A Primer on Blockchains, Protocols and Token Sales

Blockchains are changing the foundation of the global economy, and everyone is freaking out.

Opinionated tribes are invoking civil wars and “forking” into new tribes. Multiple heated battles are happening simultaneously (mostly on social media) while only a few people on the outside understand what’s going on.

Even the terminology being thrown around is confusing; there’s been a rapid evolution of new technology that has resulted in a lack of lingo consensus. What is an ICO vs. a token sale vs. a TGE? Are coins, tokens, and cryptocurrencies the same thing? What exactly is a “protocol”, “DAO”, “dApp”, etc…

Still, a growing number of people are realizing that there is a way to transfer assets to each other securely, reliably, publicly, irreversibly, and globally — without the need for a central authority.

Marketplace corporations that profit from buyer/seller transactions (like Uber, AirBnB, eBay, Amazon, etc…) are paying especially close attention.

We’ve entered a world where “trust” is moving toward distributed networks of machines that no one person, group, corporation, or government owns. These networks have rock-solid data integrity, zero downtime, and financial incentives for anyone who participates.

The implications of this are enormous.

How we do business with currency, credit, contracts, real estate, computing, communications, fundraising, and pretty much anything else you can think of is changing.

It’s happening right now, and it’s happening fast.

What is a blockchain?

Simply put, a blockchain is a very long list of transactions stored on a network of computers (i.e. a distributed ledger).

As an illustration, imagine writing on a piece of paper that resulted in thousands of other pieces of paper around the world magically showing what you wrote. A blockchain is like that.

What is special about this distributed ledger is:

  • It can smartly establish consensus of state (i.e. the transactions) even if huge swaths of computers on the network go down.
  • The transactions are organized as a list of timestamped line items in the ledger inside blocks of data.
  • Cryptography ensures that past transactions cannot be deleted or modified.
  • Everyone can see the history of every transaction since the beginning of the chain’s creation.
  • People are financially incentivized to spin up computers in the network, which secures the network’s ongoing existence.

Thus, math secures a blockchain’s past, and money secures a blockchain’s future.

What are protocols?

Most of us are used to entrusting a bank to maintain a proper ledger and ensure money is not “double spent.” But with blockchains, that trust is in the network of computers and their cryptographically secure protocol (i.e. the software that governs rules, operations, and communication between network nodes).

As an example, the Internet today runs on a relatively “thin” layer of protocols such as TCP/IP, SMTP, HTTP, and HTTPS. These protocols establish rules for computers across a network to follow in order to communicate effectively.

On top of these protocols, a relatively “fat” layer of applications like Google, Facebook, Twitter, etc.. have been built to interface between data and people. With this setup, centralized organizations own all the data and users are required to create new accounts and enter credit card info (and other sensitive data) in multiple places. This is annoying, especially when our identity and payment info gets hacked.

Enter blockchains

Blockchains are also operated and accessed via protocols, but the critical difference is that instead of a few “thin” protocols, a huge number of niche protocols are being built (i.e. “fat protocols”) that will support the next generation of applications. This setup allows users to “own” their data, and it allows developers to leverage a multitude of protocols and blockchains that will enable more powerful, secure, and inexpensive solutions for everything from payments and banking, to healthcare and telecommunications.

How can people transfer assets to each other on a blockchain?

Functionally, to create a transaction, applications like a “wallet” are used to create a “public address” (derived from a “public key”) on the network, which indicates identity. An associated “private key” — which is kept hidden from the network — grants ownership.

Importantly, each transaction sent into the blockchain must:

  • be sent to a public address
  • be “signed” by a private key
  • include a transaction fee to be collected by the computers powering the network (or just “burned”, depending on how the blockchain is setup).

This process keeps the ledger growing in size indefinitely, and properly signals to the operators of the machines (i.e. “miners”) that transactions are worth processing. Without a sufficient fee sent with a transaction, miners won’t bother processing it (this helps prevent spam and denial-of-service attacks).

What happens if I lose my private keys or they get stolen?

Then unfortunately you are out of luck. By design there is no central authority to call and help you get it back. This is why various offline solutions for storing and securing private keys have become popular recently.

(As an amusing aside: while promising to disintermediate institutions like banks, ironically a growing number of people in the crypto world are storing some or all of their private keys within safe deposit boxes… in banks.)

What can blockchains do?

Blockchains have utility to disintermediate central authorities & disrupt corporations that silo data and hoard profit. With new protocols for things like payments, file storage, computing, lending, identity, reputation, etc… a boom in marketplace innovation is already happening:

  • Bitcoin, Litecoin, Ripple, etc… record financial transactions (threatening traditional “fiat” currencies).
  • Filecoin, Storj, Siacoin, etc… enable file storage (threatening cloud storage providers).
  • Steem, Golos, Leeroy, etc… pay people to make & discover social media content (threatening Twitter, Facebook, Snapchat, Instagram, etc… ).
  • And so on…

Of course, these are just examples of blockchains in each market. There are plenty of other currencies, file storage protocols, and social media platforms rising up as well (not to mention every other market vertical!).

Can blockchains be more complex than this?

Yes, they can. As an example, blockchains like Ethereum can store transactions that contain executable code. This means it is now possible to write a computer program on a decentralized network to move assets around based on pre-defined contract conditions.

In other words, anyone can now do things like make a will that distributes money over defined time intervals, create their own digital token to represent something (like an asset, part of an asset, or a club membership), or write a series of rules to govern a company (or even a whole country!).

Overall, my friend Conrad (@lisperati) recently wrote to me the following important observation:

“One of the things that makes blockchains different from databases (and makes ethereum different from a “computer”) is that blockchains are essentially nothing more than protocols: Anybody could write their own piece of software, connect it to the ethereum network, and it would process transactions exactly like ethereum, as long as the ethereum protocol is followed. Blockchains are a sort of ‘protocolization’ of computer software.”

What are coins?

Cryptocurrencies like Bitcoin, Litecoin, Ripple, etc.. that are recorded in transactions on blockchains indicating only a change in a numerical value (e.g. “X sends 1.34 to Y”) are “coins.”

What are tokens?

While it’s common to hear the word “coin” and “token” used interchangeably, the terms are importantly different. A coin is a relatively simple type of token, whereas a token can be — and often is — much more sophisticated than a coin.

Tokens are a complicated subject because not all of them are the same. The various types have different names being used currently, but — categorically:

  • Asset tokens represent ownership of an asset (e.g. a company, a venture capital fund, an expensive piece of art, etc…). An example is BCAP, which represents ownership in Blockchain Capital’s third fund.
  • Usage tokens (aka “Use”, “Protocol”, “Intrinsic”, or “Native” tokens) are used as transaction fees to write to a blockchain. Bitcoin is the most common example here.
  • Work tokens give owners permission to contribute, govern, and/or “do work” on a blockchain. An example would be Maker (MKR), which gives owners the ability to govern an organization that manages the stability of an underlying coin (DAI).
  • Hybrid tokens have properties of two or more of the above. Filecoin, for example, will be both a usage token (you need it to use the system), and a work token (you need to own some in order to provide file storage).

And to clarify some other potentially confusing terms regarding tokens and coins:

  • ERC20 tokens can be any token (most commonly created on Ethereum) that adheres to the ERC20 token standard of how to access information, transfer assets, and fire off events. This makes it easier for developers to create applications such as wallets that work with a variety of tokens.
  • App coin (or “appcoin”) is older term for a token used to fund a project. When you read articles from more than 6–12 months ago, you’ll often see this term. When you come across this term, just think “token.”
  • Alt coin (or “altcoin”) is an appcoin that powers it’s own blockchain. Any coin that does so besides Bitcoin (e.g. Ethereum) is, by definition, an altcoin.
  • Meta coin (or “metacoin”, or “colored coin” when using bitcoin) is a token that is built on top of a blockchain that already has an underlying coin. Nowadays, people just use the term token, which can be confusing when project owners generate and sell tokens in a self-described “Initial Coin Offering.”
  • And a Cryptocurrency, by the way, is any token recorded on a blockchain that is traded on a market as money.

Are all protocols created with an associated token?

No. To make you even more confused (or enlightened), protocols can be created with or without an associated token.

An example is what’s going on over at Augur:

“Augur combines the magic of prediction markets with the power of a decentralized network to create a stunningly accurate forecasting tool — and the chance for real money trading profits”

You should read more about Augur via their white paper, but for our purposes here, it’s important to know that they run their network with a protocol token (REP), which is a work token that allows owners to report on the outcomes of events that the market is attempting to predict. REP itself is a token with a tradable market value.

In addition, Augur leverages another, separate protocol (which you can also read about in their white paper) to handle the token transactions between buyers and sellers. This protocol, by design, is not associated with a token.

Thus, protocols and tokens are extremely open-ended and flexible.

What is an Initial Coin Offering (ICO)?

This buzzword — though catchy since it sounds like Initial Public Offering (IPO) — is a misnomer. Most “ICOs” these days are more accurately described as Token Sales or “Token Generating Events” (TGEs).

Practically, token sales are fundraising events that can happen (1) before a company launches, (2) while a product is being built, or (3) after a product has been in the market for awhile. Most token sales these days are done via ERC20 tokens on top of Ethereum.

A big list of past and upcoming token sales can be found here.

Soon we’ll see token sales be the default fundraising mechanism for new and existing companies. This will offer investors a more liquid asset (i.e. the ability to quickly buy things with the tokens or sell them). And, in addition to the capital they raised, founders retain the liquidity of the tokens they kept. This essentially results in a double-fundraising event!

This is an extremely powerful change to the status quo.

What are dApps?

Having discussed blockchains, protocols, and tokens, the final layer to the tech stack in the crypto-universe are consumer-facing platforms and applications. The term “dApp” refers to a “decentralized application”, which — long story short — means it’s built on top of one or more decentralized blockchains, protocols, and tokens.

A huge list of 645+ dApps can be found at https://dapps.ethercasts.com

What are Smart Contracts?

Made popular by the folks behind Ethereum, Smart Contracts (SCs) are executable code contained within transactions on a blockchain that execute predefined rules based on a set of conditions (i.e. “contracts”). SCs, therefore, are transaction protocols by definition; they move assets between parties reliably based on programmed instructions.

Indeed, new protocols in the form of smart contracts written to the Ethereum blockchain are created often. Examples are SCs baked into tokens that define the who, what, where, and when between token purchasers and generators. Many token sales these days are an agreement to deliver a token that has not yet been created (e.g. Filecoin’s recent token sale); the associated SC is programmed to distribute the new tokens appropriately once they become available.

What are DAOs?

DAOs are “Decentralized Autonomous Organizations” that are governed by Smart Contracts. DAOs should not be confused with “The DAO”, which did a token sale in May 2016 to raise ~$150m, of which ~$50m was hacked and stolen a few weeks later. This resulted in the shutdown of the organization and a fork of the Ethereum community after leaders decided to rollback the code and give the money back to the victims.

An example of an active DAO today is Maker, which — as mentioned above — grants MKR token holders the right to govern an underlying currency (DAI). The job of MKR token holders is to minimize the price volatility DAI against the IMF’s reserve asset, SDR. You can learn more about MKR and DAI in depth by reading their white paper here.

Big picture, the future of DAOs could be in autonomous organizations like, for example, a global fleet of self-driving cars using programmed rules to operate. Instead of being owned by a central authority, they could own themselves and use profits to add to their fleet, drive themselves to repair shops and pay for services, etc…

So yes, Uber & Lyft, be prepared for disruption by one or more robotic, AI-driven DAOs.

Author’s note: thanks in advance for any/all feedback, corrections, and comments to this article. Overall, it’s an amazing time ponder new protocols and solve interesting problems with decentralized infrastructure. Subscribe to my newsletter and I’ll let you know when I write more about blockchains, protocols, token sales, startups, and the occasional underlying science of health & fitness topics. Special thanks to Andrew, Dave, Mike, Tony, JC, Shay, and Conrad for reviewing this article and providing feedback! And last but not least, remember to comment below, and/or share this primer article with a friend if you’ve found it helpful. Thanks!

This was originally published here. 








Startup Financial Modeling, Part 4: The Balance Sheet, Cash Flow and Unit Economics

In the first three articles in this series, we looked at the big-picture motivation for startup financial modeling, why it’s important to start with your assumptions, and how to practically build your income statement and custom detail tabs. Today, we’ll finish off the series by examining how to construct the final components necessary to complete your model, including a quick discussion of unit economics and how to best keep your model updated. Feel free to ping us on Twitter (@wclittle, @troyhenikoff) with any questions.

Download a Copy of the Example Model

To make following along this final article easier, download a free copy of our example financial model here. Remember, understanding the nuances of your own unique business is critical, so only use this (and any other model you find online) as a guide while you build your own model from scratch.

The Balance Sheet

Create a new tab entitled “Monthly Balance Sheet” and set up the 60 month columns similar to the other tabs we created thus far in the series. The balance sheet is designed to give you a quick look at your company’s assets, liabilities, and equity situation (past, present, and forecasted).

screen-shot-2016-10-05-at-1-39-22-pm

You can see that, similar to how we rolled up numbers from custom detail tabs for the income statement, right out of the gate in cell C4 here we need to calculate the expected cash at the end of month 1 by creating a cash flow tab (which we’ll do below).

But first, in your downloaded model, take a look at the rest of the cells down column C to see how the numbers are calculated from the assumption values. While explaining the details of each formula is beyond the scope of this article, it’s helpful to Google for terms like Inventory Days on Hand and Accounts Receivable Days if you are unfamiliar with them and how they are used to calculate the applicable cells on your balance sheet.

While we left these two assumptions at 30 days for our example model for teaching purposes, take a look at what happens to your balance sheet when you drop the Accounts Receivable Days to 3, which may be more realistic for an eCommerce business with credit card payments and relatively low transaction values.

Cash Flow

Swinging back around to cell C4 (month 1 cash) on your Balance Sheet, now it’s time to create a new tab called “Monthly Cash Flow” and set it up appropriately for your business.  Here is our example (with added rows we didn’t model, for you to consider):

screen-shot-2016-10-05-at-1-39-36-pm

From top to bottom, notice how Gross Sales and Change in A/R (Accounts Receivable) affects your Total Receipts, based how your Accounts Receivable Days assumption value affects your balance sheet (i.e. with our assumed values, we are modeling collecting zero cash in the first month).  

Also notice how your Net Cash Flow from Operations is significantly less than your expenses, given that it takes time for money to leave your bank account to pay your vendors (i.e. our “Average Accounts Payable Days” assumption).

Finally, the Ending Cash Balance you calculate can then be used to populate your month 1 cash on your Balance Sheet tab.

Unit Economics

For many business types, it is extremely useful to create a separate tab called “Unit Economics” to calculate the forecasted Lifetime Value (LTV) of your unit (which is a “subscriber” in our example model) and Customer Acquisition Cost (CAC).

screen-shot-2016-10-05-at-1-39-49-pm

With our assumed 4% churn percentage, this means we can model that our average customer stays with us 25 months. Multiple these 25 months by the Gross Profit per subscription (which is also an assumption value) and that leads to our projected LTV of $213.59.  

Pulling in the CAC per outbound channel we are modeling (from our assumptions) allows us then to see how we are anticipating to tangibly grow the business. As long as the LTV/CAC ratio is (and remains) over 1, then your model is telling you that your business can grow.  Obviously, the higher the LTV/CAC ratio, the better.

Roll up your Income Statement, Balance Sheet and Cash Flow into Annual Summaries

Finally, to give you, your team, and your investors a helpful glance at what your projected five years look like, create separate tabs called “Annual Income Statement”, “Annual Balance Sheet”, and “Annual Cash Flow” and use formulas like “=SUM(‘Monthly Income Statement’!C4:N4)” to roll up 12 months worth of values from your monthly tabs.

screen-shot-2016-10-05-at-1-40-16-pm

And on your annual cash flow tab, for example, we suggest adding add a helpful row that reveals your lowest balance of the year:

screen-shot-2016-10-05-at-1-40-35-pm

Congratulations! If you’ve been following along thus far, you now have a complete financial model for your business. If you’ve done it correctly, you’ve most likely identified many assumptions that will need to be rigorously validated, and understood the key target assumptions that must be met in order for you to build a viable business.

Keeping your Model Updated

It’s hard to overstate the importance of reconciling your model to what actually happens month over month. This will help you drill down your assumption values and face the reality of the long-term health of your business.

Practically, be sure to keep a copy of your original model for reference. Then, after you’ve reconciled your books every month, open up your “active” financial model and (1) tweak the assumptions to get the current month close to actual, then (2) plug in the actual numbers in your monthly statements (i.e. replacing the formulas from your assumption values, which will remain in place for future months and now be more accurate).

Walking a potential investor through this model to show both the past performance and forecasted numbers in one place is, therefore, an immensely powerful tool.

This article was originally published over at Startup Rocket here. As a homework assignment, for those interested, see if you can add an assumption value called “Maximum marketing spend of Cash Balance” and model in this throttle to avoid burning through too much cash. This is an example of the types of custom features you can add into your own model.  

Finally, we sincerely hope you found this series helpful. Feel free to ping us on Twitter (@wclittle, @troyhenikoff) with any questions. You can also keep up Will’s writing on early-stage startup life by subscribing to his newsletter here. Thanks!

Get a head start on your own entrepreneurial journey. Apply to an accelerator program. Applications close on October 15th.








Startup Financial Modeling, Part 3: The Income Statement and Custom Detail Tabs

In the first two posts in this series, we examined why building a financial model for your startup is important and how to practically get started with your assumptions tab. Today, we’ll continue by diving into the income statement and supporting tabs used to calculate your projected revenue and expenses. The final part of this series will go over cash flow, the balance sheet and keeping the model updated.

But first, in case you aren’t yet convinced that it’s worth your time to build a financial model for your business, here’s a quick video we put together showing how a model can be used to gain insight about what assumptions drive forecasted cash most significantly:

We’ll discuss these types of learning nuances in more detail in our next (and final) post in this series. But first, we have some work to do in order to fill out the model.

Start by Building your Monthly Income Statement

To dive right in, assuming you’ve already created your first tab called “Assumptions”, create a second tab called “Monthly Income Statement” and begin to set it up like this:

screen-shot-2016-09-28-at-11-41-58-am

Copy that formula in D2 and paste it into the next 58 columns in row 2 in order to set up 60 months (i.e. 5 years).

Next, we’ll focus on revenue. Starting in row 4 you’ll want to list the high-level revenue streams for your business.

For example, with Dollar Cave Club (the demo business we introduced in the last post), we are planning on four revenue streams:

  • Subscription
  • Add-ons
  • eCommerce
  • Advertising

We’ll write these in cells B4-B7, and in order to begin filling in cells C4-C7, we’ll need to ask ourselves how will we calculate our forecasted revenue for month one. This, as you may suspect, requires starting a new tab.

Sales Detail

Create a new tab called “Sales Detail” and, from top to bottom, fill in the rows needed to roll up calculations for each revenue stream. This is where your spreadsheet will become unique for your business. As a guide, however, here’s what it looks like for Dollar Cave Club:

screen-shot-2016-09-28-at-11-42-21-am

Note that we are using the necessary sub-rows to calculate the total for each revenue stream per month (i.e. rows 11, 15, 19, and 25).

From top to bottom in month 1 (i.e. column C), you’ll want to start with your actual numbers (or assumptions if you don’t have numbers yet) and fill in each cell appropriately with formulas using your assumptions, and – you guessed it – any additional custom detail tabs you need.

Right out of the gate here with our example business, in order to calculate the number of New Subscribers in month 1 (i.e. cell C4), we’re going to need to create another custom tab around traffic numbers to handle this calculation.

Traffic Detail

In order to calculate the number of new subscribers in month one, we’re going to need to add up anticipated traffic numbers from paid search (SEM), organic search, Facebook, direct, etc… (i.e. it will, again, be custom for your business).

Here’s what it looks like for Dollar Cave Club:

screen-shot-2016-09-28-at-11-42-35-am

If you recall our initial Assumptions from the last article, here is where you’ll pull in variables like AverageCPCGoogle (for C4’s formula), AverageCPConFACEBOOK (for C7), VIsitsPerFBPicturePost (for C10), VisitsPerFBVideoPost (for C13), DirectTrafficMultiplier (for C15), and OrganicTrafficMonthlyGrowthRate (for D16 and beyond), etc…

Importantly, after a few months of testing, we’ll assume in this model that we’ll get to SteadyStateMonthlySEMSpend (i.e. $10k) and also apply that to our Facebook channel (see G3/H3, and G6/H6). This, for the purposes of our demo model, will then carry forward through our 60 months.

screen-shot-2016-09-28-at-11-42-52-am

Back to Sales Detail

Now that we have our Traffic Detail filled out, we can go back to our Sales Detail and finally fill out C4 there (i.e. New Subscribers).

That formula will look like:

=ROUND(‘Traffic Detail’!C4*SEMToSubscriber + ‘Traffic Detail’!C7*FacebookToSubscriber + SUM(‘Traffic Detail’!C15:C16)*OrganicToSubscriberConversion,0)

…which will then be applied to the rest of row 4 and allow us to, using the proper assumption values, calculate our Lost/Ending Subscribers and ultimately our Subscription Revenue, which is what is then rolled up to our Monthly Income Statement for the C4 cell that started us down this rabbit trail to begin with.

Thus, to summarize, the blue values below are taken from our assumption values (inputs) and the formulas are calculated like so:

screen-shot-2016-09-28-at-11-43-14-am

The other revenue stream totals can then be calculated similarly from our traffic detail tab and assumption values, which allows us to calculate Gross Revenue numbers on our income statement.

Back to the Monthly Income Statement

Now that we have the Sales Detail tab filled out, we can roll those numbers into our forecasted revenue cells like so:

screen-shot-2016-09-28-at-11-43-30-am

And there we have it. Copy the formulas across the 60 columns and total them up into your Gross Revenue row.

Now we can turn our gaze to costs.

Direct Costs

The bottom rows of the Income Statement will then examine the costs associated with running our business, and – similar to how we worked it for revenue – we’ll need to create custom tabs when appropriate.

For direct costs associated with the stuff we sell (versus indirect costs like salary and rent and such), we can add them in like so:

screen-shot-2016-09-28-at-11-43-40-am

Note the formula for C11, which adds our assumed base cost to a scale factor applied to our traffic that we pull in from our Traffic Detail page. Our Subscription and eCommerce values are then calculated from our SubscriptionMargin and eCommerceMargin assumptions.

We can continue then by tallying our Gross Profit (i.e. C8 minus C14), add in our direct costs from Google and Facebook ads (pulled in from the Traffic Detail tab), and calculate our Contribution Margin (i.e. revenue minus direct costs).  

screen-shot-2016-09-28-at-11-43-52-am

Indirect Costs

Finally, we’ll finish up the costs part of our income statement with some standard indirect costs for example purposes:

screen-shot-2016-09-28-at-11-44-06-am

Since these costs are calculated from the number of employees we expect to have, and the details of their salaries and benefits/taxes, this warrants the creation of a new tab.

The Employees Tab

Create a new tab called – you guessed it – “Employees”, and add in the details that are custom for your business. Like any detail tab, it really can be formatted however you’d like as long as you can send back up the numbers needed, but here is an example of how we did it for our demo company:

screen-shot-2016-09-28-at-11-44-17-am

While most of this tab is pretty straight forward (and obviously the salary numbers will adjust based on your market), one thing to note in here is how we defined “Warehouse help” to increment based on volume, like so:

=ROUND((‘Sales Detail’!C7+’Sales Detail’!C17)/OrdersPerWarehouseHelpPerson,0)

In other words, as sales increase, we’ll add warehouse help based on our assumed OrdersPerWarehouseHelpPerson value. With our current assumption values, this starts to increment in month 22, but you can see how if things pick up, we’ll need to hire this person earlier.  

Finishing with EBITDA and Notable Mentions

Finally, to finish off our income statement, we’ll add in:

screen-shot-2016-09-28-at-11-44-35-am

For those unfamiliar with the term, EBITDA stands for Earnings before interest, taxes, and amortization, which is a sufficient number to end with here for financial modeling purposes.

You can also add in notable mentions like the number of employees, capital raised, important KPIs your are tracking, etc…

Follow the Principles, Don’t Just Copy What you See

We’ll of course provide a link to the full demo spreadsheet at the end of this series. While you can (and should) study other financial models to use as additional guides, it’s critically important for you to understand the principles behind what is going on here.

In short, start with the Monthly Income Statement and then go row-by-row, cell-by-cell and create custom detail tabs (and custom sub-detail tabs) as needed to perform the necessary calculations to keep your income statement clean.

 

This article was originally posted on Startup Rocket here. In the next and final post in this series, we’ll conclude by building out the monthly cash flow and balance sheet, including annual roll-up tabs and a discussion of best practices to keep your model updated. Subscribe to Will’s newsletter to get notified when the final article is up. Feel free to ping us on Twitter (@wclittle, @troyhenikoff) with any questions.








Startup Financial Modeling, Part 2: Start with Your Assumptions

In the previous post of this series, we described what financial modeling is and why it is important for startup founders to build their own models from scratch. Today, we’ll begin by diving into how to practically start building a financial model. In coming weeks, we will also be covering the income statement as well as cash flow, balance sheet and keeping the model updated.

Good financial models are built from the bottom up. They have assumptions that flow into backup sheets that flow into monthly statements that flow into annual summaries. You literally start with the smallest component and start building up.

What is an Assumptions Tab?

The assumptions tab should be the first tab within your spreadsheet and contain variables that will be referenced from other tabs. Rather than manually entering data into your income statement, you’ll roll up variables across your assumption and detail tabs. The ONLY place you should ever type a number is into an assumption cell. Every other cell is a calculation based on the assumptions!

The types of assumptions you’ll want to define are unique for your business type. Most commonly they are related to expected revenue from each product/service you sell, expected costs, initial investment dollars in the bank, etc…

Where to Get your Data

Don’t just randomly guess your expected revenue and costs. Do your homework. Often industry averages will be published online for you to look up, or you can ask people in your network who are familiar with your business and market.

In addition, study your competition carefully. The obvious low-hanging-fruit research is if they have publicly available annual reports where you can glean specific information about cost of goods, productivity of sales reps, salary info, etc… But even if the data isn’t directly public, very often helpful numbers are posted in blogs or articles online.

But be careful – just because you find out that a large public company spends 12 percent of its annual revenue in marketing, does not mean that is the right ratio for you. Most startups spend money in marketing long before they have ANY revenue, it is how they intend to get revenue – very different!

 

Working with an Example Company

We’ll use an example company called Dollar Cave Club:

Dollar Cave Club is the latest in a series of companies that make recurring purchases easy and painless. Sign up for our $29.99 monthly subscription and we will send you a box each month with essentials for your Man Cave. One month it is a classic poker set, the next an amazing set of beer steins. If you cannot wait to get your favorite dart board, you can come to the site directly and buy anything we sell.

We get most of our new customers from SEO, advertising on Google and Facebook. We have such a highly desirable audience that many people want to advertise to our visitors on our site as well.

Subscribers can “add-on” to their monthly subscription by selecting any items in our catalogue and shipping is free! Our churn has been around 4 percent per month and our traffic to the site is growing as well as our sales.

Here are things we know about the company:

  • Subscription Price $29.99
  • Churn = 4 percent Monthly
  • Something about SEO, SEM and Facebook
  • Add-ons to subscriptions
  • Some advertising revenue

Here are things we don’t know about the company:

  • What impact does selling ads on our site have?
  • What impact will increasing Ad spend have?
  • Can we make Ad spend more realistic?  How?
  • What if we raise more money?
  • What impact does conversion rate have on the business?

 

Start by Filling in your Revenue-related Assumptions

We’ll start the assumptions tab by filling in our best guess of every aspect of our company related to revenue (obviously you’ll add/remove things as you go, but this is a good place to start):

  • Subscription Price:
    • $29.99
  • Churn on subscriptions per month:
    • 4 percent
  • Page views per visit:
    • 3.5
  • CPM growth per month on ads sold
    • 10 percent
  • Max CPM on ads sold
    • $1.50
  • Steady State Monthly SEM Spend
    • $10,000
  • Steady State Monthly Facebook spend
    • $10,000
  • SEM -> Subscription Conversion
    • 2.00 percent
  • Facebook -> Subscriber Conversion
    • 1.00 percent
  • SEM -> eCommerce Conversion
    • 4.50 percent
  • Facebook -> eCommerce Conversion
    • 2.50 percent
  • Organic Traffic -> Subscriber Conversion
    • 0.20 percent
  • Organic Traffic -> eCommerce Conversion
    • 0.50 percent
  • Organic SEO Traffic monthly growth rate
    • 10 percent
  • Direct Traffic Multiplier
    • 33 percent
  • Average CPC on SEM (Google)
    • $2.25
  • Average CPC on Facebook
    • $1.85
  • Visits per FB Post with Picture
    • 5
  • Visits per FB Post with Video
    • 25
  • Subscription Margin
    • 25.00 percent
  • eCommerce Margin
    • 30.00 percent

 

Next, Fill in your Cost Assumptions

Similar to revenue assumptions above, now you should list out common monthly costs like IT infrastructure, rent, legal, taxes, travel, etc…

  • Internet Infrastructure base cost / mo
    • $500
  • Incremental for each 1 million visits
    • $250
  • Orders / warehouse help person
    • 5000
  • Taxes and benefits load
    • 25 percent
  • Rent per employee per month
    • $500
  • Tech Cost per employee per month
    • $100
  • Legal and Accounting Cost per month
    • $500
  • Travel cost per employee per month
    • $250
  • Training Budget per employee per month
    • $100
  • Recruiting per added employee
    • $5,000

 

Finally, List out Cash and Operations Assumptions

For example:

  • Opening Cash Balance
    • $1,000,000
  • Raise at Month 13
    • $2,500,000
  • Average Accounts Receivable Days
    • 30
  • Average Accounts Payable Days
    • 60
  • Inventory Days On Hand
    • 30

 

Assign Descriptive Labels for your Assumptions Before you Use Them

For each of your assumptions, use the feature your spreadsheet software has to label your variables. Just like in programming, make your variable names as descriptive as possible so your model can be more easily read.

For example, here’s where we defined “Page views per visit” (PagesPerVisit):

screen-shot-2016-09-19-at-4-47-13-pm

Now whenever we refer to page views per visit in the rest of our spreadsheet, we can use the label PagesPerVisit rather than the ambiguous ‘Assumptions’!B18.

 

Controlling the Size of your Assumptions List

As you build out the assumptions tab for your business, it will be tempting to create tons of variables that you may or may not need as you build out the rest of your financial model. Therefore, it is best to stick with a core set of revenue and cost assumptions, and add in more when you realize you need them. Just remember, again, to assign your label names before you include the variable in an equation anywhere in your spreadsheet (a common Excel gotcha).  

This article was originally posted on Startup Rocket here. In the next post in this series, we’ll continue building our example financial model by diving into the income statement and custom detail tabs. Subscribe to Will’s newsletter to get notified when the next articles are up. Feel free to ping us on Twitter (@wclittle, @troyhenikoff) with any questions.








Startup Financial Modeling, Part 1: What is a Financial Model?

screen-shot-2016-09-13-at-11-37-55-am

This series is the result of a friendly debate I had recently with Troy Henikoff (former Techstars Chicago Accelerator Managing Director) regarding the best approach for founders to take when building a financial model. More accurately, the “debate” was a strong adverse reaction from Troy after I shared a template I built for Prota Ventures’ portfolio companies. His feedback was, essentially, to never use a template and instead build each model from scratch.

He invited me to a 90-minute lecture he gave where he overwhelmingly convinced me and the room that, indeed, founders need to take the time necessary to build their models from scratch. After I asked him where I could find his lecture material online, he suggested we co-author this article series since there weren’t many solid resources available. We sincerely hope you find this series helpful. 

 

Here’s What We’re Planning to Cover

Our plan is to break this out into a four-part series and guide you through the components necessary for building your own financial model from scratch:

 

What is a Financial Model?

In short, a financial model is an abstract mathematical representation of how a company works (and more importantly, how it will work going forward). The model has inputs and outputs. The inputs are the assumptions that drive the model, things like what drives your customer acquisition cost, what your churn rates are, how much you pay people, etc. The outputs are a set of projections that show how the company will perform if the assumptions are true. One model can produce multiple sets of projections given different assumptions.

Based on a set of assumptions, a financial model is used to make smart decisions (e.g. how many sales people to hire and what to pay them). The model includes financial projections that are tied mathematically to the assumptions, which allows operators to “play with the variables” in order to understand how certain decisions might affect the future health of their company.

 

Why Should Founders Care about Building a Financial Model?

Troy has an important story to share on this topic:

“When fundraising for SurePayroll, we had some very high level financials in the pitch deck. Inevitably, VC’s would ask where the numbers came from. I would tell them that we had a very detailed financial model that drove it, I was setting the bait…

They would ask to be sent a copy of the model and I would refuse. I would only share it by first sitting down with them and an associate and reviewing the model in person and after that 90 minute session, I would leave them a copy of the model to play with further.

They would insist that they could figure it out without the meeting, but I ALWAYS held my ground. I wanted the meeting not just to save them time and frustration learning a new model, but more importantly to get more face time with them in a situation where I was going to shine.

I knew the model inside and out since I built it; I could answer any question about any cell and look like a genius. In the end, I did eight of these meetings and EVERY ONE of the firms that did the 90 minute meeting with me on the financial model either made an investment in the company or made an offer to invest in the company. Every single one.”

 

Why is it Important for Founders to Build a Financial Model from Scratch?

While it’s easy to search around and find a template to use, those templates were built by someone with a particular business in mind. Since every business is unique, this will lead you into trouble.

While it’s often helpful to learn from other people’s models to ensure, for example, that you aren’t missing anything important, you should never build your model using their template. You’ll end up banging your head against a wall when you need to change things, and you’ll inevitably be confused about some nuance that will come back to haunt you since you don’t understand it.

In other words, while you may think that a template will help you save time, what you are actually doing is acquiring “technical debt” that will end up costing you more time in the long run.

Plus, it’s critical to understand every column, row, cell and tab in your spreadsheet for two key reasons; it will help you better manage your business, and when the time comes to explain it to an investor, you’ll be able to explain exactly how it works and increase your odds of landing funding.

 

What are the Components of a Solid Financial Model?

Since most people are using the financial model to communicate projections to investors, it is critical that you speak the investors’ language. They are used to having financials in Excel, so you should build your model in Excel.

Google sheets is convenient for making changes and having multiple people editing, but sending an investor a model in Google sheets signals that you are not financially savvy.

Investors are also used to seeing three standard statements; an income statement, a balance sheet, and a statement of cash flow.  Each of these is more credible if it has BOTH the past performance and the future projections in the same spreadsheet.

Your spreadsheet should contain a tab for each of these outputs along with an “assumptions” tab and custom detail tabs needed to help calculate the main outputs. We’ll walk through a specific example later in this series so you have a better understanding of what this should look like.

 

Profit is a Matter of Opinion, Cash is Fact

Because of various accounting nuances – such as fixed asset depreciation and deferring revenue – if you assign ten accountants to finish your books at the end of the year, you’ll get ten different answers for how much profit (or loss) you had in the year. While hopefully not far off from the others, each will have a slightly different report of your “profit” based on their accounting opinions.

However, the balance of your bank account is a specific number to point at; it’s a fact that your ten accountants should agree on.

Therefore, it’s important to remember that your financial model will have your own opinions baked in regarding your profit. This means that examining your cash flow carefully as you fine-tune your business assumptions is critical.

 

Why Investors Care about your Financial Model

Having a solid financial model is a significant step in communicating to investors that you are a logical thinker with a defensible plan and clearly understand your business and the levers that drive it.   

Nobody expects your model to be perfect, as a matter of fact, when we present a model, we always open with the same line:

“The only thing we know for sure about this model is that it is wrong. But, if we look critically at it we can better understand the drivers of the business and what we need to be focused on to reduce our risk.”  

Keep in mind, investors are looking for the big home runs, but they are also looking at reducing their risk. The model can help them get comfortable with the risk.

 

– – –

This article was originally published over at Startup Rocket here. In our next post in this series, we’ll dive in a step-by-step guide of how to build a financial model, starting with the assumptions tab. Subscribe to Will’s newsletter to get notified when the next articles are up. As we mentioned above, feel free to ping us on Twitter (@wclittle, @troyhenikoff) with any questions.