Operational Model (Tue Nov 12, lect 19) | previous | next | slides |

How to actually build a financial model

Reminder: Readings and videos are your responsibility. You will be expected to come to class prepared, having read or viewed the material, and ready to participate in the discussion

Overview

  • Information is organized logically into several sheets
  • Columns are units of time - usually months
  • Overall span is 1 (or more) years
  • Rows are variables of various kinds
  • Use color to distnguish between input vs computed cells

Operating Model for your business

Financial model

  • A model is a simulation of the future
  • It is not the same as an analysis of the past
  • Although they are closely related

Building a financial model

  • Not a universal recipe (none can exist!)
  • These notes provide one set of guidelines and rules of thumb
  • Your mileage may vary!
  • Use common sense in your thinking.
  • You are not google or facebook so your model should reflect that.

What do you want out of the model?

  • You can think of your financial model as an MVP
  • What are the hypotheses that are being tested?
  • How much money do we need and when?
  • How many people do we need, and when?
  • How many customers do we need, and when?

General Structure of a Financial Model

  • Use google sheets (excel is ok too)
  • Columns will be time.
  • It depends, but a good start is to have 1 year in months (and perhaps additinal years in quarters)
  • Rows will be various kinds of metrics

Cells: Types of information

  • Some are inputs to the model (“assumptions) and some are outputs (“results”)
  • Don’t be uncomfortable about very broad generalizations for your assumptions (mere guesses)
  • Most of the time the best you can do is an educated guess
  • Assumptions should be clearly set off
  • Color coding can be helpful

Assumptions - Input parameters or independent variable

  • Input to the model (e.g. Cost to make one unit)
  • Basis of testing scenarios
  • Sometimes an input parameter is a number, a percentage, an amount of money
  • Sometimes its a varies over time (i.e. Number of employees in Q1, Q2, Q3, Q4)
  • The key is that it is not computed by the model, it is supplied by you

Output parameters, dependent variables or results

  • Output from the model (e.g. Operating Profit in Q2)
  • There will be lots of assumptions.

Customers or units and the Growth Engine

  • A very typical “driver” for your model will be the number of customers served or units sold
  • Your assumptions there will have a big effect on the overall model
  • Consider where your new customers will come from
  • Is it through search in which case you will spend money on google or facebook ads?
  • Is it inherent in the use of the product, then you should be able to explain how that would work

Pricing

  • As you know, pricing models can become complicated
  • You need to make a judgement call about how much of it you try to model at this stage
  • Annual vs. Monthly subscription
  • Single purchase (what about repeat purchase?)

People Costs

  • It is very common that salaries and overhead are the biggest costs for a startup
  • You can choose to enter number of people “by hand” for each month
  • Or make it a function of something else, like units sold or number of customers

Milesones

  • Most startups go through a few phases from Day 1 to having customers and/or having funding
  • E.g. Pre-revenue, Post-revenue, Pre-Funding, Post-funding
  • It is useful to build your model with that in mind

Charts (graphs)

  • The classic chart shows time in months (like the model) against:
  • Money in the bank
  • Number of customers
  • Number of units
  • Revenue
  • Expenses
  • Profits

Thank you. Questions?  (random Image from picsum.photos)