Build VC Ready Financial Model for Singapore Startups with 3 Scenarios

Published: 4 October 2026 · Reviewed by Ray Tay, Co-Founder & Managing Director, VIVOS PTE. LTD. (ACRA Filing Agent FA20240323 · MOM EA Licence 24S2425)

A startup financial model is a linked three-statement projection, income statement, balance sheet, and cash flow, driven by assumptions about revenue, hiring, burn, and runway. Investors expect one from the seed round onward, and most founders also use it internally for hiring decisions, budget control, and board reporting. Done well, it becomes the single reference document for the business.

TL;DR:

• A seed-stage financial model must include a fully linked three-statement projection with monthly detail for at least the first 18 months, covering revenue, COGS, headcount, and working capital.
• Assumptions should be built bottom-up, clearly documented, and separated into dedicated tabs, with scenario planning (base, upside, downside) to test business robustness.
• Regular monthly updates using actuals, variance reconciliation, and rolling forecasts are vital to keep the model relevant and prevent cash shortages.
• Incorporate jurisdiction-specific inputs like tax rates, GST thresholds, employer CPF, and incorporation timing, especially for Singapore startups.
• A clear cap table linked to fundraising scenarios and dilution projections helps assess ownership impacts, while validation against actuals maintains model credibility.

Table of Contents

• What should a seed-stage model include?
• How do I build the core inputs and assumptions?
• How do I build a startup financial model step by step?
• What tabs should a founder’s spreadsheet have?
• How do I run scenarios and sensitivity analysis?
• How do I keep the model useful month to month?
• What Singapore-specific inputs feed the model?
• How does equity dilution fit into the model?
• How do fundraising rounds affect the cap table?
• How do I validate the model against real data?
• What founders consistently get wrong about their own models
• How VIVOS supports founders building and feeding their model
• FAQ
• Sources

What should a seed-stage model include?

A seed-stage model needs all three financial statements linked together, not just a revenue spreadsheet with a growth curve attached. Each statement answers a different question, and investors read them as a set because each one checks the others.

The income statement, or P&L, shows revenue and expenses over time and answers whether the business model works at all. It is driven by the same assumptions that drive the rest of the model: pricing, unit volumes, headcount costs, and operating expenses. The balance sheet shows what the company owns and owes at a point in time, including cash, receivables, equipment, and any debt. It is the reconciliation layer: if the balance sheet does not balance, an assumption elsewhere is wrong.

The cash flow statement tracks the actual movement of money in and out of the bank account, and for an early-stage company it is the most consequential of the three. A company can show a profit on paper and still run out of cash because customers pay late or because a large capital purchase hits the bank account before revenue catches up. Investors commonly ask for a three-to-five year forecast with monthly granularity for at least the first 18 to 24 months, tapering to quarterly detail further out.

A model built to this standard typically includes:

• Monthly detail for the near term, with quarterly or annual detail further out.
• A dedicated assumptions tab so every driver is visible and auditable in one place.
• At least three scenarios: base, upside, and downside.
• A headcount schedule tied to hire dates and fully loaded costs.
• Working capital assumptions, not just revenue and cost lines.

Investors weigh the defensibility of a founder’s assumptions more heavily than the precision of a five-year revenue number, because nobody, including the founder, can forecast year four with real accuracy. What they are testing is whether the founder understands the levers of the business well enough to run it under pressure.

How do I build the core inputs and assumptions?

Every number in the model traces back to an assumption, and the quality of the model is really the quality of its assumptions. Vague inputs produce a vague model, no matter how polished the formatting looks.

1. Build revenue bottom-up. Start from units times price, or from sign-ups times conversion rate times average revenue per user, rather than picking a top-line growth percentage and applying it blindly. A channel-level build (direct sales, self-serve, partnerships) makes the revenue line defensible because each channel can be checked against a real conversion funnel.
2. Separate cost of goods sold by business type. A SaaS business carries hosting, customer support, and payment processing fees as COGS; a physical product business carries materials, manufacturing, and fulfillment. Mixing these into one generic “cost of sales” line hides the gross margin story investors want to see.
3. Build a headcount schedule, not a headcount total. List each role, the planned start month, and the fully loaded cost, which includes salary, statutory contributions such as CPF for local hires, and any benefits. A model that ties payroll jumps to specific hire dates is far easier to defend than one that shows a lump “headcount cost” growing by a flat percentage each quarter.
4. Model working capital explicitly. Days sales outstanding, days inventory outstanding, and days payable outstanding determine when cash actually lands, independent of when revenue or expense is recognized. Ignoring these produces a model where the P&L looks healthy while the cash account quietly drains.
5. Use input-cell conventions from day one. Color assumption cells blue and formula cells black, lock formula cells, and keep every hardcoded number confined to the assumptions tab. This single habit is what separates a model an investor can audit quickly from one they have to pick apart line by line.

Pro Tip: Keep every assumption in one visible cell with a label next to it: an investor who cannot find the input behind a number will assume you cannot either.

How do I build a startup financial model step by step?

Building the model in the right order prevents the rework that comes from discovering a missing input halfway through. Each tab should feed the next, so sequence matters as much as content.

1. Set up the assumptions tab first. List every driver you will need: pricing, growth rate, churn, headcount costs, rent, software spend, and tax rate.
2. Build the revenue model next, using the bottom-up drivers from the assumptions tab, broken out by month.
3. Build the headcount schedule, with each role’s start date and fully loaded cost pulling from the assumptions tab.
4. Build operating expenses, separating fixed costs like rent and software subscriptions from variable costs that scale with headcount or revenue.
5. Build the capex and balance sheet inputs: equipment purchases, deposits, and any financing.
6. Assemble the P&L by linking revenue, COGS, headcount, and OpEx into a single monthly statement.
7. Build the cash flow statement from the P&L and balance sheet movements, since cash flow is derived, not independently invented.
8. Add a dashboard and cap table summary so the output is readable without opening every tab.

Keep the whole build at monthly granularity for at least the first two years. Quarterly models hide the month a company actually runs out of cash, which defeats the purpose of building one. Run a parity check every month: does the balance sheet balance, and does the ending cash balance on the cash flow statement match the cash line on the balance sheet. If it does not, an assumption somewhere is double-counted or missing.

Common pitfalls to avoid:

• Hardcoding a growth rate instead of deriving it from unit economics.
• Forgetting working capital, which causes cash flow to diverge silently from the P&L.
• Mixing formulas and hardcoded numbers in the same cell, which breaks the audit trail.
• Building annual detail when monthly detail is what investors and operators actually need.

For investor review, document every assumption in plain language next to the number it drives, and be ready to explain why the churn rate or the sales cycle length is what it is. A widely used modelling checklist recommends exactly this structure: bottom-up drivers, a separate assumptions tab, monthly forecasts, and at least three named scenarios, because that combination is what lets an outsider follow the logic without a walkthrough call.

What tabs should a founder’s spreadsheet have?

A consistent tab structure makes the model usable by anyone who opens it, including a future finance hire or a due diligence team. The exact order matters less than having every tab present and clearly labeled.

Google Sheets works well for most seed-stage founders because collaboration with co-founders and investors is simpler and version history is automatic. Excel remains preferable for more complex models with heavy scenario toggles or when an investor’s finance team expects a specific file format. Whichever tool you choose, keep one master version, name file copies by date, and avoid emailing spreadsheet attachments back and forth, since conflicting versions are one of the more common sources of investor confusion during diligence.

How do I run scenarios and sensitivity analysis?

A single-scenario model tells an investor nothing about how the business behaves under stress, which is exactly the condition most startups eventually face. Three scenarios, base, upside, and downside, give a far more honest picture.

The base case reflects your most likely assumptions for growth, churn, and hiring pace. The upside case tests what happens if a channel outperforms or an enterprise deal closes early. The downside case is the one investors read most carefully, and it should include specific, named stress tests rather than a vague “slower growth” toggle:

• A 20% miss on revenue against the base case, applied from month one onward.
• A three-month delay in a planned key hire, such as a head of sales.
• A longer sales cycle that pushes collections out by 30 to 60 days.

Under each scenario, the numbers investors check first are runway, the burn multiple (net burn divided by net new revenue), and the month the business reaches break-even or its next milestone. A downside case that still leaves 12 months of runway is one of the more reassuring things a founder can show a term sheet negotiation. That single number often carries more weight than a strong base-case growth curve, because it demonstrates the business survives a miss rather than just performing well when everything goes right.

Present sensitivity with a small toggle on the dashboard tab that switches the whole model between scenarios, rather than three separate files. This lets an investor flip between cases live on a call and ask “what if churn is two points higher” without waiting for a follow-up email. Keep the toggle simple: a single dropdown or input cell that feeds every downstream tab is enough, and it signals that the model is built to be interrogated, not just presented once and filed away.

How do I keep the model useful month to month?

A financial model that is built once for a fundraise and never opened again stops being useful within a quarter, because assumptions drift and actuals diverge from plan almost immediately. Treating the model as a living document is what keeps it relevant.

1. Set a monthly reporting cadence: close the books, pull actuals, and update the forecast within the first two weeks of the following month.
2. Pull actuals from accounting software such as Xero or QuickBooks directly into a dedicated actuals column next to the forecast, so variance is visible at a glance rather than buried in memory.
3. Reconcile variances line by line: if marketing spend ran 15% over plan, trace it to the specific channel before adjusting the forward forecast.
4. Roll the forecast forward every month so you always have a 12-month view ahead, rather than a fixed forecast that gets shorter every month until it expires.
5. Set decision rules tied to the output, not just gut feel: for example, if runway drops below 12 months, cut discretionary spend by a fixed percentage or pause the next planned hire until the trend reverses.

A rolling 12-month forecast updated on a monthly cadence materially reduces the risk of an unexpected cash shortfall, because problems surface as a trend across two or three months rather than as a crisis in the fourth. Founders who skip this step tend to notice cash problems only when the bank balance itself looks low, which is usually too late to fix without an emergency raise.

What Singapore-specific inputs feed the model?

Foreign founders setting up a new Singapore company need several jurisdiction-specific figures in the assumptions tab, since getting these wrong distorts both the tax line and the cash flow timing. IRAS sets the prevailing corporate tax rate at 17%, and qualifying new start-up companies can access a partial tax exemption on the first chunk of chargeable income, which lowers the effective rate in the early years. As of 2026, CIT rebates announced in recent budgets may further reduce the cash tax payable, so the model’s tax line should reference the current exemption and rebate schedule rather than a flat 17% applied from year one.

GST registration becomes mandatory once taxable turnover crosses the current threshold, assessed on either a retrospective or prospective basis, and IRAS is the authority to check for the current threshold and assessment method. A model should flag the month projected turnover approaches that figure so GST collection and remittance can be built into cash flow before registration becomes due, not after.

Other inputs that belong in the assumptions tab:

• Employer CPF contributions for local hires, which add to fully loaded headcount cost and should be checked against the current CPF contribution rates.
• The Employment Pass salary floor for any foreign hire who will need a work pass, which sets a minimum headcount cost line for that role.
• Incorporation and compliance timing: ACRA filing, nominee director appointment, and registered address setup all affect when the entity can open a bank account and start billing.

An ACRA-registered filing agent provides Singapore company incorporation for foreign founders, including nominee resident director, registered address, and corporate secretary services, with accounting support that feeds these figures directly into the model; similar services are also offered in Malaysia, Hong Kong, and the UAE, in English and Mandarin.

Founders raising into a Singapore entity often underestimate how much incorporation timing affects their first six months of cash flow. Getting the entity, the bank account, and the accounting set up correctly from day one means the model’s early assumptions hold up instead of needing a rebuild once the real invoices start arriving.

Ray Tay, Managing Director, VIVOS

This section is informational, not legal or tax advice. Confirm current rates and thresholds directly with IRAS or CPF, or with a qualified advisor, before finalizing the model.

How does equity dilution fit into the model?

Every funding round changes who owns what percentage of the company, and a model that ignores this gives founders a misleading view of their own long-term stake. Dilution modelling sits alongside the cash flow forecast, not inside it, but the two inform each other directly.

Each time new shares are issued, whether to investors, to an option pool, or to new co-founders, existing shareholders own a smaller percentage of the same company, even though the value of that smaller percentage may grow. A simple dilution table should show, for each planned round, the pre-money valuation, the amount raised, the resulting post-money valuation, and the ownership percentage of every shareholder class before and after.

Option pools deserve particular attention because they are often created or expanded at the same time as a priced round, and investors typically want the pool sized before the new money comes in, which dilutes existing shareholders more than it dilutes the incoming investor. Founders who model this in advance can negotiate pool size and timing instead of discovering the dilution after the term sheet is signed. Keep the dilution table on its own section of the cap table tab so it updates automatically as round assumptions change, rather than being recalculated by hand each time a scenario shifts.

How do fundraising rounds affect the cap table?

The cap table and the financial model should be linked, because the amount raised in each round directly determines the runway extension the model shows afterward. A fundraising rounds forecast lets founders test how much to raise, at what valuation, and how long that capital lasts before the next round is needed.

Build the cap table tab to list every shareholder, their share count, and their percentage ownership, updating automatically as each new round is added as a line item with its own raise amount and valuation. Link the raise amount from each planned round directly into the cash flow statement as a financing inflow in the month the round is expected to close, so the runway calculation reflects the actual timing rather than treating all future capital as already in the bank.

This matters most when testing how early or late to raise the next round. A model that shows runway dropping to three months before the next round is projected to close is a planning failure the founder can catch and fix, either by extending the current round, cutting burn, or moving the fundraising timeline forward. Keeping the cap table and the cash flow statement on separate tabs but linked by formula, rather than updated independently, is what prevents the two from drifting apart as rounds get renegotiated.

How do I validate the model against real data?

A model built entirely from assumptions, with no connection to actual performance, is a forecast in name only. Validating it against historical data and outside benchmarks is what turns a spreadsheet into a planning tool investors can trust.

Once the company has a few months of actual revenue and expense data, compare it line by line against the original forecast. A consistent gap in one direction, rather than random noise, usually means an assumption was wrong rather than the business underperforming, and the fix belongs in the assumptions tab, not in a one-off adjustment to a single month’s output. Early-stage companies with no revenue history yet can still sanity-check assumptions against public benchmarks for comparable business models, such as typical SaaS gross margins or typical conversion rates for a given channel, flagging any assumption that sits far outside the normal range for further justification.

Re-run the validation every time a new quarter of actuals becomes available, since a model validated once at the seed stage will drift as the business matures and its unit economics shift. Treat each validation pass as an opportunity to tighten the assumptions tab rather than as a one-time exercise done only before a fundraise.

What founders consistently get wrong about their own models

The most common mistake is not technical. Founders build an impressively detailed model and then stop checking it against reality, because once the fundraising deck is finished, the model feels like it has served its purpose. The opposite is true: the model earns its value after the round closes, when it becomes the tool for deciding whether to make the next hire or cut the next cost.

The second mistake is modelling headcount optimistically and burn conservatively, which is backwards. Hires almost always start later than planned, recruiting takes longer, and onboarding adds a ramp period before a new salesperson or engineer contributes fully. Modelling hire dates a month later than the plan, and modelling new hires at zero productivity for their first month or two, produces a far more honest burn number than assuming every hire starts exactly on schedule and is productive immediately.

A rolling 12-month forecast, updated monthly against actuals, has changed more than one founder’s hiring decision by showing a burn trend three months before the cash account itself would have shown a problem. That lead time is the entire point of building the model in the first place.

— Ray

How VIVOS supports founders building and feeding their model

A financial model is only as reliable as the inputs behind it, and several of those inputs depend on how and where the company is incorporated. An ACRA-registered filing agent handles Singapore company incorporation for foreign founders, including nominee resident director, registered address, and corporate secretary services, so the entity, banking, and compliance timeline that feed the model’s early assumptions are set up correctly from the start.

Services that map directly onto model inputs include consulting an App developer Singapore for SMEs for accurate project delivery timelines and cost inputs.

• Incorporation, which determines the timing of entity setup, bank account opening, and the first billable month.
• Accounting, which turns monthly actuals into the figures that reconcile against the forecast.
• Payroll management, which keeps headcount cost assumptions, including CPF contributions, accurate as hires are made.
• Tax planning and structuring, which keeps the effective tax rate and cash tax timing in the model aligned with current rules.
• Banking and finance support, which affects when financing inflows actually hit the account.

Similar services are offered in Malaysia, Hong Kong, and the UAE, with support in English and Mandarin, for founders structuring operations across more than one jurisdiction. Founders who prefer to build and maintain the model themselves can still use these services selectively, for incorporation or accounting alone, while keeping the forecasting work in house. For current service details and pricing, visit the VIVOS pricing page.

FAQ

What is a financial model for a startup?

A startup financial model is a linked three-statement projection, income statement, balance sheet, and cash flow, built from assumptions about revenue, hiring, burn, and runway. Investors typically expect one with monthly detail for at least the first two years as part of fundraising due diligence.

Can ChatGPT do financial modelling?

ChatGPT and similar tools can help draft formulas, structure tabs, or explain modelling concepts, but they cannot generate the assumptions that make a model credible, since those depend on the founder’s own knowledge of their market, pricing, and cost structure. Treat AI tools as drafting assistance, not as a substitute for building and defending the assumptions yourself.

What are the 7 stages of startup?

Definitions vary across sources, but a common version lists pre-seed, seed, early stage (Series A), growth stage (Series B and C), expansion, maturity, and exit. A financial model evolves at each stage, moving from rough assumptions at pre-seed to a fully actuals-reconciled forecast by the growth stage.

Will AI replace financial modelling?

AI tools are increasingly used to speed up formula building and scenario testing, but the core judgment behind a startup model, choosing which assumptions matter and defending them to investors, remains a founder-level task. The model’s value comes from the reasoning behind the numbers, which AI can support but not originate.

Sources

• Startup Financial Model: VC-Ready 3-Statement Projection
• How to build a VC-ready startup financial model
• IRAS | Corporate Income Tax Rate, Rebates & Tax Exemption Schemes

Recommended

• Company Valuation in Singapore: A Guide for Entrepreneurs
• Seed Funding in Singapore: A Practical 2026 Guide
• Angel Investors Singapore: Your 2026 Funding Guide
• Top Venture Capital Firms in Singapore: 2026 Guide

Related guides

• Avoid Shortfalls: 13-Week Cash Flow Forecasting for Singapore Founders
• Carta vs Pulley: Which Cap Table Platform Fits You?
• Convertible Notes in Singapore: Guide for Founders and Investors
• $900 at Stake per $10,000 Expense: Zero Rated vs Exempt in Singapore