the Sales Forecast
by Dave Garwood
Everyone at Hi/Lo Tech, Inc. puts their own spin on the sales forecast.
"I discounted the sales forecast by 15% because
Sales is always optimistic and they've never hit the forecast yet,"
sighs Sally Dollar, Controller at Hi/Lo Tech.
"I bumped up the forecast by 20% since I know manufacturing always
discounts my forecast at least 10%. We are always late delivering to our
customers and we could use some extra inventory," declares Sam Seller,
VP of Sales.
Ronnie Ram, Product Manager, admits, "I used a highly optimistic
$5 million per year forecast for the new Hi-Speed Laser product to get
above the minimum ROI threshold that is required to justify proceeding
with the engineering development work. We need this product to have a
complete product lineup."
"Bless our sales people. They work hard, but forecasting our business
is almost impossible. I lowered the overall forecast by 20% because Wall
Street is brutal when you miss projections - and our Board of Directors
can get nasty when the stock price drops in half!" laments Dudley
Bucks, CFO at Hi/Lo Tech.
"I lowered the forecast by 25% when I placed the purchase orders,
"declared Patti, the Purchasing Planner. "When the forecast
is too high, I am the one they yell at when we get stuck with excess inventory."
"I low-ball the forecast at least 20%," confesses Jack Daniels,
Eastern Regional Sales Manager. "Corporate Sales Management uses
the forecast to set sales quotas. I want to make sure my bonus is a slam
"If we don't show at least a 25% increase in EPS next year, the stock
price will tank. My options will be under water! We need more sales volume
to reach that non-negotiable 25% goal. I insist we plan at least 30% sales
growth next year," shouts T.G. Highspot, CEO at Hi/Lo Tech. (Silently,
most of the troops are thinking about the Emperor's new clothes).
What are the chances for high-quality forecast performance at Hi/Lo Tech?
Nil. Obviously, well- intentioned people have intentionally biased the
forecast. While the forecast "tweaking" may appear to help individuals
meet their goals, bias insures a steady stream of "misses" --
actual sales always significantly above or below the forecast. Each individual
acted in good faith when they artificially inflated or deflated the forecast.
But they are all about to become victims of putting bias into the forecast.
The impact of this bias at Hi/Lo Tech is devastating. The sales forecast
credibility is shot. It becomes open season for everyone to make his own
forecast -- they're all different and all wrong! Measuring forecast accuracy
(I prefer to measure forecast quality) is a joke. Which forecast do you
measure? And it is no surprise when we have a significant plan to actual
delta - that was the plan! If you work in sales, why spend much time on
forecasting? Everyone second-guesses it and uses their own forecast anyway.
But what about the people who have a job to do? They need to decide every
day about ordering more material, hiring people, buying equipment, adding
product development resources, adding or deleting floor space, asking
banks to borrow money, changing expense budgets or talking to financial
analysts. They all need to know the same thing -- how much and what are
we going to sell? Make no mistake - they will get an answer. If we don't
develop a consensus, no-spin forecast, they will independently answer
the questions, sometimes just by default and do nothing. These separate
islands of forecasts lead to chaos and costly bad surprises are the result.
The symptoms are highly visible. Too much of the wrong stuff. Not enough
of the right stuff. Negative cash flow. Late and long deliveries to customers.
Lost sales. Shrinking gross margins as they scramble to recover. Wall
Street analysts lined up looking for answers while they change the stock
recommendation from Buy to Sell. Not a pretty picture. Spin-proofing the
sales forecasts will wipe out these costly symptoms.
Eliminating Biased Forecasts
The solution to this costly bias is not complicated. An effective demand
planning process and clear accountability for meeting the plan is the
key to the solution. Not the traditional "gun-to-head" accountability
where we search for guilty parties and embarrass (or fire) them. Heartlfelt
accountability is used. The focus is on what went wrong, not who screwed
up when the forecast is missed. Seeking root causes of significant deltas
and taking corrective action is the tactic.
Think of demand planning as a process, just like a manufacturing process
to bend, weld, drill parts or mix, sew or print materials or fill bottles
or assemble a product. The process has both input and output.
The input to demand planning includes historical sales, customer data,
quotes, planned price increases, new regulations, promotions, competitive
intelligence, new product impact etc. This input comes from many sources
that have different angles on the view of the future. For example, the
field sales people are close to the customer and should have a good idea
of near-term demand plans. Marketing has a better long-term view of the
total market and anticipated activities such as promotions and new products
that will cause future demand to be different than historical sales. Therefore,
a forum is needed to gather info, listen to different opinions and reach
a consensus on the future demands. This is an integral part of an effective
Sales ands Operations Planning Process.
The output of the process includes more than how many we think we will
sell in April, May, June, etc. The objective is not only a numerical forecast
but also to get buy-in from the entire organization, eliminating second-guessing
the numbers. The participants, i.e. process owners, need to present a
clear, logical case about how they arrived at the recommended numbers.
This includes the results of significant deviations from prior plans.
It also includes major assumptions used to arrive at the plans such as
signing up a new, major distributor, an end to a large customer's inventory
depletion or the effect of additional sales people. Ownership of the new
demand plan is transferred to the entire organization during this process.
Planning is essential. However, good results come from six sigma-style
execution. Clear accountability to execute a realistic demand plan is
what drives out the bias. The key is to make sure the accountability is
transferred to individuals and that they clearly understand their "piece
of the action."
Demand Stream Accountability
At Hi/Lo Tech, the demand planning process resulted in an annual forecast
of $120 million. The forecast for the first few months of the year was
around $8 million per month and the forecast ramped up to around $13 million
per month in the last quarter. These numbers made sense, i.e. passed the
sanity check. Last year they sold $105 million, including almost $24 million
in the fourth quarter. They added the impact of some new products, two
new markets and six additional sales people to justify the year-to-year
increase. The $120 million plan is broken down into a handful of meaningful
product families. This helps Hi/Lo Tech see where the business is coming
from and understand the profitability and resource impact. Company performance
now hinges greatly on selling the $120 million.
So who is going to sell and therefore be responsible for the $120 million
sales? The answer is no one! That has been the problem. Sure, you can
try to pin down Sam Seller, VP of Sales, as the one responsible but he
calls on very few customers directly. T.G. Highspot, CEO, is ultimately
responsible but he calls on even fewer customers. If sales fall short,
we can fire these two but that really doesn't solve the problem. Firing
people is an after-the-fact action and not constructive. We need to get
accountability for executing the plan properly aligned with the guys and
gals who are tracking down customers and getting the orders, making sure
they clearly understand how much of the plan they "own" and
are accountable for executing. We need to make sure they participate in
developing the plan, making them part of the process, not cramming it
down their throats. They get a chance to raise their hand and provide
feedback and suggestions before the plan is finalized.
Once the total forecast is final, we need to break down the $120 million
target into product families, then into demand streams for each family
and assign the pieces to individuals. For example, pet food is sold in
dry (biscuits and snacks) and wet (cans). Dry and wet become the two product
families. The company sells both dry and wet food to three specific markets
-- grocery stores, mass merchandisers and veterinarians. Each market has
its own needs and thus different sales people are assigned to focus on
these separate customer groups. Each of these three markets is a demand
stream. The forecast is made for each individual demand stream and rolled
up to get the total dry pet food demand. The same applies to cans.
Once the grocery sales team makes their forecast for both wet and dry,
actual sales are measured each month and compared to the plan. If the
delta is significant, the grocery sales team identifies the root causes,
takes corrective action and updates the plan. They are compensated on
the "quality" of their demand plans as well as sales growth.
They stand in front of their peers every month, report performance and
explain root causes of significant deviations. This ownership, measurement,
feedback, peer review, clear accountability and reward system drives the
bias out of the sales forecast.
Notice that the grocery team is not responsible for the $120 million or
even the $25 million forecast of dry pet food sales. They are responsible
for the $15 million forecast for grocery stores sales that was used to
get to the $25 million total.
Identifying the correct demand streams is not an exact science. Sometimes
it is dictated by geography. Other times it might be by major customer
or means of distribution. The best approach is to look at the sales organization
and insure accountability is aligned with the sales organization structure.
Some people are concerned about letting the sales people set their own
targets. The fear is that they will low-ball the number. A valid concern.
This is where senior management needs to step in.
The same sales forecast numbers derived from the demand stream roll ups
MUST be used for making the financial projections. Once the sales groups,
who are responsible for executing the plans, are used by the boss for
financial plans, the planning focus and performance improve.
After the demand stream forecasts for each product family are complete,
the numbers are rolled up in dollars. Senior management is responsible
for making sure the total forecast passes the sanity check. Do the future
numbers make sense when compared with recent history and overall market
growth (or decline) expectations? If not, go back and work with the sales
people to adjust the plans.
Senior management is also responsible for communicating the importance
of meeting the plan within reasonable tolerances. They should insist on
an explanation of root causes of significant deviations during the monthly
SOP meeting. They should offer any help to improve future plans, including
a compensation program that discourages bias. Putting this much thought
into demand planning takes time. Responsible executive leadership must
insist on taking whatever time is needed to establish a logical demand
plan in the first place and consistently meet it. The organization simply
cannot afford to not take the time to effectively manage demand. Pay now
or pay later when it is more expensive is the only choice.
The old paradigm was manage supply and demand just happens, i.e. hold
the plant tightly accountable for producing to plan, but little accountability
for meeting the forecast. Most companies can't afford the luxury of this
old paradigm any longer.
The new paradigm is to manage both. Try it. You'll like it!