneurs through Meridian Consulting, but regardless of whether you prepare the projections on your own or with help, the most important thing stays the same: that in the end you have a model that is true and that genuinely helps you make a better decision.