Financial projection is an integral part of any
business plan, but for a new business you cannot use historical trend to come up with a financial projection. How do you solve this dilemma?
2. Financial projection for a
completely new product
● Financial projection is an integral part of any
business plan
● If it is an existing line of business with
known trend, it is not too difficult to project
the trend into future
● But if it is a completely new product or
service with no equivalent to benchmark
with, what do you do?
● Is linear trend good model to use?
3. Diffusion of Innovations
● In "Crossing the Chasm", Geoffrey A. Moore
builds on the Diffusion of Innovations
concept to define a market adoption model
for hi-tech products
● The market adoption curve follows an S
shape
● More suitable than linear models because
○ it models the early slow adoption period
○ it models the hypergrowth period
○ it models the market saturation period
4. Linear vs. S curve
Comparison ● Problems with
Linear
○ too aggressive in the
early period
○ too conservative in
the hypergrowth
period
● Linear is simpler to
model, but S curve
can be modeled as
Logistic functionLinear model overstates
early period
Linear model understates
hypergrowth period
5. Modeling Adoption in Excel
Market penetration = Saturation/(1+Curve^
((Hyper+Stable/2-t)/Stable))
Link to the spreadsheet
Used only for
parameter
estimation
Used only for
parameter
estimation
minimize this
in Solver
Determines the penetration %
at the start of hypergrowth.
● if 5%, set to 360
● if 10%, set to 81
● if 20%, set to 16
6. Example
Hypergrowth starts in Year 3 at 5% penetration,
ends in Year 5
● Build this up with
estimated market size
● Pricing model
● Units sold/subscribed
● etc.
And you have a financial
projection for your
business plan!
7. Parameter Estimation
If you have an actual data, you can use Solver in
Excel to estimate parameters
● Install Solver
● In the Solver dialog: ● Set Target Cell: select the cell
for the sum of squared errors
● Equal To: Pick "Min" radio
button
● By Changing Cells: select cells
for Saturation, Curve, Hyper,
Stable (or subset)
● Click "Options" and check
"Assume Non-Negative"
8. It is only as good as
assumptions you make
In reality, the actual adoption never follows the
model faithfully
● Use the model to develop assumptions and
test various scenarios to refine your
assumptions and models
● it is a good idea to come up with a three-
point estimation of the best case scenario,
the worst case scenario and the most likely
scenario.