Making a financial model for your early-stage startup?

You’re not alone… and some insights on what investors look for when they ask for one…

Written with George Okpamen (Bloomberg Beta Fellow) and Mark Schulz (Bloomberg Beta CFO)

You’re a few months into your startup, and you’re meeting investors for the first time. They seem to be properly warmed by your intro, you’ve laid out the case for why your service is desperately needed by some initial customers, and they like your early progress — they even seem to like you!

Then, the dreaded moment: “Do you have a financial model you can share?” (Panic. Do I? I guess I should? I can probably make one fast enough to get it to them in time. Why are they even asking me??”) “Of course I do. I’ll send it as a follow-up.”

We know that moment. As venture investors, depending on the exact stage you’re at, we might care deeply about your financials — and we know that requesting a first financial model can be like asking a founder to recite poetry in Klingon. Financial models bedevil founders, especially those who skipped a career episode making spreadsheets — early founders are busy making real products and talking to real customers, so why exactly should they now stop and make up an imaginary numerical future for their business? (So many investors also want financial models that are far too precise or forward-looking than it’s reasonable to ask of an early founder… how can you predict revenue four years from now if you only started your company four months ago?)

Financial models also trigger real worries for founders: “Do I have enough money? Does the investor understand my business better than I do? What do they really want me to say here? (I’ll say whatever I need to say, so long as it’s true, to just TTFM.) Surely there’s some moment before the IPO when I can worry about this?”

Founders’ suffering might be eased if they understand why, exactly, a well-intentioned VC might want a financial model for a startup — and therefore which aspects of making a model are worth worrying about. (The only thing worse than a nonsensical financial picture of the business is one that’s overbuilt with false precision, planning the cost in pennies of things you’ll buy in eight years.)

And yes, it’s never too early for some financial model (even before you know how you’ll make money, you’ll know how you’ll spend it!) — the question is how much modeling should you do? Let’s walk through the considerations.

Why build a financial model?

In the early days of a startup, founders should want a financial model:

1. To track where you’re spending money (and, if you have them, revenues) — a financial model can serve as a budget so you understand how much you’re spending and on what.
2. To avoid running out of money — a financial model lets you know how much investment capital you need and when (even more important during these times!).
3. To identify important drivers of the business — which are the metrics that determine whether your business will survive? What will it take to build a thriving business in the future?
4. To raise money — by projecting the future value of your business based on its current financial performance, or in some other way, and telling a story that shows you have command of the numbers.

On the flip side, here are a few questions potential investors are trying to answer when they ask for a financial model:

1. Are you going to run out of cash? Startup models are all about cash and not just about “P&L.” (More on the difference later in this post.)
2. Given what you’re raising now, what progress will you have made in the business using the capital you’re raising now? Can you raise the money necessary (from investors or your customers) given what will you have proven by that point? Put differently, what, as an investor, do I need to believe will happen for this business to survive? (Pro tip: great companies often raise the next round after spending only half the money they raise in this current round.)
3. How attractive does the business look at scale? — Once you’re big, will this business be valuable? This requires an understanding of your “unit economics” (e.g., how profitable is each individual customer, what does it cost to sell to a customer, how big is the market opportunity, etc.)

Advice on actually building a financial model

“That’s awesome, I see all this, but what actually goes into a financial model and how do I really make one?!”

After investing in extraordinary founders that started out just like you, we’ve curated a working list of “founders’ financial model questions” and helpful resources.

What exactly IS a financial model? A financial model is a summary of the expenses and revenues of your business, in the past, present, and future, usually in the form of a spreadsheet (though you can find better tools out there, like this runway calculator you can modify). In an early startup, a financial model usually shares how much cash you have on hand at a period in time, and what you’ve spent and earned (and will spend and earn). The key word here is “cash.” As early-stage investors, we’re most interested in your black-and-white cash flow at this stage, not some complex accounting of your profits and losses (aka “P&L”) that’s about theoretical profit more than what will happen to your bank account. As you get further along in your startup and into later stages (i.e. start to approach raising your series A) you can worry about getting more detailed beyond the cash flow.

How do I start building a model? Avoid starting with modifying a pre-made template you don’t fully understand (though there are many out there), because you want to learn how this works from the ground up. You might want to do some research by looking at templates and playing with them to get a feel for how they work. (We’ve provided a few resources below to get you started.)

Start with what you already know:

• Bank balance today — [from your startup’s bank account; strong founders tend to know their bank balance and watch it religiously, even daily]
• Cash coming in during a typical month — [on your bank statement, or if you’re using accounting software then you can get it from that; the usual convention is to think of early startup financials by month since there are so many expenses that only arrive once a month like rent]
• Cash going out during a typical month — [Again, simply on your startup’s bank statement or in your accounting software]

Then, you need to calculate some things. (This will take more time. You will have to do the hard work to look under the hood of your business, understand the numbers, and do some math.) The standard in financial models is to have columns indicate units of time (say, months) and rows indicate types of revenue or expense. Start with:

• Burn — [how much cash in total are you spending monthly]
• Runway — [how many months (or days!) you have left before you run out of cash]
• If relevant, revenue growth rate
• As you need them, “operating” metrics that relate to these “financial” metrics — for example, if your burn follows from how many users you have, because each user costs you a lot of money in customer service costs or bandwidth, then you’ll need to include the number of users.

(Kirsty Nathoo, Y Combinator’s CFO, explains this approach in this video: Managing Startup Finances.)

When you build the model, use cash flows, not “profit and loss.” Technically, there are three basic kinds of financial statements.

• A balance sheet shows a snapshot in time of what you own and owe.
• A profit and loss statement (P&L) shows the accounting revenue and expenses across time (e.g., if you spend money upfront for a one-year contract, it will “spread that expense” across the contract period).
• A cash flow statement shows how the money actually enters and exits your bank account each month (including what you earn/spend from operating the business, longer-term purchases or investments, and financing received from investors or other sources.)

Only cash flows really matter in (most) earliest-stage startups, as opposed to “accounting” calculations, so at day zero you can mostly defer building a P&L or a balance sheet (other than the one balance sheet item you can include on the cash flow… how much cash you have on hand in any given period).

For many startups, the P&L and cash flow are effectively identical (because revenues and expenses might all arrive monthly), which can be confusing. The typical breakdown of expenses shown in a P&L can be helpful to use (e.g., separating out different categories of expenses, especially separating staffing from non-staffing expenses) in a cash flow. Some founders show a P&L and then a cash flow that is an “adjustment” to that P&L (a common convention in bigger, and especially public, companies) — this usually makes it much harder to follow, though that depends on the startup (e.g., companies in the hardware business or with unusually “lumpy” payments from customers or to vendors may sometimes find a P&L is a better way to understand the real economics of their startup, or may need to use the P&L and cash flow in parallel).

Once you have revenue really flowing, and it helps to track when payments arrive vs. when you recognize the revenue, you may need to switch to having both a P&L and a cash flow (this may happen as early as your Series A, and later for some founders).

We hear some founders even keep multiple bank accounts — i.e., one for expenses and another for revenue — to make it easier to track things.

How do I use a model to run the business? The basics: Compare your bank account at the end of the month with what your model showed you would have — if there are differences, you’re missing something in your model. Try to figure out what caused the difference.

Let’s say you’re a few months into your startup and are pre-revenue. In that case, you might have a monthly financial model that shows how you spend money every month (meaning, it breaks down your expenses into categories like “salary,” “contractors,” “rent,” and “other expenses”).

If there are big differences, then you should take the time to analyze what you’re actually spending money on. You’ll build some of your initial model on assumptions, and some on actual operational experience. As you build the business, fewer things will be assumptions and more will be known.

Should I create a different version of the model for investors vs. what I use to run the startup? No. Keep one model, both for running the business and for presenting to investors. You can have different “views” (i.e., summarize the same underlying model differently) based on your investors’ preferences (be careful here though, you should not have more than a couple of different views). If you use different models (i.e., different projections of the future or, worse, different breakdowns of how you’ll spend money), you can end up getting confused and telling fundamentally different truths to different audiences.

For investors, how optimistic should I be in making my model projections? “If I show a rosy future, will investors think I’m unrealistic? If I do something more realistic, will the business be unexciting?”

Good investors know to put little stock into your long-term revenue projections — how could you possibly know? Chances are things move slowly until, in success, they grow much faster than you would have been comfortable projecting (because it would have seemed unrealistic).

So, for the short term (i.e. the next few months to a year of projections) be conservative. Show numbers you feel certain you can hit. Missing your numbers in those early months, when you should have more certainty, can destroy credibility (even worse, missing numbers *during* a fundraising process can ruin an otherwise successful fundraise; conversely, exceeding expectations within a raise can buoy a fundraise and create urgency for investors).

For those early months, show a high level of “resolution” in terms of timing (i.e., show monthly numbers, at a minimum; seeing annual expenses for the first two years after a round in funding gives investors little insight into how you’ll spend money).

For the longer-term projections, you should be aspirational (yet realistic). If you’re still pre-revenue, or unsure of your business model, then it’s OK to show only the current rate of burn! The only numbers you should show are the ones that are relevant in your mind to assessing your business as it is now. If you’re just getting started, ask yourself if you really need to project out more than a couple of years. If the numbers are phony, they won’t help much with investors. If your investor has a different view, they can ask, and you can discuss it.

Consider these thoughts from another early-stage investor and friend of our firm, Ben Cmejla, formerly at First Round Capital and now at The General Partnership:

“I want to know what your first chess move is, your options for your next move, and then your end game — especially in relation to the market. What I don’t want to see are five and ten-year projections for a pre-revenue company because those are just assumptions. Nobody has any real idea yet.”

If you do have a sense of your business model (maybe even a guess at pricing, or some early customers), show what the business looks like once it gets to scale. Does it have better economics than other companies in the industry? (Meaning, is it more profitable per dollar of revenue, can it get to larger scale, does it have higher prices, or some other economic advantage over competitors?) This is where unit economics matter, because the “out years” really just take your unit economics and multiplying them by more units (customers, for example). Often times those unit economics will evolve (e.g., your customer acquisition cost may grow or shrink as you scale, depending on the market you’re in and the channels you’re using).

It helps to research some industry benchmarks for whatever your assumptions are — is there a standard percentage of revenue that bookings tends to cost for startups with your particular business model? (Avoid using benchmarks that relate an expense of yours — like the cost of sales — to a backward-looking metric like revenue, when you could use a forward-looking metric like bookings.)

How fast are the market leaders growing? (Showing you understand your industry and have done your homework helps investors get comfortable, and helps make your future feel less foggy for you. It can also give you a goal to hit as a benchmark.)

You can show lower resolution timing (e.g., annual) for these longer-term “out years” — and sometimes early startups just show an illustrative single out year without showing the years in between (so long as the first couple of years are in high resolution, this can be OK).

What about for audiences other than investors? Sometimes you’ll want to use a “stretch goal” to motivate your team. It’s OK to have your model include multiple scenarios and to tell your team you’ve promised something different to investors than what you want them to try to achieve. That’s honest.

What view do I show in my pitch deck for investors? As early-stage investors, at first we want to understand a simple summary of how you plan to spend money, and (if it’s relevant to your stage) how fast you need to grow (users, revenue, or some other metric) to be able to raise more money. Different businesses have different knowns and unknowns (some may know exactly how they’ll monetize, and some not), so you need to make judgments about what’s appropriate to show for the kind of startup you’re creating.

Your summary must show unit economics. Those are the building blocks that make up your model.

If we want to dig into the model that generated the view you’re showing us, then we can ask.

What are the pro tips that will make my model look like I know what I’m doing? (Put differently: how can I build a model that reads as real vs. cosmetic?)

Less is more — An impressive model is often a simple one, where the details are few, and just right. So avoid adding for the sake of adding. The goal of your model is to provide clarity of thought and insight, as our friend Villi Iltchev (one of the most astute readers of models we know) reminds us.

Spreadsheets, please — Make sure your model is a spreadsheet, not a PDF. Yes, you can share a view in your investor deck, and that can be a PDF, slide, etc. Your actual model should be a living and breathing spreadsheet so that we can verify the numbers, calculate accurately, and follow your logic. Avoid pasting “values only” (i.e., without calculations) into a version to share with investors, because that defeats the point of sharing a spreadsheet.

Avoid hardcoded assumptions — Calculate and show your work. The only “hardcoded” numbers should be assumptions that drive other results in the model (some of these might be “knowns” like your current pricing, headcount, team salaries, etc., and others “unknowns” like the rate at which you will add new customers). Everything you can calculate from those inputs, you should calculate. Make it clear which inputs are known vs. unknown (often times you can include unknowns in a different color, like blue, to tell the reader “you can change this”).

Use real labels — Use actual month and year names (e.g., Aug ‘22, Sep ’22 or 2022, 2023 vs. Month 1, Month 2 or Year 1, Year 2). That contributes to the sense that your numbers are real, and not a theoretical business.

Organize expenses — Make sure the line items you use actually reflect the way you spend money (for example, one line item that says “people” and includes 80% of your cost, and 20 line items for random incidentals that are all tiny provides less insight than breaking down “people” into different functions or some other way of understanding the spend). In most startups, it makes sense to separate “people” expenses from “non-people” expenses.

Avoid visual clutter — As your model gets more complicated, use separate tabs for separate areas of the business, if needed. One page for staffing, for example, and another for customers and revenue. Keep one tab (ideally the first one) as the summary, and add any notes that a reader might need to understand the model. At the earliest (i.e., before seed) stage, it’s rare you’ll need more than one tab.

Be consistent across time — Keep the same categories for actual (i.e., historic) cash flows vs. projected cash flows. And make the sure the historic results are the actual results (some models show past projections without updating them to reflect what really happened!). We often see models with totally different categories of spend for the past vs. the future — or one set of statements that covers the past and a completely unrelated model that covers the future — mostly due to blind copying of previous templates vs. actually doing the hard work to understand the model). That makes it difficult to see how your spend and revenue are changing over time, and specifically before and after our investment.

Test the model yourself — Experienced investors pressure test your model by using common sense to make sure the numbers seem real (e.g., if you have a startup’s headcount rising from 2 to 15 in one of your early months… that seems unlikely, not too many founders can really hire that fast). Many models fail the “do you really plan to spend money that way?” simple question test. (Founders say no, and then we ask them to show us their financial model as they really plan to run the business.) Investors may also “audit” select calculations by trying to follow them through to their conclusion (e.g., “how did you calculate salary in August 2023”). So you might as well do that kind of work yourself and make sure the model stands up to simple scrutiny. Lots of money has been lost on simple calculation errors (like models that left out an expense line in the monthly totals, dramatically undercounting their costs).

Should I get a contractor or accounting firm to build my financial model? Our take might seem counterintuitive: the more experienced you are with building a financial model, the more comfortable we are with outsourcing. In the early days of a startup, you as a founder should be hands-on and knowledgeable about the most important parts of your business (and the financial model is certainly one of them). Delegating that knowledge — which will drive the success of your startup — to someone else can be dangerous. If you’ve built models before, then you can more easily manage someone else doing the work. (This logic applies to other areas of hiring, and we call it the cell division model of early hiring.)

Either way, you as the founder are still responsible for knowing the numbers and drivers of your business. This is both for your own benefit as the ultimate business leader and as you fundraise with investors.

What does my financial model have to do with my cap table? A capitalization table is a breakdown of who owns what pieces of your startup — between you and your co-founders, employees, and investors. It determines how much each person gets paid if you have profits to distribute, or (more likely in venture-backed startups) an exit. While it’s often also shown as a spreadsheet, it rarely has any direct tie to the financial model at all.

What do all these financial words even mean?

• pro forma // a Latin term that means “for the sake of form” or “as a matter of form,” where you make calculations using projections or presumptions. Every financial model that deals with the future (or cap table that illustrates ownership after closing a round) is a pro forma.
• financial forecasting // the process by which a company thinks about and prepares for the future. Forecasting involves determining the expectations of future results.
• financial modeling // the act of taking a forecast’s assumptions and building a predictive model that helps a company make sound business decisions.
• profit & loss (P&L) statement // a financial statement that summarizes the revenues and expenses incurred during a specific time period, adjusting for the period where the revenue or expense actually is relevant to the performance of the business (as opposed to the contractual mechanics of when an expense might actually be paid). Based on accrual accounting principles, or (for public companies, especially) generally accepted accounting principles (GAAP) in the U.S., a P&L assigns different costs and revenues to different time periods. For public companies, this is one of three financial statements shared quarterly and annually, along with the balance sheet and cash flow statement.
• balance sheet // a financial statement that reports a company’s assets, liabilities, and shareholder equity at a specific point in time. Since it only provides what a company owns and owes at a specific date, this cannot give financial trends for your startup in and of itself.
• cash flow statement // a financial statement that provides aggregate data on all cash inflows and outflows a company receives and pays — for its ongoing operations, external investments, and other business activities.
• revenue recognition // a GAAP (generally accepted accounting principle) that describes which period of time revenue appears in the P&L. This uses accrual accounting, which recognizes revenues when “realized” and earned — not necessarily when cash is received. For example, if a client pays you today for advertising on your website that will run in six months, the cash flow will show the funds arriving today, and the P&L will show the funds arriving six months from now.

Usually we say #thisisnotadvice, but actually this time this is advice.

Just like any good financial model, we’re still evolving (ideally you do maintenance on your model every month vs. doing a batch adjustment all at once when it’s time to raise another round) and we’ll update this piece from time to time. We’ve included some more resources below, and please let us know (in the comments or otherwise) if there are other questions we can answer at Bloomberg Beta.

Additional resources:

[Helpful posts & best practices]

• foresight: best practices and resources for financial modeling (including free templates) // strength: provides a plethora of resources from templates, to spreadsheets, to pitch decks, etc. // weakness: founders can get caught in a rabbit hole and get even more confused across the well-intentioned overabundance of information.
• visible blog: building a financial model that works // strength: one of the more well-written, thorough, and applicable posts // weakness: Some of the context is a little dated and may be more fitting to later-stage startups
• recurring company: making financials public // strength: provides step-by-step guidance on how to build a financial model // weakness: only applies to a specific type of business and mostly trying to create demand for their templates for other types of businesses
• rocket pro forma: pitch deck financials // strength: gives tangible advice on the purpose and parts of a financial model // weakness: advice may be overly general, trying again to create demand for paid resources
• efinancialmodels: financial modeling for startups // strength: broadens the scope of financial models to include non-venture backed businesses and global examples // weakness: a little too broad for US and venture-backed businesses
• bessemer venture partners: publicly available investment memo of pagerduty // strength: provides a real look at how investors evaluate an “early-stage” company’s financials in the context of entire company // weakness: the company is closer to a growth-stage company after product-market fit and could be out of scope for most pre-revenue/pre-product market fit companies to glean direct learnings
• prophet models: request a demo // strength: provides a variety of solutions for modeling, including templates for venture-backed startups and even modeling venture capital funds // weakness: requires a demo to see the templates
• venture ready models: startup financial models for sale // strength: provides a focused set of financial model templates to preview // weakness: blog and templates are a bit dated and have not been refreshed since early 2017

[Accounting basics]

• swartz center for entrepreneurship: accounting for startups // strength: provides a rudimentary guide for startup founders trying to understand and apply beginners concepts in accounting // weakness: guide might be a bit too beginner level for founders who have a stronger command of fundamental accounting and finance concepts
• halon tax: bookkeeping basics // strength: delivers a step-by-step presentation with hands-on examples to drive home understanding for founders // weakness: the examples used reference small-business vs. being tailored specifically for high-growth startups

[Financial model courses & tutorials]

• mike lingle: financial projections live youtube walkthrough // strength: clearly breaks down components of a financial model // weakness: may be too general for some founders, though shows models in ways that may be less than ideal (e.g., uses a P&L instead of a cash flow, and theoretical “month 1” vs. actual calendar months)
• slidebean: financial modeling 101 tutorial // strength: gives a real founder’s perspective of how and why he uses a financial model // weakness: it is a direct sales pitch from the founder for the service he sells (financial modeling help for startups)
• lili balfour: how to build a startup financial model // strength: gives a relatable framework of how to think about a financial model, though also covers other areas like growth metrics // weakness: may include extraneous information for founders who just want to build a model

Note: The financial forecast and runway calculator, to which we also linked in the text above, was built by Summit, a Bloomberg Beta portfolio company.

Thanks to Jennifer Dennard, Villi Iltchev, Anthony Xu, Craig Thompson (and others Medium automatically thanks at the bottom of this post) for reviewing drafts. This post came in part from a talk to Eve Blossom’s investor training program, Material Change.

If you know a founder who would benefit from this guidance, of course feel free to share. You can also find this guide at bit.ly/earlystartupfinancialmodel