How to Build a Seed Stage Financial Model for a UK SaaS Raise

A seed financial model is not a Series A model in miniature. It is something different - a credible story about how a relatively small amount of capital (typically £500K to £2M in the UK) buys you enough runway and product progress to make a Series A investable.

UK seed investors - Octopus Ventures, Notion Capital, LocalGlobe, Episode 1, Seedcamp, Hambro Perks, and the active angel network - typically write cheques of £100K to £1M each and lead rounds of £500K to £2M. They are betting on the founding team and the market opportunity, with the financial model serving as proof you have thought rigorously about how the money will be spent and what it will produce.

At seed, the financial model has three jobs:

1. Show how much capital you are raising and what milestones it funds
2. Demonstrate that you understand the cost of acquiring customers and serving them
3. Prove there is a credible path from where you are now to a Series A in 18-24 months

The model does not need 60 months of granular projections. It needs to be honest about what you know, transparent about what you are assuming, and operationally connected to the team you plan to hire.

What seed investors actually look for in the model

UK seed VCs evaluate the model differently than Series A investors. They are not yet expecting cohort-level retention data, mature unit economics, or a quality-of-earnings-ready set of accounts. What they want is:

• A defensible capital ask. Why £500K and not £300K or £1.5M? What specifically does each tranche of capital fund?
• A credible 18-month plan. What does the company look like at month 18? What ARR threshold makes you Series A-ready? UK Series A typically requires £80K-£100K MRR (~£1M ARR), so seed funding should plausibly get you there.
• Sensitivity to the things that matter at seed. Burn rate, runway, churn assumptions, sales pipeline conversion. Less so DCF, terminal value, EV/Revenue multiples.
• An honest founder. Investors are watching how you handle the assumptions you cannot back with data. A model that hides uncertainty behind hardcoded inputs is a red flag; a model that explicitly names assumptions and stress-tests them is a confidence builder.

Before you build the model - what data you have at seed

At pre-seed and early seed, you may have very limited historical data. That is fine. The model still needs to be grounded in what you do know.

Pull together:

• What you have already sold. If you have signed customers, every contract: ACV, billing frequency, payment terms, churn status. Even 3-5 customers tells investors a lot.
• What is in the pipeline. Real, named opportunities. Investors will probe these in diligence: who, what stage, expected close date, what is blocking the deal.
• Current burn rate. Last 3-6 months of P&L showing actual costs by department. If you have no formal accounting, even bank statements are enough to back into a burn figure.
• Headcount. Who is on the team today, in what role, at what cost. Including the founders if drawing a salary.

If you have less than 6 months of trading data, build the model around pipeline + the unit economics of the few customers you have closed. Do not pretend to have cohort data you do not have. UK investors at seed will respect honesty about data limitations.

Define the question the model is answering

A seed model has fewer questions to answer than a Series A model. The most important ones:

• How much capital do you need to reach a credible Series A inflection point?
• What does that inflection point look like - in MRR, customers, gross margin, or product progress?
• What is your monthly burn at the highest-spend point of the plan, and how does that compare to the cash you will hold?
• If revenue comes in 30% lower than plan, how long does the cash last?

The answers to these determine the model's structure. A 24-month model with monthly granularity is sufficient at seed. Anything beyond 24 months is speculative and the precision implies false confidence.

Build revenue from the pipeline up

For seed, "bottom-up" means starting from what is in your sales pipeline rather than from your closed customers (you may not have enough). Build the revenue model in three layers:

Existing MRR. Whatever you have today, with realistic assumptions about retention. At seed, single-month churn assumptions are sketchy without 12+ months of data - use a quarterly view if cohort data is thin.

Pipeline conversion. List every real opportunity, its expected ACV, and your honest probability of close. Apply a conservative win rate (20-30% for B2B SaaS at seed) and stagger close dates realistically. A deal in "Discovery" today does not close in 30 days for B2B SaaS - typical seed-stage cycles are 60-120 days.

New pipeline generation. Once you have hired the planned sales / marketing team, how much pipeline does that team generate per month? Tie it to specific outbound activity (calls, demos, conversion to opportunity) or content / inbound assumptions. Do not assume the pipeline magically doubles after the round.

The result is a month-by-month MRR build with three drivers: starting MRR, identified pipeline conversion, and post-raise generated pipeline conversion. Each driver has its own probability assumption that you can stress-test.

For UK seed founders, see unit economics for UK SaaS for how to think about CAC, payback, and LTV when the data is thin.

Cost structure - simpler than Series A

Cost build at seed is simpler than at Series A because the team is smaller and the headcount plan is more concrete.

Build a named hiring plan: who is on the team today, who you plan to hire month-by-month, at what salary, in what department. For UK hires, use fully loaded cost (base salary + 13.8% employer NIC + 3-5% pension). A £60K hire costs the business approximately £70K-£72K all-in.

Beyond headcount, the main cost lines:

• Hosting and infrastructure. AWS / GCP / Azure spend, plus any third-party APIs your product depends on. At seed, this is typically modest (£500-£3,000/month) unless your product is compute-heavy.
• Sales and marketing tools. CRM, email, analytics, possibly some early paid acquisition tests. Keep this realistic for a £500K-£2M raise - paid acquisition at this stage is usually about learning unit economics, not buying revenue.
• Professional services. Accountancy, legal (especially for the SEIS/EIS round itself), R&D tax credit prep. Budget £20K-£40K in the year of the raise.
• Office and operations. Co-working, software subscriptions, business insurance. Typically under £2K/month per person for a fully remote / hybrid team.

Two scenarios, not three

Series A models need bear / base / bull. At seed, two scenarios are usually sufficient:

Base case. Your best honest estimate. This is what you present.

Downside. Pipeline conversion is 30% lower than base. Hires are delayed 1-2 months. Churn (or contract cancellations) is 50% higher. What happens to runway? When do you need to raise again? What is the minimum viable outcome at the end of the funding window?

The downside scenario at seed is more important than the bull case. Seed investors are partly underwriting the risk that things go wrong - they want to see you have thought about it.

The three financial statements - integrated but simpler

A seed model still needs three integrated financial statements. The complexity bar is lower than Series A but the integration must still hold.

Profit and loss. Monthly for years one and two. Revenue, COGS (hosting and payment processing typically), gross profit, S&M, R&D, G&A, EBITDA, net profit.

Balance sheet. Cash, trade debtors, deferred income (if you bill annually upfront), trade creditors, share capital, retained earnings. Must balance every period.

Cash flow. Operating cash flow, investing cash flow (capitalised dev costs, equipment), financing cash flow (the seed round itself, any SEIS / EIS tranches, any loan drawdowns). Closing cash equals balance sheet cash every period.

If you bill annually upfront on a £12K ACV, you receive £12K of cash on day one but only £1K of recognised revenue per month. The other £11K sits as deferred income on the balance sheet. UK seed investors will check that this is reflected correctly. Getting it wrong is one of the most common errors in seed financial models.

For the full mechanics, see our guide on building a financial model for UK SaaS startups.

Unit economics - emerging, not mature

At seed, the unit economics are still being discovered. The model should not claim a precise LTV / CAC ratio - the data simply does not exist. What it can show:

• Customer Acquisition Cost (CAC) estimate. Based on your pipeline conversion logic - the implied cost of acquiring each closed customer given your sales and marketing spend.
• Gross margin per customer. Revenue minus the direct cost of serving that customer (hosting, support, payment fees).
• Payback period sketch. Months to recover the estimated CAC at current gross margin. This will be rough at seed but signals you have thought about it.
• Sensitivity to churn. If your assumed monthly churn doubles, what happens to ARR by month 18? This matters more than precise LTV at seed.

UK seed investors do not punish you for not knowing your true unit economics. They punish you for pretending to know them when you do not.

UK-specific factors at seed

UK seed rounds have specific tax and regulatory features that the model must reflect.

SEIS and EIS. Most UK seed rounds use SEIS (£250K company limit, requires gross assets under £350K and fewer than 25 employees) and EIS (up to £12M per company lifetime, up to £5M per year). These give investors 50% income tax relief on SEIS and 30% on EIS - which is often what makes the deal happen. The cap table and use-of-proceeds section should show clean SEIS / EIS allocations.

HMRC Advance Assurance. Most UK seed deals require Advance Assurance from HMRC before completion. The pre- and post-money cap table should show no investors who breach EIS rules (for example, no single pre-money holder above 30%).

R&D tax credits. Many UK SaaS seed companies qualify for HMRC R&D relief under the merged scheme. The credit is typically received 6-12 months after the financial year-end. The model must reflect this timing - many founders book the credit as Day 1 cash which overstates near-term runway.

Employer NIC. UK payroll costs include 13.8% employer NIC above the secondary threshold (£5,000 from April 2025). Models that use gross salary only understate true cost by 15-20%.

VAT. UK B2B SaaS charges 20% VAT on UK customers. EU customers may be reverse-charge. The P&L is net of VAT, but the cash flow needs to account for VAT collection and payment timing.

What goes alongside the model in the data room

At seed, the data room is lighter than at Series A but still expected:

• 6-12 months of monthly management accounts (or bank statements if no formal accounts yet)
• Customer list with ACV and billing terms
• Current sales pipeline with named opportunities
• Cap table - current and pro-forma after the round
• SEIS / EIS Advance Assurance letter (or evidence it is in progress)
• Founder agreements and vesting schedules
• Top 5-10 customer contracts (where signed)
• Any IP / patent filings if applicable

The financial model is the centrepiece - everything else either supports or reconciles to it.

Timeline - how long to build a seed model

A seed financial model takes 1-2 weeks of focused work to build from scratch. The timeline depends on data quality:

• Week 1: Audit historical data (or assemble what exists), pipeline review, define the question the model is answering
• Week 2: Revenue build, cost / headcount plan, three-statement integration, downside scenario, narrative alignment with the pitch deck

If you have less than 6 months of trading data, the "audit" step is shorter but the pipeline conversion logic takes longer because you are building it from first principles.

The model should be a living document. Each month post-build, update with actuals, compare against plan, and revise assumptions as the data improves. By the time you are pitching investors, you should have 3-6 months of actuals against the model - that demonstrates operating rigour as much as the model itself.

For founders preparing for Series A in 18-24 months, the seed model becomes the foundation for the Series A financial model. The data structure, MRR schedule, and cost categorisation you set up at seed are the same ones investors will scrutinise at Series A - getting it right now saves significant work later.

If your seed model needs building from scratch and the underlying data is unreliable, that is the immediate priority. The fractional CFO services at ScaleWithCFO are built around this exact problem - cleaning data, building the model, structuring the cap table for SEIS / EIS, and getting the financial section of the seed pitch ready for investor scrutiny.