Every transaction is recorded twice — where the value came from and where it went — so the books always balance and every euro is traceable. Think of it as conservation of money, enforced by bookkeeping.

The bank balance shows liquidity today — nothing about money customers owe you, bills coming due, or whether this month was profitable. The three statements exist to answer those questions.

Most startups need 30–80. Every account should map to a report line you actually want to see; anything else is noise. Too few accounts and you can't answer questions, too many and coding gets inconsistent.

Cash accounting records money when it moves. Accrual records revenue when earned and expenses when incurred — which makes months comparable and is the basis investors expect. The bank statement tells you liquidity, never performance.

Accrual profit includes revenue you've earned but not collected, and excludes cash you've paid for future benefits. The gap lives on the balance sheet — usually in receivables, prepayments, or deferred revenue.

As soon as you invoice ahead of delivery, sign annual contracts, or plan to raise within eighteen months. Retrofitting history under diligence pressure is far more expensive than starting right.

Net income from the P&L flows into equity on the balance sheet; the cash flow statement starts from net income and adjusts for non-cash items and working-capital changes to explain the real cash movement. One system, three views.

Customers are paying you slower. Watch DSO — if the trend continues, growth is being funded by your runway instead of your customers.

The cash your core business generated or consumed — net income stripped of non-cash items and adjusted for changes in AR, AP, and inventory. It's the lie detector of the three statements.

Because the customer has paid and you still owe the service — it's an obligation to deliver, not income yet. It converts to revenue month by month as you deliver. It's also good news: customers prepaying is the cheapest financing a startup can get.

Ratably over the service period — €1,000 per month on a €12,000 annual deal — regardless of the upfront invoice. Recognizing it all at signature fabricates growth that diligence will unwind.

Bookings are contracts signed, billings are amounts invoiced, revenue is value delivered. Only revenue belongs on the P&L; the other two are operational metrics.

Only if the setup has standalone value delivered at that moment. If it's inseparable from the ongoing service, spread it over the expected customer relationship.

As usage occurs. At month-end, estimate unbilled usage and book it as accrued revenue, then true up when the metered data lands. Waiting for the invoice understates every month.

Time & materials: Recognize as billed. Fixed-fee: Recognize by percentage of completion or delivered milestones. Recognizing 100% at invoice on a half-finished project overstates the month and borrows from the next one.

No — deposits and advances are liabilities until earned. Rental deposits in particular are never revenue; they're money you may have to give back.

Straight-line over the lease term, even if payments are uneven. Free months and escalating rents get smoothed across the full term.

Days sales outstanding — the average days between invoicing and collection. Judge it against your own terms: net-30 with a DSO of 35–40 is normal; 60+ means collections are broken. The trend matters more than the level.

Systematize it: a courteous reminder before the due date, then at +7, +15, and +30, escalating in firmness. Automated dunning pulls weeks out of collection cycles without damaging relationships — the machine is politely relentless so you don't have to be.

Provision when collection is doubtful — commonly 90+ days overdue with no engagement — and write off when it's realistically dead. Carrying dead invoices overstates assets and the quality of your historical revenue.

No. VAT is money you collect on behalf of the tax authority — a €10,000 invoice plus 22% VAT is €10,000 of revenue and €2,200 of liability. Booking gross amounts inflates revenue and sets up a nasty surprise at filing time.

Legally it's in your account, but it was never yours — it's owed at the next return. Treat the VAT float as off-limits cash. Spending it is one of the fastest ways small companies die of a tax bill.

For cross-border B2B services in the EU, the buyer self-assesses VAT instead of the seller charging it. It applies both ways: your invoices to EU business customers, and the US SaaS tools you buy. Missing it is a common audit finding.

Cash paid for a future benefit — annual insurance, yearly licences. It's booked as an asset and released to expense over the covered months, so June doesn't carry twelve months of insurance cost.

A cost you've consumed but not yet been billed for — legal work in progress, earned bonuses, utilities. You book it in the month the benefit happened, and it reverses when the real invoice arrives.

No — only above your materiality threshold. Spread the €12,000 licence, expense the €40 tool. Write the threshold down and apply it consistently; a fast, 95%-right close beats a perfect one that's six weeks late.

Everything that scales with serving customers: hosting, embedded third-party software and APIs, payment processing, support, and customer success. Sales, marketing, and R&D stay in opex. An honest definition matters more than a flattering margin.

First check the COGS definition — most 'margin gaps' are definition gaps. If support and CS salaries are in opex while the benchmark includes them, you're comparing different numbers.

Yes — COGS, sales & marketing, R&D, and G&A. One undifferentiated personnel line makes gross margin meaningless and benchmarking impossible. It's the single change that unlocks real margin analysis.

On the due date — not on receipt. Paying everything same-day quietly shortens your runway for zero benefit. Payment terms are free financing; use them.

Three cheap controls: all invoices flow into one system, two-person approval above a threshold, and out-of-band verification of any new or changed supplier bank details. Most invoice fraud fails against any one of them.

Substantially more than gross salary — employer social contributions add roughly 25–35% on top in much of the EU, before bonuses, equipment, and tools. Budget hires at total employer cost or your plan is fiction. Verify the rate for your country.

Usually a 13th-month salary or annual bonus that was never accrued monthly. Spread the obligation across the year and December becomes a normal month — the cost belonged to all twelve.

Yes, and for the law more. Contractors are simpler to book but misclassifying de-facto employees creates back-tax and penalty risk that surfaces at diligence. Review classifications annually.

Because you bought years of future use, not a month of cost. The purchase becomes an asset and depreciates over its useful life — €120k of hardware over four years is €2.5k per month, matching cost to benefit.

You can when criteria are met, but think strategically: capitalizing flatters burn today and gets unwound by skeptical investors tomorrow. Many startups expense everything for simplicity and credibility. Decide deliberately and document it.

Earnings before interest, tax, depreciation, and amortization — operating profit with the non-cash and financing items stripped out. Useful for comparing operating performance; dangerous as a proxy for cash flow.

Inventory is an asset until sold — the cost hits the P&L as COGS only when the sale happens. Expensing purchases on receipt makes margins swing with order timing instead of sales performance.

The true unit cost: purchase price plus freight, duties, and handling. Ignoring it overstates unit margins — sometimes by enough to make an unprofitable product look like a winner.

When stock is obsolete, damaged, or unlikely to sell at full value. Review slow movers quarterly. 'Profitable' companies with a warehouse of dead stock are neither.

Building the forecast from operational drivers — hires, pipeline, pricing, conversion — instead of 'revenue plus 10%'. When reality diverges, you can see which assumption broke and fix the plan, not just the number.

A weekly cash view thirteen weeks out, built from actual receivables, payables, and payroll dates. It's the tool that catches a payroll miss two months early, while there's still time to act.

Tax you'll save later — mostly accumulated losses that can offset future profits. It only goes on the balance sheet when those profits are probable; recognizing it on hope is a classic audit reversal.

Potentially — many jurisdictions limit carryforwards after significant ownership changes, and some expire. Check before each funding round, not after.

Many startups are — credits and patent-box-style incentives can be worth six or seven figures. Tag eligible costs through the year instead of reconstructing them at filing time.

As a financial instrument on the balance sheet — typically a liability or equity-like item depending on terms and jurisdiction — never as revenue. Yes, booking a SAFE as income happens; yes, diligence notices.

Because equity given to employees is compensation — the P&L should reflect the cost of the work it pays for, even though no cash leaves. Ignoring it until an audit means restating history.

Yes — cap table, corporate registry, and ledger must tell one story. Reconcile after every transaction; divergence found at diligence costs weeks.

Revenue quality (is it recognized right, is it recurring), margin reality, completeness of liabilities, and whether your deck metrics tie to the ledger. Most valuation haircuts trace to findable, fixable issues.

Monthly accrual statements for 24+ months, deferred revenue and AR/AP schedules that tie to the statements, metric definitions, material contracts, cap table, and tax filings — maintained continuously, not assembled in a panic.

Before — always. Restating during diligence looks like concealment even when it's just housekeeping, and it hands the other side negotiating leverage on a plate.

When the founder spends more than a day a week on finance ops, or complexity (multi-entity, inventory, 20+ people) outgrows the external accountant. Usually the first hire is a controller — not a CFO.

Before it's forced to — by statute, a lender, or an acquirer. A first audit under deal pressure is slow and expensive; a voluntary one a year earlier is a dress rehearsal that surfaces issues cheaply.

From doing to reading: At €1M you touch everything, at €20M your job is to read the monthly pack in thirty minutes, trust it, and ask the three questions that matter. Building the system that earns that trust is the work in between.