DCF

A discounted cash flow model answers a question the other valuation methods sidestep: what is
this business actually worth based on the cash it will generate, independent of what the market
happens to be paying for similar companies right now. That independence is the appeal — a DCF does
not need a comparable company or a recent transaction to exist, which matters for a genuinely novel
AI business that has no close peer to borrow a multiple from.

The independence is also the danger. Because the answer comes entirely from assumptions about the
future — how fast revenue grows, how margins evolve, how long the growth persists, and what discount
rate reflects the risk of getting there — a DCF can be made to say almost anything by someone
adjusting those assumptions, deliberately or not. That is why experienced negotiators on both sides
treat a DCF as a sanity check on a valuation reached some other way, rather than as the number
itself.

The model forecasts free cash flow for a number of future years, then discounts each year back to
today's value using a rate that reflects the risk of the business — a rate that in a founder's own
model is a chosen assumption, not an observed fact, and one a buyer's advisors will challenge
directly. Beyond the explicit forecast period, a terminal value captures everything the business is
expected to be worth afterward, and that terminal value very often accounts for the majority of the
entire result — a company forecast for five years and given a terminal value beyond that can find
most of its valuation sitting in the assumption about what happens in year six and onward, not in
the five years actually modelled.

Because the terminal value dominates, and because it depends on an assumed long-run growth rate and
discount rate that are themselves debatable, a DCF is unusually sensitive to small changes in inputs
that feel minor when you are choosing them. Two analysts using almost identical operating forecasts
can land on very different valuations purely from disagreeing about the discount rate or the terminal
growth assumption — which is the central reason this method resists being reduced to a benchmark
number that travels between companies.

Small assumption changes produce large valuation swings. A discount rate or terminal growth
assumption that looks like a rounding difference between two analysts can move the resulting
valuation substantially. Never accept a DCF output without seeing the assumptions that produced it,
and never present one of your own without being ready to defend each input individually.

The terminal value usually does most of the work. If the majority of the valuation sits beyond
the explicit forecast period, the model is really a bet on a long-run growth and discount
assumption dressed up as a detailed multi-year forecast. Ask what share of the total value comes
from the terminal value before trusting the headline number.

A DCF forecasting hypergrowth off a small base is fragile by construction. Early-stage
companies, including most AI startups, have thin operating histories to forecast from, and a model
built on a short track record inherits all the uncertainty of that history, compounded forward for
years. Treat the output as one input among several, not as the answer.

Use it to check other methods, not to replace them. A DCF is at its most useful when its result
is compared against a revenue multiple or EBITDA multiple
derived valuation — a large gap between the two is worth understanding, but neither number should
be trusted purely because it exists.

The AI analysis report does not lean on a single discounted-cash-flow output as the answer; the
eight-model pipeline reads your actual history and public market data to produce a valuation range
with per-section confidence, which is the honest response to exactly the sensitivity described
above — one point estimate from one set of assumptions is not a number worth building a negotiation
on. That range is a useful anchor heading into the Valuation Up-Lift stage on the M&A track or the
Valuation Alignment stage on the investment track, where the two sides converge on an actual figure.

Try the free tool at /valuation to see where your business lands before you build a
DCF of your own.