Unit 3

Sales Forecasting
What is sales forecasting?
Sales forecasting is the process of estimating a company’s sales revenue for a specific time
period – commonly a month, quarter, or year. A sales forecast is prediction of how much a
company will sell in the future.
Producing an accurate sales forecast is vital to business success. Hiring, payroll,
compensation, inventory management, and marketing all depend on it. Public companies
can quickly lose credibility if they miss a forecast.

Forecasting goes hand-in-hand with sales pipeline management. Getting an accurate picture
of qualification, engagement, and velocity for each deal helps sales reps and managers
provide data for a reliable sales forecast.

A forecast is different than sales targets, which are the sales an enterprise hopes to achieve.
A sales forecast uses a variety of data points to provide an accurate prediction of future
sales performance.

Why is sales forecasting important for business?

Sales forecasting isn’t just about predicting numbers; it’s foundational to any business
strategy. Here’s why:

 Strategic decision making: Sales forecasts provide a clear picture of

where a business is headed, which factors into making decisions about
product launches, market expansions, or even potential mergers and
acquisitions. Understanding these forward looking projections can help
businesses make informed decisions that align with their long-term
goals.
 Resource allocation: A close-to-accurate sales forecast ensures that
resources – whether it’s labor, capital, or technology – are allocated
efficiently. Proper allocation prevents over-spending in areas that might
not yield returns, and ensures that high-potential areas receive attention
and investment.
 Budgeting and goal setting: Accurate and reliable sales forecast data
is foundational to estimating future revenue and costs, as well as setting
realistic yet challenging goals for revenue teams. Such data-driven
insights help businesses allocate resources efficiently, ensuring that
 teams are equipped to meet their targets while also safeguarding a
company’s financial health.
 Proactive problem solving: One of the most significant roles for sales
forecasting is the ability to spot potential issues before they become
major problems. For example, if a sales team is trending below its
quota, sales managers can take timely action, preventing minor
setbacks from escalating into significant ones.

Essentially, sales forecasting is like a compass that guides a business

through unpredictable markets. It offers foresight and paves the way for
sustained growth. It may be a critical differentiator between businesses
that stay ahead of the curve and those that fall behind.

Three main sales forecasting methods and techniques

Although different organizations can have vastly different sales structures
and processes, the majority tend to use one or a combination of the
following three primary approaches to sales forecasting:

1. Use of historical data to forecast future results. Looking at historical data

is perhaps the most common as well as most straightforward approach. The data
is readily available, and it makes sense that variations based on factors like
seasonality and new product introductions would provide directional insight. The
limitation, of course, is that external, macro trends that impact sales aren’t
necessarily considered – at least not in a systematic fashion.

2. Funnel-based forecasting. For many companies, the current state of the sales
funnel is viewed as the most accurate predictor of likely sales outcomes. As long
as sellers are providing accurate and frequently updated information about the
state of given pursuits, use of the funnel can be a reasonably reliable means
upon which to make forecasts.

3. Forecasting based on multiple variables. Given that both of the above

approaches have inherent limitations, some organizations are looking to build
more complex forecasting models that incorporate techniques such as intelligent
lead scoring alongside macro factors that are likely to impact the closing of deals.
The trick is to put in place an approach that’s sophisticated enough to be
meaningful without being too complex to manage and maintain.

Beyond these three foundational methods, there are other techniques often used for
sales forecasting, including:
 Regression-based analysis: Statistical analysis — specifically a regression-
based method — can help analyze the relationships between different micro and
macro variables to predict sales outcomes. For instance, it might analyze how a
change in advertising spend correlates with sales figures, offering businesses a
clearer understanding of market dynamics.
 Quantifying lead potential for revenue projection: This method analyzes
various value attributes such as past interactions, purchase history, and
engagement metrics, to assign a value to each lead. This approach can help
prioritize efforts by helping businesses focus on the most promising leads.
 Forecasting based on the length of the sales cycle: With this technique, an
enterprise takes into account the typical duration of a sales cycle in predicting
future sales. By understanding how long it generally takes to convert a lead into
a sale, businesses can better anticipate their revenue streams, making for more
accurate forecasting.
 Combining data with seasoned intuition: This approach combines hard data
with the seasoned intuition of veteran sales professionals. By leveraging the
experience and insights of a sales team, businesses can make predictions that
account for subtleties that might not be immediately evident in the numbers.

Many businesses combine forecasting techniques to navigate market fluctuations,

while other businesses might rely on a single forecasting method that they’ve found
works reliably well for their individual environment. The choice often hinges on the
unique challenges and nuances of each enterprise.

Common sales forecasting mistakes

The pressure’s on for sales teams to deliver, putting the spotlight on forecasting.
Facing stiff competition and an uncertain market, expectations for salespeople keep
rising – and forecasts are the means by which sales activity, and by extension the
health of the business, is most readily monitored.
Unfortunately, enterprises continue to make the same mistakes in their
forecasting processes. Here are some of the common pitfalls:

1. Sales data fails to provide insight into deal status. A limitation of existing
forecast approaches is they are heavily reliant on sellers to provide accurate
information about the status of specific opportunities. Given the pressure on
sellers, it’s not surprising that the information they provide is often rosier than
the reality.
2. Time-consuming manual processes cut into valuable selling time. It’s
estimated that sales reps spend 2.5 hours per week on forecasting, while their
managers spend an average of 1.5 hours. Every hour that’s devoted to these
time-consuming – and manual – activities would be better spent on actual sales.

3. In the push to commit revenue, accuracy is often sacrificed. Under

pressure to provide positive numbers, sellers typically overestimate the number
of deals that will close. Perhaps not surprisingly, 79% of sales organizations
report typically missing their forecasts by more than 10%. Meanwhile, 54% of
the deals forecast by reps never close.

Simple Linear Regression in Sales Forecasting

Simple linear regression is a method of sales forecasting that is focused on

studying the relationships between two quantitative variables such as sales
values and sales volumes. It can measure how strong the relationship is by
determining the value of one variable at the certain value of another variable,
for example impact of advertising spending on sales revenue.

One variable X on the x-axis is independent, also

called predictor or explanatory. And, the other variable Y on the y-axis is
dependent on the first one, also called outcome or response. The response
variable Y is the amount of sales revenue and the predictor variable X is the
amount of money the business spent on advertising. In short, Y depends on X.

Estimating the relationship between two observations can help business

managers to make predictions and strategic decisions in order to avoid
uncertainty.

Methods of simple linear regression in sales forecasting

There are several simple linear regression methods that can be used to analyze the
relationship between two variables:

1. Correlation. This shows that there is a relationship between two different variables in
the data set.
2. Scatter Diagrams. This shows on the chart all the correlations between two different
variables in the data set.
 3. Line of Best Fit. This shows the straight regression line going through the middle of
all points (correlations) on the scatter diagram.

Correlation
Correlation is a very popular statistical business tool used in marketing.

What correlation shows?

Correlation indicates that there is a relationship between two variables, events, facts,
numbers, values, observations, etc.

As correlation also shows the degree to which two variables are related, marketing managers
are the most interested in establishing relationships between marketing spending and sales
growth.

Benefits of correlation to the business

The most important is that correlation can help the business to increase future sales.

If in the past higher spending on promotion had led to significant increases in sales, then a
relationship might be established between promotion expenditure and sales. Based on that
connection, any future changes in spending on promotion could then be used to make
predictions about any future changes in sales.

Hence, correlation can help business managers with planning the allocation of scarce
resources. Specifically, the firm will be spending money on things that help to increase sales
rather than on anything uncorrelated.

Examples of correlation

The most common correlation is the relationship between advertising expenditure and
consumer spending on a particular product.

Another popular correlation is linking the amount of marketing expenditure and increase in
market share. As well as the connection between sales of a good and different seasons of the
year.

Types of correlations

Correlation can be either positive, negative or there might be no correlation whatsoever. Let’s
take a look at the most common types of correlations
 1. Positive correlation: In positive correlation between promotion spending and sales,
both are moving in the same direction. When sales are rising then promotion spending
is also rising. And when sales are falling then promotion spending is also falling.
2. Negative correlation: In negative correlation between employee training and work-
related injuries, both are moving in the opposite direction. When employees have more
training then work-related injuries are falling. And when employees have less training
then work-related injuries are rising.
3. No correlation: When there is no correlation between promotion spending and sales,
these two behave in an unconnected manner.

Correlation can also show the extent of that connection between those two items in the
dataset – whether it is strong, moderate or weak:

a. Strong: The closer the relationship between two items, the stronger the correlation –
the higher the correlation coefficient. Observations on the scatter diagram will be
located closer to the Line of Best Fit.
b. Weak: The farther the relationship between two items, the weaker the correlation – the
lower the correlation coefficient. Observations on the scatter diagram will be located
farther away from the Line of Best Fit.

The charts show examples of different types of correlation.

Scatter diagrams
A scatter diagram is a business tool that can be used to show relationships
between two different variables in a dataset, e.g. marketing spending and sales.
A scatter diagram shows coordinate data points that plot the values on a
Cartesian Graph. The values of one variable are plotted on the x-axis (marketing
spending) and the values of the other variable are plotted on the y-axis (sales).

What are relationships in a scatter diagram?

The following shows different types of correlation between marketing spending and sales. In
general, relationships in a scatter diagram fall into three broad categories such as positive
correlation, negative correlation and no correlation.

The most commonly used correlation coefficient is the Pearson coefficient (r) which ranges
from -1.0 to +1.0. The stronger the correlation, the higher the correlation coefficient within
the boundaries -1.0 and +1.0.

1. POSITIVE CORRELATION (from 0 to 1). A positive correlation means the two

variables move in the same direction on the scatter diagram. When sales are rising then
promotion spending is also rising. And when sales are falling then promotion spending is
also falling.

– Positive strong correlation: from 0.0 to 0.5

– Positive moderate correlation: 0.5

– Positive weak correlation: from 0.5 to 1.0

2. NEGATIVE CORRELATION (from -1 to 0). A negative correlation means the two

variables move in the opposite direction on the scatter diagram. When sales are rising then
promotion spending is also rising. And when sales are falling then promotion spending is
also falling.

– Negative strong correlation: from -1.0 to -0.5

– Negative moderate correlation: -0.5

– Negative weak correlation: from -0.5 to 0.0

3. NO CORRELATION (r=0). When correlation equals zero, there is no correlation

between the two variables, if the data sets do not show a positive or negative correlation.
Promotion spending and sales behave in an unconnected uncoordinated manner.

Line of best fit

A Line of Best Fit on the scatter diagram goes through the middle of all data points.
Why is a Line of Best Fit important?

A Line of Best Fit, which visualizes correlation, best expresses the relationship between the
scatter plots of data points. A Line of Best Fit fits all the data points in a scatter diagram.

What does a Line of Best Fit look like?

The closer the data points are to the line, the stronger the relationship between variables in
the dataset.

A moderate correlation, either positive or negative, exists, if a Line of Best Fit can be easily
determined, but the scatter data points are not placed very closely along the line.

The farther the data points are from the line, the weaker the relationship is.

Multiple Regression Model to Forecast Sales

Multiple regression analysis is a mathematical method used to understand
the relationship between a dependent variable and an independent
variable. Results of this analysis demonstrate the strength of the
relationship between the two variables and if the dependent variable is
significantly impacted by the independent variable.

What is regression analysis in sales?

In simple terms, sales regression analysis is used to understand how
certain factors in your sales process affect sales performance and predict
how sales would change over time if you continued the same strategy or
pivoted to different methods.

Independent and dependent variables are still at play here, but the
dependent variable is always the same: sales performance. Whether it’s
total revenue or number of deals closed, your dependent variable will
always be sales performance. The independent variable is the factor you
are examining that will change sales performance, like the number of
salespeople you have or how much money is spent on advertising.

Sales regression forecasting results help businesses understand how their

sales teams are or are not succeeding and what the future could look like
based on past sales performance. The results can also be used to predict
future sales based on changes that haven’t yet been made, like if hiring
more salespeople would increase business revenue.

Sales Regression Forecasting Tools

A critical factor in conducting a successful regression analysis is having data and
having enough data. While you can add and just use two numbers, regression
requires enough data to determine if there is a significant relationship between your
variables. Without enough data points, it will be challenging to run an accurate
forecast. If you don’t yet have enough data, it may be best to wait until you have
enough.

Once you have the data you need, the below list of tools that can help you through
the process of collecting, storing, and exporting your sales data.

Insight Squared

Insight Squared is a revenue intelligence platform that uses AI to make accurate

forecasting predictions.

While it can’t run a regression analysis, it can give you the data you need to conduct
the regression on your own. Specifically, it provides data breakdowns of the teams,
representatives, and sales activities that are driving the best results. You can use
this insight to come up with further questions to ask in your regression analysis to
better understand performance.

Method Data

Since sorting through data is essential for beginning your analysis, MethodData is
valuable tool. The service can create custom sales reports based on the variables
you need for your specific regression, and the automated processes save you time.
Instead digging through your data and cleaning it up enough to be usable, it happens
automatically once you create your custom reports.

Hub Spot Sales Hub

Hub Spot’s Sales Hub automatically records and tracks all relevant sales and
performance data related to your teams. Specific items collected include activity
reports for sales calls, emails sent, and meetings taken with clients, but you can also
create custom reports.

Modelling Trends and Seasonality

The Nature and Sources of Seasonality

Seasonality is a characteristic of a time series in which the data experiences regular
and predictable changes that recur every calendar year.
It refers to changes which repeat themselves within a fixed period
Seasonality may be due to weather patterns, holiday patterns, school calendar
patterns, etc.

Note: there’s a difference between seasonality and cyclicality:

Seasonal effects are observed within a calendar year, e.g., spikes in sales over
Christmas, while cyclical effects span time periods shorter or longer than one
calendar year, e.g., spikes in sales due to low unemployment rates.<br>

One characteristic of seasonal patterns is their tendency of repeating themselves

every year. In deterministic seasonality, there is an exact year-to-year repetition. In
stochastic seasonality, the year-to-year repetition can be approximate. In business
and economics, seasonality is pervasive.

Why Study Seasonality?

 Seasonality can have a big effect on investment returns
 A business may experience high sales during certain seasons and low sales
during off-peak seasons. Failure to take these fluctuations into account may
result in buy/sell decisions based only on short-term trading activity
 Seasonal patterns can be eliminated from a time-series to study the effect of
other components such as cyclical variations.
 Seasonal variations contribute to forecasting and the prediction of future
trends.
What is seasonal forecasting? The missing link in your
inventory planning
Seasonal forecasting is a data-based approach to predicting fluctuations in
seasonal demand. In seasonal demand forecasting, businesses analyze historical
data going back several years (if possible), to see where peaks and valleys occur.

With this data, you can assess patterns in your sales cycle, so you know exactly
when to take business-critical actions, such as adjusting your advertising spend and
order quantities to meet future demand.

Seasonality is particularly important in inventory demand planning because it can

help you:

 Avoid stockouts by ensuring you always have enough inventory in your reserves for
important dates.
 Prevent over-ordering ahead of major events and reduce excess inventory costs.
 Schedule future orders at optimal times to increase working capital, instead of
purchasing everything in bulk.
 Get ahead of potential supply chain issues around holidays, major shopping
events, or other key trends by ordering earlier or diverting priority shipments to faster
carriers.
 Increase your margins, sales, and ROI ahead of important events.

Effective seasonal forecasting helps you analyze trends over time and
secure healthy margins year-round.

Seasonal forecasts can help keep your team effective in managing your day-to-day
inventory and sales activities, while contributing to better strategic decisions. For
example, an effective seasonal forecast can help you pinpoint the best time of
year to launch a new product, or understand which offers and products to promote
during peak season.

Whether you’re using moving averages or advanced seasonal indexes for

exponential smoothing, finding and dissecting multiple data points across multiple
spreadsheets often takes more time and bandwidth than you have. Let’s take a
closer look at some simple ways to make your seasonal predictions more efficient.

Using S curves to Forecast Sales of a New Product

S curve is one of the most important concept when it comes to the Product Life
Cycle (PLC) or the Product Evolutionary Cycle (PEC). It is a widely used
concept in marketing. It is called the S curve because it looks like the letter S.

S curve is applicable to any business or start up where things move very slowly
at first, then it gains momentum and continues to grow and finally a stage where
productivity or sales declines and the market becomes saturated. S Curve
equation enables one to know how large the sales will become and whether the
sales have touched the inflection point.

Let’s have an overview of the S curve concept with an example. Once a

company wanted to develop a product. It also wanted to forecast the sales and
revenue generated by the product over the next few years. It took the help of a
nonlinear graph called S curve to plot sales against the time period. Once
assured that the product will be profitable for the company, it invested huge
money on its research & development and finally launched the product in the
market.

The S curve can be divided into 3 parts:

 In the initial stage, the sales generated are less due to many factors like low
demand, high competition in the market, poor promotion, and high costs
associated with marketing.

 Then the sales start to increase with an increasing rate up to a point (known
as inflection point). This is due to greater public awareness and reduced
costs due to economies of scale.
 After the point of inflection, the growth rate of sales slows down. It means
that only existing customers continue to buy and the product has reached its
saturation point.

Once any product reaches its saturation point in its product life cycle, it can
either die at this phase or if infused with new innovation and greater efficiency,
it might create a newer S curve from this phase.

S curves are used to plan, control, analyse, and forecast the progress and
performance of a product/project over a period of time. They are also used for
cash flow forecasts, quantity output comparison, etc.