The problem isn't the tool. It's that most startup spreadsheets are built the wrong way: hardcoded assumptions scattered across 4 tabs, no version control, formulas that break when someone adds a row. This guide covers how to build Sheets models that hold up under investor scrutiny and scale with your business as it gets more complicated.
The 5 Sheets Every Startup Actually Needs
You don't need 20 tabs. You need 5 that talk to each other cleanly.
1. Assumptions — one tab, all inputs. Every hardcoded number lives here: monthly growth rate, CAC, churn rate, headcount by quarter. Everything else references this tab. When your investor asks "what if growth drops to 8%?", you change one cell and the whole model updates.
2. P&L — revenue minus costs, monthly, 24 months out. Build it as a formula-driven projection, not a hand-typed table. Revenue should pull from your unit economics model; expenses should pull from your headcount plan.
3. Cash Flow — different from P&L. This shows when money hits the bank versus when it's recognized. For SaaS with annual contracts, the gap between recognized revenue and cash received can be significant. Investors who've seen companies die solvent care about this one specifically.
4. Unit Economics — LTV, CAC, payback period, contribution margin per cohort. A single tab that shows you understand your business at the unit level matters more to most seed and Series A investors than the full 3-year projection they know you made up.
5. Cap Table — ownership percentages, option pool, pro-forma dilution for your current raise. Google Sheets handles this fine at pre-Series B. After that, move to Carta.
Building Your Runway Model
Runway is the number every founder needs to know on any given Monday. The formula that breaks most startup models is treating burn as a fixed monthly number. It isn't — it changes every time you hire, add a tool, or expand office space.
Build it dynamically. Your monthly burn should sum from a headcount tab where each hire has a start date, salary, and benefits load (typically 15-20% on top of base in the US):
=SUMPRODUCT((C$2:C$50<=$A3)*(D$2:D$50="")*E$2:E$50)*1.175
This formula sums salaries for all employees hired before the current month who haven't left, and applies a 17.5% benefits load. Your Assumptions tab holds the 0.175 constant so you can change it in one place.
Your runway cell then becomes:
=IFERROR(MATCH(TRUE,MMULT(N(ROW(INDIRECT("1:"&COUNTA(B5:B28)))),N(CUMULATIVE_CASH<0)),0)-1, "24+ months")
Or, more readably, just calculate the month where cumulative cash hits zero and surface it on a summary dashboard.
As of March 2026, the median pre-seed startup had 14 months of runway at close, according to Carta's state of private markets data. Investors generally want to see 18 months post-raise modeled in your deck. If you can't get there on the current raise size, show the milestones that justify a smaller number.
Revenue Modeling for SaaS
The standard SaaS revenue model has three inputs: beginning MRR, new MRR added each month, and churn. Everything else derives from those.
=B2 + C3 - (B2 * $Assumptions.$B$12)
Where B2 is prior month MRR, C3 is new MRR, and $Assumptions.$B$12 is your monthly churn rate. Lock that reference with $ so you can drag the formula across 24 columns without it drifting.
The mistake most founders make is modeling churn as a percentage of total MRR rather than separating logo churn from dollar churn. A 2% monthly logo churn rate with net dollar retention of 110% is a completely different business than 2% logo churn with NDR of 85%. Your model should show both, especially if you have any expansion revenue.
One useful benchmark: according to OpenView's 2024 SaaS benchmarks report, median net dollar retention for B2B SaaS companies at $1-10M ARR is around 102%. If your model shows 120%+ without a clear explanation of why, an investor will push on it.
What Investors Actually Look For
Investors have seen thousands of pitch decks and models. The things that immediately signal a weak model:
A single row for "Revenue" with no supporting calculation. Show your math — cohort by cohort, channel by channel, contract by contract if you're early enough to have that visibility.
Expenses that grow linearly with revenue without explanation. In reality, headcount scales in steps (you hire an engineer, then another 6 months later), infrastructure costs scale roughly logarithmically, and sales costs should be modeled as a function of pipeline, not revenue.
CAC that improves every single quarter. This happens in models because founders extrapolate early efficiency gains indefinitely. Real CAC tends to increase as you exhaust early channels. A model that shows declining CAC through year 3 without a mechanism (a new low-cost channel, outbound improving, etc.) will get called out.
A Non-Obvious Advantage of Sheets Over Dedicated Tools
Most startup finance tools (Mosaic, Runway.com, Causal) are built for finance teams that need audit trails, role permissions, and accounting integrations. Pre-Series B, you need none of that.
What you actually need is speed. The ability to restructure your model in 10 minutes when your co-founder has a new idea about pricing, or when an investor asks a question in the room that changes your assumptions about the market.
In Sheets, restructuring a model is just moving cells and updating references. In a purpose-built tool, it's often a multi-day project because the tool's data model has opinions about how your business should be structured. The flexibility Sheets gives you in years 1-3 is worth more than the polish of a dedicated tool.
The limitation hits when you're pulling in real data. Manually updating Sheets with numbers from Stripe, HubSpot, or your data warehouse is where founders lose 2-3 hours a week. That's where ModelMonkey helps — it pulls live data from your connected tools directly into the sheet, so your P&L and unit economics tabs stay current without the copy-paste ritual.
Structuring for Investor Sharing
Before you share a model with an investor, do three things.
Lock your Assumptions tab. Use Data > Protected ranges to prevent edits to your core inputs. Add a note explaining what's protected and why. Investors who dig into models sometimes accidentally break things; this prevents that.
Add a Summary tab as the first sheet. It should show: current ARR or MRR, monthly burn, runway in months, and the key milestones the current raise is designed to hit. Make this one tab something a non-technical person can read in 30 seconds.
Remove the cruft. Every model accumulates dead tabs — old scenarios, test calculations, abandoned approaches. Delete them before sharing. An investor poking through 12 tabs looking for the actual model is an investor getting annoyed.
In summary: your Google Sheets model should have clean inputs, formula-driven projections, and a one-tab summary built for a fast reader. Everything else is optional.