Forecasting and planning in today’s fast-changing business landscape is a challenging task for any organization. To survive, organizations require agile forecasting and planning processes to proactively manage change. Progressive CFOs and their teams recognize that the forecasting and planning processes, when done right, are powerful tools for informing the future of the organization.
3. History of Planning
Corporate Investment
Model
Discounted cash flow model to
stack rank R&D/Capex
investments
Fra Luca Bartolomeo de Pacioli,
"The Father of Accounting and
Bookkeeping“ was born. He was
the first person to publish a work on
the double-entry system of book-
keeping on the continent.
1447 1921
Corporate Investment
Model
Discounted cash flow model to
stack rank R&D/Capex
investments
GM almost runs out of money
because they overbought inventory
before the slump in ’20-’21. Alfred
Sloan developed a new
management principle that could
prevent this from happening
1972
Corporate Investment
Model
Discounted cash flow model to
stack rank R&D/Capex
investments
SAP was started in 1972 by five
former IBM employees with a vision
of creating a standard application
software for real-time business
processing.
Corporate Investment
Model
Discounted cash flow model to
stack rank R&D/Capex
investments
Microsoft released the first version
of Excel for the Macintosh on
September 30, 1985, and the first
Windows version in November
1987.
Corporate Investment
Model
Discounted cash flow model to
stack rank R&D/Capex
investments
Software as a service (SaaS)
model for corporate performance
management begins to take shape
1985
2003
+
Corporate Investment
Model
Discounted cash flow model to
stack rank R&D/Capex
investments
Machine Learning; Artificial
intelligence, predictive and
prescriptive Analytics; Robotics;
Blockchain etc.
4. Why do we need planning in the modern era
Uncertainty:
1. Globalization and the rise of new global competitors
2. “Software eating the world”
3. Increased speed of technology change and consumer preference
4. The large number of high growth startups entering every industry
The Startup Way – Eric Ries
6. Align to Your Strategy
1. Diagnosis
What is the challenge
2. Guiding Policy
Overall approach to
overcoming or coping with the
obstacle in the diagnosis
3. Coherent Actions
Designed to carry out the
guiding principle
7. The Reality
“Our research
reveals that, on
average, 95% of a
company’s
employees are
unaware of, or do
not understand, its
strategy.”
https://hbr.org/2005/10/the-office-of-
strategy-management
8. The reality… FP&A Empowerment Survey 2017
FP&A Empowerment Survey: 311 respondents (Due for release Nov’17) – Survey opening soon
Time on the Budget Process Time on the Forecast Process
9. The reality… FP&A Empowerment Survey 2017
Forecast Frequency
Not Applicable
Forecast Accuracy
FP&A Empowerment Survey: 311 respondents (Due for release Nov’17) – Survey opening soon
10. Traditional Planning & Budgeting
Bottom up Planning & Budgeting Top down Planning & Budgeting
16% of companies*
* FP&A Empowerment Survey (Due for release Nov’17)
14% of companies*
68% of companies* 2% of companies do nether*
Business Groups
Business Units
Business
Segments
11. Rolling Forecast
Decoupling Planning from Budgeting
Q1
* FP&A Empowerment Survey (Due for release Nov’17)
Q2 Q3 Q4 Q1
Budget
Forecast M1
Forecast M2
Forecast M3
58% of companies have rolling forecasts *
59% of companies feel they need to improve forecast accuracy*
12. Driver Based forecasting
Rolling up the driver not the result
* FP&A Empowerment Survey (Due for release Nov’17)
Drivers 1
e.g. HC
Drivers 2
e.g. Average
Salary
Drivers 3
e.g. Foreign
Exchange
X X =
Forecast
13. Salaries Cost – Driver model
GL data
HR data
(masked)
Dimensions:
• Period
• Function
• Region
Dimensions:
• Period
• Function
• Region
Uplift Cube
Dimensions:
• Period
• Function
• Region
c
HR data
(masked)
Output:
• Uplift factor
Salaries
Cube
Dimensions:
• Emp ID (masked)
• Period
Salaries Ref
data
Average
Salary Dimensions:
• Period
• Function
• Region
cPeriods
Output:
• Average salary
cTBH/Exits
Output
14. Integrated Planning
Traditional planning
* FP&A Empowerment Survey (Due for release Nov’17)
Sales Supply
chain
OPEX
Integrated planning
Sales
OPEX
Supply chain
Data from one forecast process feeds another
15. Business Planning Transformation
Current challenges
Business Planning currently performed
off-line and the processes differs by
Geo and Business Group
Cross BU, Cross Geo data analysis is
currently not possible until it is
consolidated in the corporate tool
Difficult to perform any business
analysis on internal and external
drivers
Future State
Business Planning performed in a single
solution, managed by a Center of
Excellence to ensure consistency
across Geo’s and BU’s
Business planning results are loaded
into the data lake on a near real time
basis to enable visibility across BU’s
and Geo’s
Solution to capture current underlying
assumptions, to enable more detail
understanding of how the numbers were
derived
Benefits
Standardized process ensure
efficiencies of scale and less
repetitive work reducing time
spent on forecasting
Improved decision making
across Geo’s and BU’s as all the
data is available in on place
Analysis can be done on
underlying assumptions, driving
more analytical forecasting and
improving forecast accuracy
16. Excel – The good, the bad and the ugly
https://www.forbes.com/sites/metabrown/2017/12/31/3-reasons-why-replacing-microsoft-excel-is-worth-money-to-a-boss/#2b61466737a9
Reasons to switch (Forbes):
• Hands-on involvement sucks up expensive labor
• Spreadsheet error leaves business open to risk and liability
• When spreadsheet creators leave, knowledge leaves with them
My reason’s to switch (Planning Focus):
• Valuable information is stored off line
• Data capture by multiple people & consolidation is prone to error
• Versioning is very difficult to manage
• Data is not secure
17. Excel – Why it’s here to stay
1. The ease and speed you can create a model
2. The barriers to entry are low – no IT involvement, no PO required
3. Accessible training online
4. Excel has advanced:
• PowerQuery – connects, shape and/or cleanses data
• Power Pivot – data modelling
• Power BI – data vizualization
18. Building the perfect Excel model
Data Source
e.g. ERP
Power
Query
Power
Pivot
Excel Pivot
Table
Pivot data
Build cubes
Filter data
Error checks
Formulas
Link tables
Lookup’s
Sum
GetPivotData
19. Options
Technology Pro’s Con’s
Excel based
solutions
• Easy to build and maintain
• Low Cost solution
• Data collection prone to error and requires it to be
consolidated
• Requires multiple emails and co-ordination to collect all
submissions
• Security could be a challenge
Database with web
interface solutions
• Data loads inputs directly into SQL
• Interface is online reducing emails
• Integrated reporting - real time feedback
• Highly customizable
• IT Support required for setup
• Risk of loosing data if there are no proper backups
• Interface may be difficult to change easily
Cloud purpose built
planning solutions
• Interface is online reducing emails and tighter
control over access
• Low IT requirements and easy setup
• Low cost
• All online, so instant feedback on results
• Limited ability to customization – typically your processes
need to fit into their solution
• Limited scenario modeling capability
Cloud planning/
modelling solutions
• Interface is online reducing emails and tighter
control over access
• Low IT Support Required
• Strong Modelling capabilities
• Strong audit capabilities
• Model maintenance requires expertise
• Data size limitations
• SaaS application so data controlled by someone else
ERP solutions
• Data is captured in batch, but updates speeds are
within seconds
• Enterprise ready – Audit etc.
• Most Expensive
• Highly dependent on IT support and set up
• Deployment times slow
• Limited or costly customization
• Limited scenario modeling capability
20. Rise of the Machine – AI and the impact of
Finance
23. Moravec Paradox
Moravec Paradox states that while "it is comparatively
easy to make computers exhibit adult level
performance on intelligence tests or playing checkers,
and difficult or impossible to give them the skills of a
one-year-old when it comes to perception and
mobility".
24. The real advantage of using ML
Advantages:
1. Reduced natural bias
2. Identify things we never thought of – the unknown unknowns
3. Analyze large amounts of data and learn
Barriers:
1. Volume of Finance data available
2. Data can be messy
3. ML is not readily accessible or available
25. Readily available types of machine learning
Sentiment Analysis
Scale 0-1
(0 being 100% negative and
1 being 100% positive)
Key word
Identifies the key words in a
set of text
Boolean prediction
Predicts if an event will
happen or not e.g. a pipeline
deal won or lost
Forecasting
Linear time based forecast
AutoML is a suite of machine learning products that enables users with limited machine learning
expertise to train high-quality models specific to their business needs
26. Interpredictive Planning – The future of financial
planning
Sales
OPEX
Supply chain
Human
Machine
Accountability
Interpredictive planning focuses time and
accountability driver inputs. These drivers form
the inputs into the forecasting model. Once
entered, the model in then optimized according
to specific requirements e.g. Revenue growth;
EBITDA
The advantages are:
1. More responsive change to the P&L driven by
changes in drives e.g. demand
2. Faster planning cycles, driving better decision
making
3. Improved forecast quality as the solution
learns over time and improves
Reducetime
Machine
ExternalfactorsOptimization, refers to the study of decision problems in which one seeks to minimize (min) or maximize (max) a
function by systematically choosing the values of variables within their allowed sets.
Optimization
27. Thank You
For blogs – link in at:
www.linkedin.com/in/jamesmyersw
3 Key Predictions for 2019 for FP&A
How can finance leaders really embrace their role as a strategic
advisor, connecting business to strategy and data?
Artificial Intelligence and the Future of Finance
28. Bibliography
1. FP&A Empowerment Survey: http://bit.ly/FPAEmpower
2. The Startup Way – Eric Ries
3. To be a Machine – Mark O’Connell
4. https://www.forbes.com/sites/metabrown/2017/12/31/3-reasons-why-replacing-microsoft-excel-is-
worth-money-to-a-boss/#2b61466737a9
5. Wikipedia
Editor's Notes
In Alfred Sloan’s My years with General Motors (1963), he describes how GE almost ran out of cash
Co-ordinated control of decentralized operation – rigorous production estimates for each division manager of the precise number of car the GM should sell in an ideal year. Using these estimated in combination with internal targets and external macro economic factors, the company would product a forecast of how many car each division was responsible for selling
Step 1: Align to the business Strategy
Step 2: Break these down into strategic Priorities
Step 3: Define the Key Objectives for each of the priorities
Step 4: Define what the measures of success
Step 5: Identify the key drivers of the measures
Step 6: Create actionable metrics around the key drivers
Issues with traditional planning
Strategy is not integrated
Multiple level – each with non standardized process
Difficult to get ownership of the numbers if there is no bottoms up
Too much time rolling up the numbers and not enough time on the analytics
Issues with traditional planning
Strategy is not integrated
Multiple level – each with non standardized process
Difficult to get ownership of the numbers if there is no bottoms up
Too much time rolling up the numbers and not enough time on the analytics
Hands-on involvement sucks up expensive labor
Building and maintain spreadsheets cost time and money
Spreadsheet error leaves business open to risk and liability
Most conservative research-based estimate indicates that at least 1 in every 100 spreadsheet cells contains an error
$6 billion dollar loss attributed to the London Whale was partly due to a spreadsheet error
When spreadsheet creators leave, knowledge leaves with them
Spreadsheets are created informally, and they lack elements that make professionally-developed software desirable, such as:
Designers with a solid understanding of user needs
Design review, to prevent problems before the software is built
Programmers who are well-trained and use appropriate processes to work efficiently and minimize flaws in the end product
Quality assurance, to ensure the program does what it’s supposed to do
User testing to detect and correct problems in the product before it gets into general circulation
Documentation, both internal documents used by developers and documents such as manuals and training guides for end users.