There’s a phrase worth knowing from the start: all models are wrong, but some are useful. Your financial model will never perfectly predict the future – no startup’s can – and trying to make it do so is a trap that costs founders weeks they don’t have.

What a good financial model does is force you to think through your business plan with enough rigour that the numbers become believable. It’s a discipline tool as much as a forecasting one. Get it broadly right, then get on with delivering it.

Start with the Plan, Not the Spreadsheet

The most common mistake founders make is opening a spreadsheet before they know what they’re actually building. A financial model is a numerical representation of your plan – which means the plan has to come first.

What does customer acquisition look like? What are your unit economics? How long does it take to convert a lead? When do you need to hire, and into which roles? The answers to these questions don’t live in a spreadsheet – they live in your understanding of your business. The spreadsheet records them.

This is also why going through the modelling process is valuable even when the initial numbers don’t work. If the model shows you need to acquire customers at a cost that isn’t commercially viable, that’s the process doing its job. Go back, adjust the plan, and let the revised plan drive the next version of the model.

“A startup is a temporary organisation in search of a sustainable business model.” – Steve Blank

By the end of the process, you should be able to see – and show – that your business is sustainable, and what financial and other resources it will take to get there.

What UK Investors Expect to See

Whether you’re pitching to angel investors, a UK VC, or applying for a British Business Bank-backed loan, you need to know your numbers before you’re in the room.

Your model is supporting material, not presentation material – investors won’t necessarily ask for the spreadsheet during the meeting. But you should assume it will be requested either during or, more commonly, after your pitch. We’ve seen funding conversations fall apart because a founder took three weeks to clean up a model after an investor asked for it. By that point, the momentum was gone.

For UK early-stage raises, investors are typically looking at:

Revenue model. How does the business generate money? For SaaS businesses this means MRR, ARR and churn. For marketplaces or transactional models, GMV and take rate. Be explicit about how you’re modelling revenue – not just what the final number is.

Cost structure. What are the major cost drivers and how do they scale? Investors want to understand unit economics, not just the total figures.

Burn rate and runway. At your current rate of spend, how long can the business operate on existing cash? Most UK investors at seed stage expect to see at least 18 months of runway post-raise modelled clearly.

Headcount plan. When are you hiring and into which roles? Headcount is typically the largest cost driver for early-stage startups, and investors will scrutinise it closely.

If you’re raising under SEIS or EIS, there are additional requirements around how the raised funds are deployed – your model should reflect this, particularly the timeline for deploying capital into eligible trading activities. Investors relying on those reliefs will want to see that the spend plan is compliant.

The Fundamentals of a Good Startup Model

A financial model that an investor can actually use requires more than correct numbers. Here’s what separates a credible model from one that raises more questions than it answers.

Make your assumptions easy to find. Every model needs a dedicated assumptions tab where all the key inputs sit in one place. Never bury hard-coded numbers inside a formula. If an investor finds a hidden assumption, they’ll immediately wonder what else they haven’t found – and your credibility takes a serious hit.

Make your assumptions defensible. Every growth rate, conversion rate and cost assumption needs to be grounded in something real. If you’re pre-revenue, use industry benchmarks as a starting point. If you have trading history, use it. The question investors are asking is not “is this right?” – it’s “is this credible?”

Build from the bottom up. Rather than assuming a percentage of revenue for a cost category, itemise the actual spend. Instead of “marketing: £50k,” show your channel breakdown – paid search, social, events – with a cost and expected return for each. This level of detail signals that you’ve actually thought the plan through rather than applied a rule of thumb.

Make it easy to read. A spreadsheet that only you can navigate is a liability. Colour code consistently – for example, blue for enterable assumptions and black for derived formulas – so an investor can immediately see what’s an input and what’s calculated. Use charts to illustrate key trends, particularly when you can layer actual data in alongside the forecast.

Reflect how your business actually works. If your revenue comes from a mix of subscription and professional services, show each stream separately. If customer acquisition cost varies by channel, break it out by channel. The model should read like your business, not like a generic template someone has filled in.

How Much Time Should You Spend on It?

Enough to feel like you’ve modelled the most likely outcome for your business. Then stop.

Analysis paralysis kills more financial models than bad assumptions do. An honest, directionally correct model delivered promptly is worth considerably more than a highly detailed one that arrives three months later, after the investor’s attention has moved on.

Build it, sense-check it, get feedback from someone who has seen enough startup models to know whether it holds together – then get on with executing the plan underneath it.