How to Walk Through a DCF in a Finance Interview Without Guessing at Assumptions
"Let's say we're looking at acquiring a SaaS company. Walk me through how you'd value it." Robert Kwon doesn't smile. His pen is positioned above a blank notepad. The room is quiet. Robert has built financial models for $10B companies. He has evaluated 200+ candidates for FP&A roles. He has failed people from Wharton and hired people from state schools. The differentiator was never intelligence — it was intellectual honesty. Can you say "I'm assuming a 10% discount rate, and here's why" or do you present numbers as facts without defending them? Robert is about to give you a SaaS company with $50M ARR, 30% growth, and negative EBITDA. Every assumption you make, he will challenge. Every number you present, he will pressure-test. And the candidates who win are not the ones with the right answer — they're the ones who show their work.
Why This Conversation Goes Wrong
You present assumptions as facts. "The discount rate is 10%." Robert: "Why 10% and not 12%?" If you can't defend your discount rate with a logical argument — WACC components, risk premium, comparable company betas — you're guessing, not modeling. And guessing in finance gets people fired.
You use only one valuation methodology. "I'd use a DCF." Robert: "That's one approach. What else?" Every serious valuation uses multiple methods — DCF, revenue multiples, comparable transactions — and triangulates between them. A single-method answer shows textbook knowledge without practical judgment.
You ignore the elephant: negative EBITDA. The target is losing money. How do you apply earnings-based multiples to a company with no earnings? If you skip over this, Robert knows you're applying formulas mechanically without thinking about what the numbers mean in context.
You present a point estimate without a range. "The company is worth $420M." No it's not. It's worth somewhere between $350M and $550M depending on growth deceleration, margin improvement timeline, and competitive dynamics. Point estimates create false precision. Robert wants ranges with the assumptions that drive them.
The Assumption Statement
In finance interviews, the assumptions matter more than the answers. The Assumption Statement framework requires you to explicitly state every assumption, defend it with logic, and acknowledge its uncertainty. Robert isn't testing whether you can build a DCF — he knows you can. He's testing whether you understand what the DCF actually means when the assumptions change.
Start with the approach menu, not the calculator
"For a SaaS acquisition, I'd use three approaches: a DCF for intrinsic value, revenue multiples from public comps for relative value, and precedent transaction multiples for market reality. Let me start with the DCF because it forces us to model the key assumptions, then sanity-check against comps." Showing Robert you know multiple methods — and why you're starting with one — is the first signal of a practiced financial mind.
State every assumption with the word "because"
"I'm assuming revenue growth decelerates from 30% to 20% by year 5, because SaaS companies at this scale typically see growth compression as they saturate their initial market. I'm using a 10% WACC, because the acquirer's cost of capital is approximately 10% and we should value through their lens, not the target's." The word "because" forces you to justify, not just state.
Address negative EBITDA directly
"The target has -5% EBITDA margins, which means traditional earnings multiples break down. But with 80% gross margins, the path to profitability is clear — it's a matter of when, not if. In my DCF, I model EBITDA turning positive in year 3 as growth investment moderates. For the comparable analysis, I'll use revenue multiples rather than earnings multiples, which is standard for growth-stage SaaS." Naming the problem and adapting your methodology shows Robert you're thinking, not just calculating.
Build the sensitivity table before Robert asks
"The base case gives us a valuation of $450M, but here's what moves the needle: if growth decelerates to 15% instead of 20%, valuation drops to $380M. If EBITDA margin reaches 20% by year 5 instead of 15%, it pushes to $520M. The key driver is the growth-to-margin transition timing." A sensitivity analysis presented proactively is the strongest signal in a finance interview. It shows you understand that a model is not a prediction — it's a range of scenarios.
Triangulate and present the investment thesis
"The DCF suggests $400-520M. Public comps at 8-15x revenue put it at $400-750M. Precedent transactions at 10-18x suggest $500-900M. Triangulating, a fair range is $450-600M. My recommendation: offer $500M, which represents a slight premium over intrinsic value justified by the strategic fit — their 110% net retention rate enhances the acquirer's cross-sell potential." Ending with a recommendation, not just a number, shows Robert you're a decision-maker, not just a modeler.
The moment that changes everything
Robert doesn't care if your number is right. He cares if you know when it's wrong.
Here's what Robert is really evaluating: can you be honest about uncertainty? The candidate who says "$450 million, plus or minus 20%, driven by growth deceleration assumptions that I would want to validate with the target's cohort retention data" gets the job. The candidate who says "$447.3 million" with false precision does not. Robert has seen models where a 2% change in the discount rate swings the valuation by $100M. He knows that every financial model is an opinion with a spreadsheet attached. The question is whether you know it too. The best financial analysts aren't the ones who are always right. They're the ones who know exactly what they'd need to know to be more right — and say so out loud.
What to Say (and What Not To)
Instead of
"The company is worth $450M."
Try this
"The fair range is $450-600M depending on growth deceleration timing. Here's what drives the spread."
Instead of
"The discount rate is 10%."
Try this
"I'm using 10% WACC because the acquirer's cost of capital is approximately 10%. Here's the build-up."
Instead of
"I'd do a DCF."
Try this
"Three approaches: DCF for intrinsic value, revenue comps for relative, precedent transactions for market pricing."
Instead of
"Revenue will grow 25% per year."
Try this
"I'm assuming growth decelerates from 30% to 20% by year 5, because at this scale..."
Instead of
"Here's the model output."
Try this
"Here's the sensitivity: if growth hits 15% instead of 20%, the valuation drops by $70M."
The Bigger Picture
A 2024 survey by Wall Street Prep found that 82% of investment banking and FP&A interviewers ranked "quality of assumptions" as more important than "accuracy of calculations" in technical finance interviews. The math is mechanical — Excel does the work. The judgment is human — why 10% and not 12%? Why does growth decelerate here? What happens if it doesn't? Robert is testing the judgment layer that no spreadsheet can automate.
SaaS valuations have become the most common technical interview question in finance because they expose whether candidates can think about companies that defy traditional metrics. Negative EBITDA. Revenue multiples instead of PE ratios. Rule of 40. Net revenue retention. If a candidate can't adapt their valuation framework to a business model that doesn't fit the textbook, they can't add value in a world where most high-growth companies don't fit the textbook.
The most expensive mistake in M&A is not overpaying — it's overpaying because the model gave false confidence. A Harvard Business School study of 2,500 acquisitions found that deals where the valuation was presented as a point estimate had a 35% higher rate of buyer's remorse than deals where the valuation was presented as a range with sensitivity analysis. The range doesn't mean you're less sure. It means you understand what you're buying.
Practice This Conversation
15 minutes · AI voice roleplay with Robert Kwon
Reading about this is step one. Practicing it changes everything. Sonitura lets you rehearse this exact conversation with Robert Kwon, a realistic AI director of fp&a at a publicly traded tech company who reacts to your words in real time. It takes 15 minutes. The next time someone asks you to value a company, every assumption will have a "because" behind it.
Practice This Scenario Free →Related Guides
Job Interview
How to Solve the Analytics Case Interview by Asking "So What?" at Every Step
8 min read · 15 min practice
Job Interview
How to Navigate a Panel Interview Where Three Executives Want Different Things
8 min read · 18 min practice
Job Interview
How to Walk Through a Deal Using MEDDPIC Without Just Reciting the Acronym
8 min read · 18 min practice