Lecture 7

Spreadsheet Modelling

Dr Fui-Theng Leow (FT)

[email protected]
 Class activities

2.00 – 2.15 pm – Reminder: Templates, strength (lecture 6)

2.15 - 2.50 pm – Lecture 7: Spreadsheet Modelling
• Types of spreadsheet model
• Building a Spreadsheet Model
10-minute break
3.10 – 3.30 pm – Lecture 7: Spreadsheet Modelling
• 10 rules to guide spreadsheet modelling
• 5 good spreadsheet modelling practices in Excel
3.30 – 3.50 pm – Live Q&A session

2
 Types of Spreadsheet Model

Prescriptive model Descriptive model Deterministic model -

prescribes a course of predicts the number of all inputs and outputs
action, e.g. to propose the participants/hours the are known.
total cost of the workshop can host based
Assuming that we know
workshop. on the allocated budget.
the unit cost and the
Input (e.g., cost, quantity) Output (e.g.. total amount demand, so we can find
is known. of allocated budget) is the result-maximising
known. option (e.g., optimising
the value of the software
license by maximising
its usage during the
workshop).

3
 Building a Spreadsheet Model (Decision-relevant)

Organisations often use a spreadsheet model to aid in decision-making, or to determine

how changes in the external environment will impact their success.

Situations which a spreadsheet model is useful:

• Should Bakery A open a new shop to sell its products?
• What price should Bakery A charge for a pastry?
• What percentage of revenues should Bakery A spend on advertising?
• How many bakers does Bakery A need to hire to be able to keep up with projected
demand for its products?

4
 Building a Spreadsheet Model (Decision-relevant)

Decision-relevant Spreadsheet Model

Q: What is the decision?

A: Whether or not to launch the new line of tire swings.
Q: What are the decision alternatives?
A: Option A. Launch new line of tire swings / Option B. Do nothing.
Q: What are the decision criteria (condition)?
A: The product line must be profitable.
Q: What is the system (formula) that converts inputs into outputs?
A: start at the decision criterion (profit) and work backwards to the inputs.
Profit = revenue (price × quantity sold) – costs (cost of goods sold + fixed costs)
5
 Think & Apply: Individual Assignment

What are the decisions to make when planning the budget for organising an IT
skills enhancement workshop (for the employee of your selected company)?
1. Venue:
 Location: host the workshop in-person? Online? hybrid event?
 Cost for equipment rental, room rental, furniture rental
2. Salaries:
 Fees for hiring trainers
 Cost for travel and accommodation for trainers
 Salaries for staff who handling the administrative tasks
3. Training materials:
 Copyrighted content
 Software Licenses: any specific software used in the workshop?
 Cost of printing workbooks, handouts, or other hardcopies
6
 Think & Apply: Individual Assignment

What are the decisions to make when planning the budget for organising an IT
skills enhancement workshop (for the employee of your selected company)?

4. Technology & Infrastructure:

 Online platforms subscription fees (virtual meetings, polls, quiz, etc. )
 Videography, recording and editing services
5. Participant Engagement:
 Costs for creating workshop kits or goodie bags
• Cost for certificates and awards
6. Provide refreshments or meals to participants?
7. Logistics & Travel:
• For in-person events, provide transportation and accommodation for participants?
7
 Think & Apply: Individual Assignment

What are the decisions to make when planning the budget for organising an IT
skills enhancement workshop (for the employee of your selected company)?

8. Post-event Evaluation:
 Survey Tools: using paid tools for feedback collection?
 Analytics: using paid tools to measure the success of the workshop?
9. Sponsorships and Partnerships:
 partner with organisations or get sponsorships to offset some costs?
10. Revenue Generation:
• Selling tickets to other participants (non-employees)?
• Consider selling workshop-related merchandise (i.e., T-shirt)?

8
 Building a Spreadsheet Model

Each quantity involved in a spreadsheet model can be classified as an output, input,

decision variable, or calculation.

Model output
- The quantity/result computed by the model.
- A model can have multiple outputs.
- The expected time a customer spends waiting in line at a bakery.
- The annual profit and market share from a new bakery.
Model input
- An assumption / a key quantity needed to compute a model output.
- Models often have many inputs.
- An annual growth rate for units sold for a bakery
- The average number of customers arriving at the bakery per hour 9
 Building a Spreadsheet Model

Decision Variables
- The quantities that can be controlled by the decision maker.
- The bakery owner can control price, the number of staff serving the customers.

Model Calculations
- The set of relationships (Excel formulas) that are used to transform model inputs
and decision variables into outputs.
- The formula to determine how customer waiting time depends on arrival rate and
service rate (the number of staff).

10
 Building a Spreadsheet Model (Example)

Harold owns a playground fabrication company called Play Inc. He is considering

whether or not to expand his business with a new product line of tire swings shaped like
animals (e.g., a horse).
He expects to sell each swing at a price of $80. The materials to make each swing will
cost $10. He expects that it will take three hours of labour to make each swing at $15 per
hour. Storing tires requires a special storage facility with special insurance because of the
risk of fire. Harold estimates that the storage facility and insurance will cost $1,200 per
month.
In the first year, Harold expects to sell 500 units. He expects sales to be fairly constant at
750 units every year after that.
Develop a spreadsheet model to assist Harold in deciding whether to launch this new
product line.

11
 Building a Spreadsheet Model (Example)
 Fatima wants to begin planning for her retirement. She wants to determine how much she
will have saved for retirement by depositing $1,000 per month, every month for the next
30 years, at a monthly interest rate of 0.6 per cent.

 Build a spreadsheet model that Fatima could use to calculate her retirement savings in
30 years with inputs that include Monthly Deposit, Monthly Interest Rate, and Months
Until Retirement.

12
 Building a Spreadsheet Model (Strategies)

Strategies for organising the spreadsheet model

 State inputs and assumptions (to provide information relevant to the decision)
 To prepare the budget sheet for cost comparison, for a decision to reduce the cost
of printing for publicity, you need to information of paper type, paper size, print
item, print quantity, supplier, etc.
 Prepare the dataset as the input
 Separate model inputs from analysis
 Leave the dataset in its original sheet (for reuse purpose, for Vlookup purpose)
 Create the item list
 Perform analyses in steps
 Design model layout – other users can understand and use (labels, notes, messages)
13
 Think & Apply: Individual Assignment

To provide information relevant to the decision, what are the inputs/attributes

(columns in the datasets) you need?
• Quantity of tickets sold
• Price per ticket
• Sponsor name
• Sponsorship amount
• Type of merchandise (e.g., t-shirts, mugs)
• Quantity sold
• Price per item

And (provide more attributes on Mentimeter)?

14
 Think & Apply: Individual Assignment

How to prepare the datasets?

Method 1 - use the sample datasets

 Refer to “Sample dataset for individual assignment.xlsx” file

Method 2 – download the datasets from the Internet

 Free Example Data Sets For Spreadsheets

\Method 3 - create random data in the worksheets

 How to Generate Random Numbers in Excel
 Random Numbers

15
 Building a Spreadsheet Model (Process)

STEP 1: What do we want to use the model to accomplish?

STEP 2: To accomplish the goal, what output quantities must be computed?

STEP 3: What inputs and decision variables need to be determined in order to compute
the needed outputs?

STEP 4: Identify the relationships between the inputs, decision variables, and outputs to
determine the calculations/formulas.

STEP 5: Set up a spreadsheet that connects the inputs, decision variables, and outputs.

STEP 6: Make the model to present the decisions for actions.

STEP 7: Identify the model's limitations (e.g., factors that the model does not consider)

16
 Building a Spreadsheet Model (Ideas)

Ideas that you can use in building your spreadsheet model?

1. The cover page (e.g. with a short note, logo, how-to-use instructions, etc)
2. The list of worksheets with links
3. Spacing and gridlines
4. Section headings (e.g. highlight the name of each category)
5. Use a different colours for input data (e.g. blue); calculated data (e.g. black)
6. Provide instructions for how users can use the worksheet (e.g. what are the cells that
need an input data from the user)
7. Display a timeline, or dates, timestamps, days
8. Formatting, including the decimal places, currency symbols, font style
9. Use Functions and formula to generate the result based on the input data
10. Flexibility - allow users to select a scenario (e.g. 1, 2, 3) to present the data model

17
 10 Rules to Guide Spreadsheet Modelling

Rule 1 - Spreadsheet inputs and decision variables should never be hard-coded in

formulas; inputs and decision variables should always be entered in their own cells.

Rule 2 - A common layout structure is to place model inputs and decision variables on
the top of a worksheet. The model outputs may be placed below the calculations or to
the right of the model inputs and decision variables.

Rule 3 - use different colours to denote each type of building block, an example the
color-coding convention: Yellow for Inputs; Blue for Decision Variables; Orange for
Calculations; Gray for Outputs

18
 10 Rules to Guide Spreadsheet Modelling

Step 4 – (Optional) If you have trouble

determining how your spreadsheet inputs
and decision variables can be used to
calculate model outputs, you can use
influence charts to develop an
understanding of a model's key formulas
and relationships, e.g. what are the
quantities needed to calculate profit.

19
 10 Rules to Guide Spreadsheet Modelling

Rule 5 - Use range names to make formulas easier to

understand, it is easier to understand to =price*demand
than =A50*B100.

Rule 6 - You can place comments (or an arrow and text

box) in cells to help the user better understand complex
formulas.

Rule 7 – Use the Copy command (or Auto-fill command)

to efficiently create formulas, make sure the use of
absolute and relative addressing (the $ signs) is correct.

20
 10 Rules to Guide Spreadsheet Modelling

Rule 8 – determine how sensitive our final decision is to changes in each input, e.g.
assuming that with an annual sales growth of 15%, it would be a good decision to build a
new bakery. But how bad a growth rate could be before opening the bakery would be
a poor decision.

Rule 9 - Models need to be flexible so that they can easily incorporate changes, e.g. the
model should allow the products to be added or subtracted from the list; the model should
enable us to vary the number of years of operation; the formulas and charts can
automatically be updated to include the new data.

Rule 10 - It is important to understand a model’s shortcomings and limitations

21
 5 Good Modelling Practices in Microsoft Excel

1. Clearly state inputs and assumptions

 Treat every numerical value (parameter) that
contributes to the result as a separate input.
 Each input should be listed in a separate cell
with a label.

2. Document your sources and assumptions

 Place the reference or assumption beside
input values that have been found through
another source, or calculated from other data
 This will ensure that other users of the model
know how you arrived at the assumptions
you are making.

22
 5 Good Modelling Practices in Microsoft Excel

3. Separate Model Inputs from Analysis

 Do not put any numbers directly into the
intermediate calculation formula (this is
called hard-coding)
 Hard-coding a model makes debugging
a model more difficult, because
changing them can only be done
manually
 List a value as the parameter in the
inputs section of the spreadsheet and
link to it using cell reference

23
 5 Good Modelling Practices in Microsoft Excel

3. Separate Model Inputs from Analysis

 Add explanatory text / inserting a
comment box can help communicate
your model to another users.

24
 5 Good Modelling Practices in Microsoft Excel

4. Reliably perform analysis

 A good spreadsheet model must be reliable - perform calculations without error.
 Model reliability can be improved and evaluated in a number of ways:
 Break down complex formulas into multiple intermediate calculations
 Assess the reliability of intermediate calculations as the model is being
developed
 Evaluate Face Validity - When you make small changes to the input values,
verify that the results are in the order of magnitude that you might expect
 Test Simple Inputs – use inputs with a known outcomes to ensure that the
expected result occurs.

25
5 Good Modelling Practices in Microsoft Excel

26
 5 Good Modelling Practices in Microsoft Excel

5. Design the model to enable understanding and use by other users – a good
model can be used with relative ease by a non-expert user.
 Label inputs and intermediate calculations using plain language descriptions
 Flow of intermediate calculations should be logical – e.g. sort calculations into
groups by new product
 Communicate the results of the analysis
 organise the outputs into a single section of the spreadsheet
 put a heading in each table explaining the output(s) being presented
 put a title and labels for the data being charted

27
 5 Good Modelling Practices in Microsoft Excel

 Multi-spreadsheet models should have clearly named sheets informing the user as to
the contents of each sheet.

 When referencing a cell across sheets, the sheet name will appear followed by an
exclamation mark before the cell reference. If the sheet name contains multiple words,
the full sheet name will be enclosed in single quotation marks

28
 5 Good Modelling Practices in Microsoft Excel
 Cell references become difficult to read with a large model.
So, naming cells and naming cell ranges can reduce the risk
of cell referencing mistakes.
 There are rules to naming cells and ranges
 Names must start with a letter, underscore(_), or a
backslash(\).
 Names cannot contain spaces
 Names cannot be cell references
 Names are not case-sensitive

 Once a cell has been named, you can use the name in
place of the cell address.
 Excel includes the named cells in the list of function name
prompts (see the picture) – to prevent spelling errors.
29
 Think & Apply: Individual Assignment

What the good practices/rules can help enhance the efficiency

of the workshop budget planning? How?

 Which of these do you focus on?

 Elaborate them in the Summary worksheet

30
The End