How DeFiGure works
Research first, advice second. The method behind our financial models, valuations and advisory work, step by step, with the sources we use and the limits we state.
Built for founders, leadership teams and SMEs across Europe and English-speaking markets
DeFiGure's corporate finance advisory method is research first, advice second. We start from public signals, then your own numbers, compare segments and competitors, show every source, state the limits and rank findings by impact. The result is a financial model, valuation or recommendation built for one company, with a summary a leader can present.
This page sets out each part of that method: how the research works, how we build a financial model, how we approach a valuation, what data we use and do not use, and how an engagement runs from first call to handover.
Research first, advice second
Before DeFiGure recommends anything, we write down what the evidence says. The same six steps apply whether the question is a funding round, a price change or a budget, and whatever the company sells.
What is intelligent research, and how is it different from a generic report?
Intelligent research is research built around one company's decision: its own numbers, its segments and the competitors its customers actually compare it with. A generic report describes a market on average; intelligent research asks what that market means for this company, and ranks what to do about it.
The difference shows in what a leader gets back. A generic report ends with trends. DeFiGure's research ends with findings ranked by impact, each with its source, its limits and a first step. For example, instead of "the market is growing", the finding reads "two of your three closest competitors raised prices in the last year; here is what a similar change would do to your margin and cash".
What does "research first" mean in practice?
- Public signals first. We start with what is already public about the company, its market and its competitors: registry filings and published accounts, published pricing, job postings, product pages, customer reviews and market data from named publishers. The first conversation then starts from evidence, not from a blank page.
- Segments and competitors compared. We split the market into segments, by customer type, size, channel or region, and compare the company with the alternatives its customers actually weigh it against, on pricing, positioning and visible signs of growth.
- Sources shown. Every figure carries its source and the date we read it, so anyone can check it. Where a number is our own estimate, it is labelled as an estimate and the arithmetic is shown.
- Limits stated. We say what the research could not see, such as private competitor figures, small samples or old data, and how much weight each finding can carry.
- Findings ranked by impact. Findings are ordered by how much each could move cash, revenue, valuation or risk, so the first item on the list is the one worth acting on first.
- A summary a leader can present. The work ends with a short summary a founder or leadership team can take to a board, an investor or their own team, with the full evidence behind it.
Only then comes the advice. Your own numbers, such as accounts, bank data, pipeline and customer data, are brought in to test the public picture, and each recommendation points back to the finding it rests on. If the evidence does not support a recommendation, we do not make it.
How does DeFiGure build a financial model?
What it is. A financial model is a spreadsheet that turns a company's assumptions (prices, volumes, hiring, costs, payment terms) into a forecast of revenue, profit and cash.
Why it matters. A model is a tool for decisions: how much to raise, when to hire, and what happens to runway (the number of months a company can operate before its cash runs out) if sales slow down. Investors read it to test how the founders think, not to check whether the forecast comes true.
Example. A company plans to hire two salespeople. The model takes their start dates, salaries, months to full productivity and expected deals per quarter. It then shows when the hires pay for themselves, how much cash is used before they do, and how runway changes if every deal closes one quarter later.
The steps we follow
- Pin down the decision. We agree what the model must answer and who will read it: a funding round, a hiring plan, a pricing change or an annual budget. That sets the time horizon and the level of detail, typically monthly for the first 12 to 24 months and quarterly after that.
- Start from actuals. Historical accounts, bank data, billing and pipeline come first, so the forecast continues from what has really happened. A company without revenue yet starts from its cost base and hiring plan.
- Build revenue bottom-up. Revenue comes from drivers the company controls and can measure, such as leads, conversion rates, prices, order volumes and churn (the share of customers who leave), not from a percentage of a market. This is what investors look for in a startup financial model.
- Tie costs to the plan. Headcount is planned by role and start date, and variable costs scale with volume, so no cost line stays flat when the plan changes.
- Connect the three statements. The profit and loss statement (P&L), balance sheet and cash flow statement are linked, so profit flows into equity and cash, and working capital (money tied up in receivables and inventory, minus what is owed to suppliers) shows when cash actually moves. Our guide explains how the P&L, balance sheet and cash flow connect.
- Test the drivers. Sensitivity analysis changes one driver at a time to find the few that move the result most. Scenario analysis then builds base, upside and downside cases around those drivers. See scenario and sensitivity analysis: base, best and worst cases you can defend.
- Check and document. Inputs sit apart from formulas, no numbers are typed into calculations, each assumption carries a source or a reason, and a check row confirms that the balance sheet balances.
- Walk through and hand over. We go through the model with the people who will use it, so they can add actual results each month and update it without us.
This is the core of our Forecasting & Modeling work. Terms such as runway, churn and working capital are explained in plain words in the DeFiGure corporate finance glossary.
How does DeFiGure approach a valuation?
What it is. A valuation is an estimate of what a company is worth on a given date, built from methods that investors and acquirers recognise, and best stated as a range rather than a single number.
Why it matters. Every method rests on different assumptions, so no method is right on its own. We use more than one, look at where they agree and where they do not, and write the reasoning down, so the founder can defend the number in a negotiation.
Example. A company raising its first round has little revenue history. We compare it with recent rounds for companies at the same stage in the same region, check what it would need to be worth at exit for an investor to reach their return target, and test both against the plan in the financial model. If one view gives a very different answer, we examine its assumptions rather than average it away.
The steps we follow
- Fix the purpose and the date. A valuation for a funding round, a share sale, an employee option plan or the sale of the business answers different questions, and every valuation holds only for the date it was made.
- Choose methods that fit the stage. Before revenue, investors rely on structured judgement, such as the Scorecard, Berkus and VC methods explained in how to value a pre-revenue startup. Once there is revenue, multiples from comparable companies and transactions come in; our guide to SaaS valuation multiples in 2026 shows how growth and customer retention move them. With stable cash flows, a discounted cash flow (the value today of the cash the company is expected to generate) adds a further view.
- Build on the model. Multiples and discounted cash flows both need a credible forecast, so the valuation uses the same financial model as the rest of the work, not a separate set of numbers.
- Triangulate and test. We compare the result of each method and test which assumptions move the range most.
- Translate the range into deal terms. Pre-money valuation is what the company is worth before new money comes in; post-money valuation is the pre-money figure plus the new money. An investor's stake equals their investment divided by the post-money valuation: €1 million invested at a €4 million pre-money valuation gives a €5 million post-money valuation and a 20% stake. Our guide to pre-money vs. post-money valuation also covers option pools and convertible instruments.
- Write down the range and the reasoning. The output states the range, how much weight each method carries, the dated source behind every comparable and what would move the number.
Valuation work sits within our Capital Raising & Valuation service. A DeFiGure valuation is an advisory estimate to support a negotiation or a decision. It is not a formal valuation opinion for tax, accounting or court purposes, which may require a qualified valuer.
What data we use, and what we do not claim
A finding is only as good as its source. This is what goes into DeFiGure's work, what stays out, how findings are ranked and where the limits are.
What data do we use?
- Public sources: registry filings and published accounts, published pricing and product information, job postings, customer reviews, and market, funding and transaction data from named publishers, each recorded with its date.
- Your own data, shared in confidence: accounts, bank exports, billing and order data, pipeline, plans and the cap table (the record of who owns which shares).
- Your team's knowledge: what founders and managers know about customers and operations, written down and labelled as assumptions to test, not as facts.
What do we not use?
- Figures we cannot trace to a source or to arithmetic we can show.
- Undated market figures, or old figures presented as current.
- Confidential information about other companies, including one client's numbers in another client's work.
- Personal data that the question in hand does not need.
How are findings ranked?
Each finding is weighed on three questions, in this order: how much could it move cash, revenue, valuation or risk; how confident are we, given the quality and age of its source; and how hard is it to act on. High impact comes first. Between findings of similar impact, the better evidenced one comes first. Each finding comes with a first step, not just a general recommendation.
What do we not claim?
- No guaranteed outcomes. A model is not a promise that the forecast will happen, a valuation is not a promise that an investor will pay it, and fundraising support is not a promise that a round will close.
- No investment advice. We advise companies on their own finances. We do not make recommendations on securities or other investments, and nothing on this site is investment, legal or tax advice. Legal and tax questions belong with your lawyer or tax adviser.
- Figures are dated. Market figures such as valuation multiples, interest rates and funding medians are true as of the date shown and change over time, so we check them again before they go into a decision.
- No invented benchmarks. Where no reliable benchmark exists, we say so instead of filling the gap with a guess.
How does an engagement run?
Four steps for every service. The scope is agreed in writing before any work starts. For CFO advisory and risk monitoring, step 4 can continue as ongoing monthly support.
Free 30-minute call
We talk through your stage, your numbers and the decision in front of you. No obligation, and no preparation needed beyond what you already have.
Fixed-scope proposal within 48 hours
A written proposal: the question we will answer, the work involved, what you will receive and the timeline. You know the scope before work starts.
Build
Research first, then the model, valuation or recommendation. You review the assumptions before the conclusions are drawn.
Walkthrough and handover
We walk your team through the logic and hand over the working files, so you can run, update and defend the work yourselves.
Questions about our method
Should I build my financial model top-down or bottom-up?
Bottom-up. A top-down model that claims a small share of a large market contains no operating logic, so investors discount it. DeFiGure builds revenue from drivers the company controls and can measure, such as leads, conversion rates, prices, order volumes and churn, so every output traces back to an assumption you can defend. A top-down view of the market is still useful, but only as a sense check.
How many years should a startup financial model project?
Three years is the working standard for early-stage fundraising, built monthly for the first 12 to 24 months and quarterly after that. The first year should be defensible line by line, because that is where actual pipeline and customer data carry real signal. Beyond that, DeFiGure keeps the logic driver-based and treats the later years as scenarios rather than predictions.
Which valuation method do angel investors actually use?
Most angels triangulate rather than rely on one method. The Scorecard method compares a company with recent rounds at the same stage in the same region, the Berkus method acts as a sanity check, and the VC method confirms the deal can still return the investor's target at exit. DeFiGure builds the same three views with founders, so they know which number they are defending and why.
How often should I update my financial model?
Monthly. Add the month's actual results, compare them with the plan and change the forward assumptions that the difference has proved wrong. A model that still shows January's assumptions in June tells investors it was built for the raise, not to run the company. DeFiGure hands models over with inputs kept apart from formulas, so your team can make this update without us.
Have a decision in front of you?
DeFiGure works with founders, leadership teams and SMEs across Europe and English-speaking markets. Tell us what you are deciding and we will tell you whether we can help. Still weighing your options? Read our fair comparison of advisory vs. templates.