How to Build a Financial Model for Your UK SaaS Startup

A financial model is not a spreadsheet exercise. It is the document that tells investors: here is how we make money, here is where we spend it, here is when we become profitable, and here is how much capital we need to get there.

Every UK SaaS fundraise -- whether it is a £500K angel round (often SEIS-eligible), a £1-3M Seed, or a £5-10M Series A -- requires a financial model. The quality of that model directly affects how much you raise, at what valuation, and how quickly. A weak model full of unsupported assumptions and hockey-stick projections will get you a polite rejection from UK VCs. A rigorous model grounded in your actual data, with proper UK line items (PAYE, Employer NIC, VAT, Corporation Tax, R&D credit) and defensible assumptions, will get you funded.

Here is the process I follow when building financial models for SaaS companies as a fractional CFO. It typically takes two to four weeks from start to finish.

Step 1: Audit your historical data

Before building anything forward-looking, you need to understand what has actually happened. This means reviewing 12 to 24 months of:

• Profit and loss statements -- month by month, with costs properly classified into COGS, sales and marketing, R&D, and G&A.
• Balance sheet -- particularly cash, debtors, creditors, and deferred income.
• MRR schedule -- customer-level recurring revenue with movement categories (new, expansion, contraction, churn).
• Headcount -- who was hired when, at what cost, in which department.

The historical data sets the baseline for your forecast. If your last 12 months show 25% year-on-year revenue growth, your model needs to explain convincingly why growth will accelerate to 60%. Without that explanation, investors will not believe the numbers.

This step also catches data quality issues. Misclassified costs, missing accruals, or an MRR schedule that does not reconcile to the P&L all need fixing before the model can be built. See our guide on financial data accuracy for what to check.

Step 2: Define the model's purpose and audience

The model serves different audiences depending on what you need:

Fundraising from angels: Angels typically expect a 10x return in five to seven years. Your model needs to show a credible path to that outcome -- usually through revenue growth leading to a trade sale or Series A at a higher valuation.

Fundraising from VCs: VCs need to see the potential for 20x to 50x returns. The model must demonstrate a large addressable market, efficient unit economics, and a path to £10M+ ARR within three to four years.

Bank lending: Banks care about cash flow and debt service coverage. The model must show positive operating cash flow and the ability to service loan repayments. HMRC's Enterprise Finance Guarantee scheme supports loans up to £1.2M for eligible businesses.

Internal planning: For operational use, the model needs more granularity -- monthly detail, departmental budgets, hiring triggers, and cash runway calculations.

The audience determines the level of detail, the time horizon (three years for angels, five years for VCs), and the key metrics to emphasise.

Step 3: Build revenue from the bottom up

The most common modelling mistake is building revenue top-down. "The UK HR tech market is £2 billion, we will capture 1% within three years, giving us £20M ARR." Investors dismiss this immediately because it is not grounded in anything concrete.

Instead, build revenue bottom-up from your actual sales motion:

Start with your MRR schedule. Your current customers, their contract values, and their renewal dates form the baseline.

Model new customer acquisition. How many leads do you generate per month? What is your conversion rate? What is the average deal size? These inputs -- grounded in your historical data -- drive new MRR each month.

Apply churn and expansion. Use your actual gross churn rate and expansion rate. If you do not have 12 months of data, use conservative estimates based on B2B SaaS benchmarks (2-3% monthly logo churn for SMB, 0.5-1% for enterprise).

Layer in pricing changes. If you plan to increase prices, model the impact on new customers (higher ACV) and existing customers (at renewal).

The result is a month-by-month MRR forecast built from specific, defensible assumptions. Each assumption can be tested and challenged, which is exactly what investors will do. For a deeper look at the template structure, see our SaaS financial model template guide.

Step 4: Build the cost base

Costs fall into four categories, and the model must handle each differently:

Headcount (typically 60-75% of total costs). Build a detailed hiring plan that links to revenue milestones. For example: hire one customer success manager for every 15 new customers, or one additional engineer when ARR reaches £500K. Each role needs a start date, salary (including employer NIC at 13.8% and pension at 3-5%), and department allocation. This is the most important cost input because it drives the majority of your burn rate.

Cost of goods sold. Hosting and infrastructure (typically 5-15% of revenue for SaaS), third-party software costs embedded in your product, payment processing fees, and customer support costs if you classify support as COGS.

Sales and marketing. Advertising spend, tools, events, commissions. These should tie directly to your revenue assumptions -- if you are modelling 50% growth, the S&M spend must be sufficient to generate the leads required.

General and administrative. Office costs, legal, accounting, insurance, software tools. These typically grow slowly and are the easiest to forecast.

Step 5: Build three scenarios

A single-scenario model is incomplete. Investors want to see how sensitive the business is to different assumptions. Build three scenarios:

Bear case (downside). Revenue grows 30% slower than base. Churn is 20% higher. Key hires are delayed by two months. This scenario answers: what happens if things go worse than planned? How much runway do we have? When do we need to cut costs?

Base case (expected). Your best estimate based on current data and achievable targets. This is the scenario you present as your plan.

Bull case (upside). Revenue grows 30% faster than base. A major partnership or product launch accelerates growth. This scenario answers: what is the return potential if things go well?

Each scenario should change the key drivers (new customer acquisition rate, churn rate, hiring pace) rather than simply scaling revenue up or down. The scenarios must be internally consistent -- if you model 50% more customers in the bull case, you also need more support staff and higher hosting costs.

Step 6: Generate the three financial statements and unit economics

The model must produce a complete set of outputs:

Profit and loss (monthly for years one and two, quarterly or annual for years three to five). Revenue, COGS, gross profit, operating expenses by category, EBITDA, and net profit. Gross margin should trend toward 70-85% for a well-run SaaS company.

Balance sheet. Cash, trade debtors, prepayments, trade creditors, deferred income, loans, equity. The balance sheet must balance (assets = liabilities + equity) in every month. Cash is the balancing item -- it falls out of the P&L and working capital movements.

Cash flow statement. Operating cash flow (from the P&L adjusted for working capital movements), investing cash flow (capex, capitalised development), and financing cash flow (fundraising, loan drawdowns and repayments). The closing cash must equal the balance sheet cash in every period.

Unit economics. CAC, CAC Payback, LTV, LTV:CAC ratio, Magic Number. These must be derived from the model's outputs, not hardcoded.

Runway. How many months of cash remain at the current burn rate? At what date does the company run out of cash without additional funding? This is the number that determines how urgently you need to raise.

Common modelling mistakes

Hockey-stick revenue without justification. If historical growth is 25% and the model shows 100% next year, you need a specific, credible reason. "We will hire more salespeople" is not sufficient -- model the actual pipeline conversion from those hires.

Ignoring cash timing. Revenue recognised is not cash received. If you invoice annually upfront, you receive cash in month one but recognise revenue over 12 months. If you invoice monthly, cash arrives monthly. If customers pay on 30-day terms, cash arrives a month late. The model must capture these timing differences, or your cash runway will be wrong.

Flat operating costs. Costs do not stay flat while revenue triples. Infrastructure costs scale with usage. Support costs scale with customer count. Even G&A increases as you add legal complexity, insurance requirements, and regulatory obligations.

No sensitivity analysis. If a 10% increase in churn turns your 24-month runway into 14 months, investors need to know that. Build sensitivity tables for the key drivers.

Overcomplicating the model. A 50-tab spreadsheet with circular references and hidden macros is not impressive -- it is a liability. The model should be clean, auditable, and understandable by anyone with financial literacy. Tabs should flow logically: assumptions, revenue, costs, P&L, BS, CF, unit economics, scenarios.

How long it takes

Building a fundraise-ready financial model from scratch typically takes two to four weeks with a fractional CFO. The first week is data gathering and historical analysis. The second week is building the revenue and cost models. The third week is generating statements, running scenarios, and stress-testing assumptions. The fourth week is refining presentation and preparing the narrative that accompanies the model.

If your historical data is clean and your MRR schedule is up to date, it can be done in two weeks. If the data needs significant clean-up first, budget four weeks.

The model is not a one-time deliverable. It should be updated monthly with actual results, so you can track variance against plan and adjust forecasts accordingly. A good model becomes your primary management tool, not just a fundraising artefact.