Here's what I look for in a seed-stage financial model.

As an investor at Commonweal Ventures , I review hundreds of pitch decks and financial models each year. ?In my view, pre-Series-A founders spend far too little time on their financial models. The typical founder would be better off spending ~20% less time on their pitch deck and reallocating this attention to their financial model.

A polished pitch deck and smooth intro meeting with a VC get you in the door and convey that you have strong founder-market fit. But once investors believe that you are the right founder to launch this particular business and that you know how your industry works, we want to see that you’ve thought about your business’s growth story with clarity and depth. Your financial model is quite likely to be the only sample of your* analytical capabilities that investors can dig into during the diligence process. A well-constructed model communicates that you can think clearly about your business - in particular, that you have well-reasoned views on how specific inputs into the business are likely to generate growth.

Put simply, investors want to see a story of your business’s growth over the next ~5 years that is simple, testable, and plausible. Below, I lay out five actionable steps that founders can take to create a top-tier model. I’d estimate that the top ~20%** of financial models I review get all five of these things right.

1. Blue and black color coding. This might sound pedantic, but I genuinely believe that using different font colors for input cells versus calculated cells helps clarify your thinking as you construct the model. Input cells should be blue, calculated cells should be black. By “input cells,” I mean literally any cells where you are typing a specific number into the cell itself rather than using a formula to generate the cell's value—e.g., the price of your product, the number of active customer accounts you currently have, etc. “Calculated cell” means any cell whose value is determined by a formula—e.g., if next month’s total customer accounts are determined by applying some growth rate and churn rate to this month’s total, then the cell containing next month’s total customer accounts should have black font. Ideally, all your blue input cells should be on a single tab at the front of your model, rather than embedded within each individual backup tab. If you want to earn bonus points with investors who previously worked on wall street, the hex code for the “right” blue font is #0000FF…IYKYK.
2. No hard-coding or hard-coded calculations. “Hard coding” refers to the (horrifying!) practice of simply entering a value into a cell when that value should in fact be the result of some calculation. If you’re forecasting the total number of customer accounts your company will close over the next five years, and you simply write these numbers into cells as “100,” “150,” “200,” etc., that would be hard coding. Instead, you should forecast your total number of customer accounts via some formula so that investors can see your underlying thought process—e.g., apply some growth rate to today’s number of customer accounts to arrive at next year’s. By “hard coded calculations," I mean the (only slightly less horrifying) practice of calculating formulas by typing the input numbers within a given cell, rather than drawing those number into the formulas with cell references. To stick with my example from above, let’s assume this year’s customer account figure is in cell B2 and you want to forecast next year’s customer accounts to be 50% higher than this year’s. Typing “= B2 * 1.5” would be a hard-coded calculation. Hard coded calculations frustrate investors because they hide your reasoning within individual cells, rather than making that reasoning explicit and easily testable. Instead, you should put the 50% growth rate in its own input cell, say B1, and calculate next year’s figure with the formula "= B2 * (1+B1).” Essentially, every cell in your model should either contain a single number (an input) or a formula that references other cells, but only contains cell references—never multiple numbers performing some operation within a given cell.
3. Basic rationale for external inputs. This recommendation is most relevant for businesses whose unit economics are dependent on some external data points, like the market-wide price or transaction volume for a certain product--and are thus likely to generate debate among investors. For instance, say your business is streamlining homeowners’ purchases of heat pumps with some cool new software platform. Your software revenue is likely dependent on overall volume of residential heat pump installations and prices in some region. For external inputs—like, the sticker price of a heat pump or the total number of heat pumps installed in Massachusetts, etc.—you should provide a simple rationale or source to back up your inputs. Literally a single url pasted into the cell next to a particular input will do. This action should take you about five seconds and will save you at least an hour (if you add up all the time you’ll inevitably spend answering emails from investors asking you to provide a source for that estimate of heat pump sales in Massachusetts).
4. Drivers-based revenue growth. In terms of conveying your clarity of thought, this is perhaps the single most important element of a seed-stage financial model. Yet it's also typically the weakest element of most models I review. Instead of an assumed growth rate of, say, 200% every year for the next five years, I’d like to see your view on how specific "drivers"--or, inputs into your business--will cause revenue to grow by 200%. The benefit of a drivers-based revenue model is that it is testable and conveys to investors that you can think about your business as a “machine” with clear inputs and outputs, rather than a magical black box that produces exponential sales growth. Let’s stick with the heat pump software example from above. Drivers in your model might include: (i) how many BD reps you’ll hire for each region into which you expand, (ii) the likely sales volume generated by a typical BD employee (+ how this scales with your hiring plans), and (iii) volume of spend you’ll allocate toward marketing (+ likely ROI on that spend). There is no single right answer for the relationship among these drivers—much of the value here lies in demonstrating to investors that you are the sort of founder who thinks about your business’s growth with structure.
5. Drivers-based gross margins. Similarly, your view on gross margins should be based on some underlying rationale about the resources you’ll need to align to each incremental dollar of revenue, in terms of technology and employee time. If I see that you’ve simply input a gross margin trajectory from 60% today to 90% in five years, I don’t know how much credence to assign to this path. I’d rather see your view on how many customer support reps you’ll need to hire over the next three years and how that personnel cost will scale with revenue to translate into margin expansion. Even if this particular projection turns out to be wildly wrong, we can look back and debate the reasons why personnel costs turned out to be much higher or lower than expected.

??

If you’ve read this far, I hope you got some value from this post. And if not, well…you got what you paid for! As a reminder: I’m an early-stage investor at Commonweal Ventures , where I invest in startups focusing on climate, government services, mobility, and the built environment. If you’re an early-stage founder building a business in one of these areas, don’t hesitate to reach out. Even if you’re not yet raising, we’re always happy to meet founders early in the company-building journey.

--

*or your team’s! Self-awareness is one of the most important traits in a founder. If you’re not the person to construct a top-tier financial model, go find the right person. Feel free to contact me for some recommendations of fractional CFOs / consultants who can get your model in gear.

** Remember, a half-decent early-stage VC firm invests in maybe 1% of the pitches they see in any given year, so top 20% is not even close to good enough. Consider the recommendations in this post necessary but not sufficient for a successful raise.