How to Estimate Startup Revenue (Without Lying to Yourself)

The most common startup revenue projection starts with the market. "The market is $50B. We capture just 1% — that's $500M." This is top-down thinking, and it's almost always wrong because it ignores how you actually get customers.

A credible projection starts from the bottom up — from the actual mechanics of how you acquire customers, how much they pay, and how long they stay. Let's build one.

▸ PROJECTION_METHODS

Bottom-Up vs Top-Down Projections

Both are legitimate. They serve different purposes. Investors expect both — but will trust the bottom-up far more.

Top-Down

TAM $10B → 0.1% capture → $10M revenue

+Shows market context

+Easy to explain to non-technical audiences

−No basis in operational reality

−1% of market is often unachievable

−Ignores customer acquisition mechanics

Use for: showing market opportunity. Not for: operational planning.

Bottom-Up

200 leads/month → 3% conversion → 6 customers → £499 ACV → £2,994 MRR

+Based on real acquisition assumptions

+Exposes where model breaks down

+Forces you to think channel by channel

−More work to build

−Assumptions may be wrong

−Harder to impress with big numbers

Use for: everything. This is the real plan.

▸ KEY_VARIABLES

The 4 Variables You Must Estimate First

Before you build any model, pin down these four numbers. Everything else flows from them. If you can't estimate these with any confidence, your projection is guesswork.

01Addressable Market Size

How many people/businesses could realistically buy from you — not the total market, the reachable segment with your budget and channel?

Example: You sell HR software to UK companies with 50–200 employees. That's roughly 18,000 companies. At £5/company in marketing cost, you could reach ~2,000 of them in year one.

Why it matters: This caps your realistic upside and tells you whether the market is big enough to build a real business.

02Conversion Rate

Of 100 people who encounter your product (visit your site, take a sales call, trial your software), how many buy?

Example: SaaS trial-to-paid: 2–8%. Inbound lead-to-close: 15–30%. Cold outbound: 1–3%. Start with the low end.

Why it matters: This is almost always the most uncertain variable and the most dangerous to overestimate.

03Average Contract Value (ACV) / ARPU

What does the average customer pay in year one? Not what you hope — what does your pricing page actually charge and what do customers realistically choose?

Example: If 70% of customers choose the £29/month plan and 30% choose £79/month, ARPU = £29×0.7 + £79×0.3 = £43.90/month.

Why it matters: ACV drives everything. A 20% increase in ACV is worth far more than a 20% increase in conversion rate if you're already volume-constrained.

04Churn Rate

What percentage of customers cancel per month? This is the variable most founders ignore until it's too late.

Example: At 5%/month churn, a customer acquired in January is 54% likely to still be paying in December. At 2%/month, they're 79% likely to still be paying.

Why it matters: Churn compounds. Even a 3% monthly churn rate destroys revenue projections after month 6.

▸ 12_MONTH_MODEL

Build a 12-Month Projection

A monthly model forces you to think about when things happen, not just what happens in total. Here's a sample subscription business model with conservative assumptions: 20 new customers/month, 4% churn, £45 ARPU.

Month	New	Churned	Total Customers	MRR	Cumulative Rev
Month 1	+20	−0	20	£900	£900
Month 2	+20	−1	39	£1,755	£2,655
Month 3	+20	−2	57	£2,565	£5,220
Month 4	+20	−2	75	£3,375	£8,595
Month 5	+20	−3	92	£4,140	£12,735
Month 6	+20	−4	108	£4,860	£17,595
Month 7	+20	−4	124	£5,580	£23,175
Month 8	+20	−5	139	£6,255	£29,430
Month 9	+20	−6	153	£6,885	£36,315
Month 10	+20	−6	167	£7,515	£43,830
Month 11	+20	−7	180	£8,100	£51,930
Month 12	+20	−7	193	£8,685	£60,615

YEAR-1 REVENUE

£60,615

MONTH 12 MRR

£8,685

ASSUMPTIONS

20 new/mo · 4% churn · £45 ARPU

▸ MODEL_YOUR_SCENARIO

Use BoringRiches to model different scenarios

50 interactive revenue calculators. Adjust every assumption and see the outcome change in real time.

Browse All Businesses →

▸ STRESS_TEST

Stress-Test Your Assumptions

Your base case will be wrong. The question is which direction it will be wrong in, and by how much. Build a sensitivity table before you commit to any assumption.

SENSITIVITY: What if new customers/month is 50% lower?

Scenario	New/Month	Month 12 MRR	Year-1 Revenue
Bear (−50%)	10/mo	£4,365	£30,510
Base	20/mo	£8,685	£60,615
Bull (+50%)	30/mo	£13,095	£91,485

The spread between bear and bull is significant. Plan to survive the bear case. Build the systems for the bull case. Don't spend the difference in advance.

▸ COMMON_MISTAKES

Common Projection Mistakes That Kill Investor Credibility

Hockey stick in month 4 with no explanation

Growth inflection points need a cause — a product launch, a PR hit, a channel turning on. If your curve goes exponential with no trigger, investors will dismiss it.

Assuming 0% churn

Every SaaS has churn. If your model assumes every customer you ever acquired is still paying in month 12, it's wrong. Sophisticated investors will know before you finish presenting.

Ignoring CAC

20 new customers/month sounds achievable — until you work out what it costs. If each costs £200 to acquire, that's £4,000/month in acquisition spend. Is that in your model?

Point estimates instead of ranges

A single number implies false precision. Ranges — with explicit assumptions driving each scenario — signal that you understand uncertainty. This builds more credibility than a confident wrong number.

▸ FAQ

Frequently Asked Questions

How do you estimate revenue for a startup with no data?

Start with analogues — businesses similar to yours that have published revenue data (Indie Hackers, SaaS benchmarks, competitor case studies). Use their conversion rates and ARPU as your starting assumptions, then adjust based on how you differ. No data doesn't mean no model — it means wider ranges.

What is a realistic revenue target for a startup in year one?

Depends entirely on the model and how much capital you're deploying. A bootstrapped SaaS with one founder might end year one at £0–£5K MRR. A funded startup with a sales team might target £50K+ MRR. Don't set a target without understanding your acquisition mechanics first.

What is the difference between a revenue forecast and a revenue projection?

In practice, people use these interchangeably. Technically, a forecast uses historical data to extrapolate forward. A projection is based on assumed scenarios ('if we do X, then Y'). For early-stage startups with no history, everything is a projection.

How far ahead should a startup project revenue?

12 months in detail (monthly). 3 years at a high level (annual). Anything beyond 3 years for an early startup is decoration. Focus your energy on making the 12-month model as honest and defensible as possible — that's what drives real decisions.

▸ MODEL_YOUR_SCENARIO

Use BoringRiches to model different scenarios

50 interactive revenue calculators. Adjust every assumption and see the outcome change in real time.

Browse All Businesses →