Why your assumptions matter more than your projections

Every investor who has reviewed more than a dozen startup models has a version of this story: a founder walks in with a three-year projection showing $15M ARR and a clear path to profitability. Everything looks reasonable on the surface. Then you ask where the 20% monthly growth rate in year two comes from, and the honest answer is that it looked achievable.

The projection was confident. The assumption behind it was not there.

This is the most common structural problem in early-stage financial models - not that the projections are wrong, but that the assumptions are either missing, vague, or buried in cells with no explanation. Projections without defensible assumptions are not financial models. They are organized optimism.

What assumptions actually are

In a financial model, an assumption is any input that you have decided on rather than calculated. Your monthly churn rate is an assumption. Your conversion rate from trial to paid is an assumption. The salary you plan to pay your second engineering hire is an assumption. The number of customers you expect to acquire from your LinkedIn channel in month 6 is an assumption.

Projections are the outputs that flow from those assumptions. Revenue in month 12 is not an assumption - it is the result of your acquisition assumptions, conversion assumptions, pricing, and churn combined. Changing any input changes the output.

The distinction matters because it tells you where to focus attention. Investors are not primarily trying to validate your revenue projections - they are trying to understand whether your assumptions are grounded in something real, and whether you understand which ones are doing the most work.

Why explicit assumptions are a competitive advantage

A founder who can say "we are assuming 3% monthly churn based on our first cohort of 35 customers at 90 days" is in a fundamentally different position than a founder who says "we assume churn will be low." Both might be using the same number in their model. One of them has made the assumption testable, which means they will know sooner when it needs to be revised.

This matters for three reasons:

Learning speed. The case for building a financial model before you have real numbers makes exactly this point: the purpose of a model at early stage is to make beliefs testable. When your assumptions are explicit, reality tells you something specific when it diverges. When your assumptions are vague, reality just feels confusing.

Investor trust. Explicit assumptions signal that you have done actual research, that you have tested things, and that you understand the mechanics of your business. "Our CAC assumption is $280, based on three months of paid acquisition data at$340 with ongoing optimization work" builds trust in a way that "CAC around $300" does not.

Model maintainability. A model built on explicit, separated assumptions is a model you can update in minutes when something changes. A model where assumptions are embedded in formulas across 12 tabs is a model nobody updates. This is the core argument behind Binokl vs spreadsheets - not that spreadsheets are wrong, but that assumptions buried in formulas are fragile by design.

What makes an assumption defensible

There is a hierarchy of evidence behind startup assumptions, roughly from weakest to strongest:

Guesswork - "We think churn will be about 3%." This is a starting point, not an assumption. Fine for day one, not for a pitch deck.

Industry benchmarks - "SaaS businesses at our stage average 4-6% monthly churn; we are modeling 4%." Better, but still not specific to your business.

Comparable companies - "We looked at three SaaS tools in adjacent categories that disclosed metrics; their churn is 2-5%. We are modeling 4% conservatively." More specific, but still borrowed.

Your own early data - "Our first cohort of 40 customers has retained at 96% through 90 days. We are modeling 95% monthly retention (5% churn) to be conservative." This is the standard to reach for. Early data, even from a small sample, is significantly more defensible than any benchmark.

Signed contracts or commitments - "Three customers have signed LOIs at $1,500/month. We are modeling ACV based on these." As close to certainty as early stage gets.

Building an "assumptions library" - a folder or document where you collect the evidence behind your key inputs - is good practice. It serves two purposes: it makes you think explicitly about what you actually believe and why, and it prepares you for due diligence, where investors will ask exactly these questions.

The assumptions that matter most

Not all assumptions have equal impact on the model. Some move the output dramatically when changed; others barely register. Knowing which assumptions are doing the most work in your model is part of understanding your own business.

For most early-stage SaaS, the highest-leverage assumptions are:

Monthly churn rate - because it sits in the denominator of LTV and has a compounding effect on the customer base over time. A difference between 3% and 6% monthly churn does not feel large, but over 24 months it produces dramatically different ARR and runway outcomes. This is why MRR, ARR, and churn tracking is so central.

CAC by channel - because it determines how much acquisition costs, which determines how much runway each month of growth consumes. Optimistic CAC assumptions are one of the most common ways models overstate financial health.

Conversion rates - from visitor to trial, trial to paid, and sales pipeline to close. These compound across the acquisition funnel and small differences add up to large revenue differences at 12 months.

Hiring timeline - because headcount is often the largest cost line and has a direct runway impact that founders underestimate. Delaying one hire by a quarter sounds small. Modeled correctly, it often adds 2-3 months of runway.

The sanity checks you should run on every model

Before presenting a model to investors, run these basic checks:

• Can revenue projections be larger than the TAM you have defined? (They should not be)
• Do operating expenses grow in proportion to revenue? (Costs that are flat while revenue triples are usually wrong)
• Are gross margins realistic for your infrastructure and delivery model? (See gross margin for SaaS founders)
• Does the funding need match the burn rate and milestones in the model?
• Are the key metrics - ARR, churn, runway - consistent with each other?

These checks do not catch every problem, but they catch the obvious ones that cost credibility quickly.

Founder checklist

• Every key assumption in your model is written down and can be explained in one sentence
• Your highest-leverage assumptions have evidence behind them, not just intuition
• You know which three assumptions, if wrong, would most significantly change the outcome
• Changing an assumption in your model updates everything downstream automatically
• You can answer "where did that number come from?" for any cell an investor might ask about

How Binokl helps