A method for creating more reliable forecast

6 Ways To Remove Uncertainty From Sales Forecasts

An Adage For Improving Forecast Quality

A sales forecast represents many things in business, but is most universally understood as the pulse of a company. From an operations perspective, a forecast lets the supply team anticipate and quickly respond to changes in customer demand, and helps manufacturing and distribution companies optimize inventory levels, minimize stock-outs and improve production capacity. Conversely, a poorly developed and inaccurate forecast negatively affects operational performance of a company and can become a barrier to achieving strategic imperatives and fiscal goals.

Many surveys have shown that successful companies are able to analyze historical sales patterns and create an accurate baseline forecast, which, at least, includes seasonality, trend and some measure of forecast error. A good forecast lends itself to lower operating cost, making better purchasing decisions, improved manufacturing schedules, optimized product mix, and higher customer satisfaction. The intention behind improving a forecast is to support increased sales. So why do some companies fail miserably at forecasting?

Companies that struggle with forecasting likely fail at it because they do not provide sufficient organizational attention to the process. The typical result (in all but the simplest and most predictable businesses) is significantly higher costs and increased levels of inefficiency creating stress on the company. Such companies tend to be very reactive, and usually a little behind in meeting customer expectations.

The results of poor demand planning ripple through an organization and are often recognizable with low levels of confidence in the forecast, lack of process ownership, no measurement of quality, and reliance on inappropriate tools such as standalone spreadsheets. The product of which can lead to lengthy, and even confrontational discussion about whom in the organization has the correct planning numbers. It is not surprising that further lack of confidence in being able to make sound management decisions is inevitable.

Operating on an unreliable, or highly volatile forecast, frequently results in a greater inefficiency within supply planning. Keeping customers satisfied, and avoiding a ‘bull-whip’ effect on your supply chain, may require compensating with higher inventory. The financial impact could result in writing off and discarding of unused product later. An unreliable forecast can result in disrupted supply lines, and increases the need to expend more energy, and money, at the risk of still leaving many customers dissatisfied.

Contrary to the common statement that all forecasts are wrong; a more precise statement would be that all predictions about future sales contain some degree of error. The legendary Yankees Baseball Manager Yogi Berra once said, “If you don’t know where you’re going, then you’ll end up someplace else.” It would be a stretch to say there’s supply chain wisdom lurking in there somewhere, but if it were, it would suggest that setting a plan and executing to it while monitoring progress, improves the chance of a company achieving its goals. So what are some things that can be done to ensure that a supply chain forecast is less ambiguous and more reliable?

A good place to start is to accept that the forecast will likely contain a degree of error. Secondly, knowing if the forecast is too high or too low will begin to expose what needs to be done to improve it. A sound forecasting process seeks to find what is creating forecast bias in order to minimize errors. There are many ways to measure statistical forecast inaccuracy, but here I will examine that part of the process where hunches and gut feelings get introduced.

I’ve commonly witnessed many demand planners override a good forecast with different numbers obtained from sales and marketing people. Very often the changes do not include the underlying business assumptions to justify forecast modifications, or the change requests are vague and incomplete. Regardless of being void of any insight, the low quality changes are accepted and introduce additional forecast error.

    I’ve commonly witnessed many demand planners override a good forecast with different numbers obtained from sales and marketing people.

The two most common reasons why sales inputs create additional error are because they are unmoored from proper historical analysis, or targeting a different goal. Ensuring that the sales person’s input originates from the statistical baseline, is the simplest way to correct the first problem. In most companies, the forecaster is a person using mathematical models to generate a detailed set of numbers that often represents a tenable estimate of the future demand. They usually perform extensive statistical analysis of sales data, accounting for trends, seasonality, and even changes in the product mix. Using the forecaster’s data as the starting point for a sales person’s input, will not only help to improve the accuracy of the forecast, but it will save sales people time as they will no longer have to compile their own statistics. Correcting the second problem requires a little more effort at uncovering the rationale behind a sales person’s request to change the baseline.

During my previous experience as a Demand Manager, I recall that getting insight from sales and marketing about their intended demand generation activities can be a challenging task. While a forecaster’s function is to sense what demand might be statistically based on history, the role of sales and marketing is to understand and communicate how they intend to shape the future. Leading companies realize that forecasting error is partially the result of incorrect forecast models and partly the result of communicating unrealistic expectations from the sales creation team. This article discusses one method for creating more reliable inputs.

Since leaving the world of planning to become a sales person, I have recognized that I tend to be a lot more coin operated these days and more focused on tasks that generate my pay. In having to prioritize the worthiness of my own sales forecast, I stumbled on the following, extremely insightful method, of qualifying my own activities. After all, not all business interactions will result in additional sales and my goal as a commissioned sales person is to make money. To avoid expending too much time and effort in the wrong places, I implemented a set of questions to qualify the priority, timing, and potential of my sales opportunities. It struck me that these are the same questions that would have helped me qualify the sales forecast inputs I used to receive as a planner. The method can easily be remembered using the following acronym - B.A.N.T.E.R.

Now, as with most things, these observations may not be directly applicable to everyone. The goal here is to ignite some creative thinking on this insight so feel free to interpret specifics. Let’s examine the points individually.

  • B – Budget
  • A – Authority
  • N – Need
  • T – Timing
  • E – Execution
  • R – Requirement

Budget – These days, it goes without saying that a sales person runs the risk of wasting a lot of time and cost without knowing if a prospect has a budget. Without a budget, there is not much likelihood of a sale occurring because the prospect has no money to spend. Many inexperienced, hungry, and overly enthusiastic sales people believe a client will eventually purchase simply because the sales person is more accommodating than the competition. However, a lot of time and money could be wasted, juggling company resources while jumping through hoops for an opportunity that will likely never happen.

In my experience, non-buying prospects are not entirely forthcoming about their budget. I’m not entirely sure why that is, but suspect it’s for fear that the sales person will disengage from discussion upon discovering there is no money. It could also be because they feel as though the sales person will take full advantage of the budget information in order to maximize a commission. Often, buyers without a budget will proclaim that if the sales person presents enough value, there should be no problem obtaining financing for the purchase. However, to experienced sales people, prospects unwilling to discuss their budget may be sending the wrong signals to the sales person. It may be translated as the prospect is simply fishing for information, or more likely, they are not even in a buying cycle. The proverbial trust issues between buyer and seller begin to surface.

At the very least, establishing an estimated procurement budget ensures that both sides begin their interaction with the right expectations. Quoting an accurate price may be virtually impossible until the sales person has fully outlined requirements, but the buyer should expect the cost is within an acceptable range. Navigating this issue is an essential sales skill, but for the purpose of supply chain planning, budget insight will greatly help qualify the forecast input.

Another way to address budget is to clarify the prospect’s capital prioritization process. For example; ensuring the purchase will not get bumped in favor of a new forklift, may require adherence to a rigorous financial prioritization process. Understanding that a company has prioritized capital for the purchase is almost more insightful than knowing the budget limit itself. There is also the reality that some buyers manage a budget, but once spent, it‘s done – and they have no further budget to allocate. Others in organizations, usually the “C” suite, can reallocate more budget and make unplanned purchases if the business requirement warrants attention. With respect to budget, some important questions to ask a sales person when qualifying their forecast inputs are;

  • Who is the person responsible and accountable for fiscal decision making?
  • What is the customer’s capital prioritization process?
  • What are the budgetary limits related to this purchase?
  • What is the value proposition being placed in front of the prospect for our product?

In many cases, if the sales person is too low in the decision tree, they may not be able to answer these questions. The sales person may be talking to people who are completely unable to establish a budget. This is a sure sign that the forecast timing is optimistic and may not even occur at all.

    In my experience, most non-buying prospects are not entirely forthcoming about their budget.

Authority – Many sales people call into C-level managers these days because of escalated company decision making policy. C-level managers have decision making capability over business direction and company budget. However, it is unlikely that C-level managers have time to answer every phone call, email, and postcard received. It’s more likely that they initiate action based on significant business issues presented to them internally. Therefore, most of the marketing directed at them is just noise. The same scenario could be happening with any person responsible for purchasing in an organization.

Most purchasing decisions require authority. If the purchaser does not have the authority to make the final decision, then the sales person’s risk of not completing the transaction is very high. The sales person’s conundrum is that decision makers often delegate investigative responsibility to their staff. The delegation probably goes something like this - “You’re the expert, find me a solution to…, but do not involve me until you have established a shorter list of suppliers to select from.” This action can create inherent risk for both sides engaged in the transaction. The risk can become particularly significant when large purchases are at stake.

The Authority questions used to qualify a sales person’s forecast, identify whether the decision maker is aware of the pending interaction, or better yet, involved in the purchasing process. They reveal whether the sales person has established enough contact to close the sale. The following question might be asked by the demand manager during communication with the sales person;

    “You’re the expert, find me a solution to..., but don’t involve me until you have established a shorter list of suppliers to select from.”

Again, if the sales person is speaking to people too low in the decision tree, they may not be able to get the answers they need to these questions. The sales person may be talking to people unable to make final decisions needed to move things forward. This is another sure sign that the sale will either take longer than expected, or again, not even occur at all.

Need – There is nothing more compelling in a sales cycle than establishing what the needs are of the customer. I suspect most people get hundreds of marketing calls per year, but very few people are likely compelled to shop this way. Clearly, a guy with a chainsaw immediately showing up to sell you on tree removal, 10 minutes after it falls on your house is more than a little sketchy, but my point is this; if the customer has a requirement that facilitates a business transaction, it should be evident. Prospects typically will not spend money without a compelling need and the sales person should have uncovered it. Prospects that vaguely answer the ‘Need’ question are usually in the early stages of a buying process. Alternatively, they could be just educating themselves and not ready to buy.

The goal in the ‘Need’ query is to help the prospect frame their requirements. This involves hard work, product knowledge and the discipline to execute without making assumptions. Discovering a customer’s need is not always exciting but it can be rewarding when properly executed. As an example; I recognize that buyers often balance how I’m able to handle their issues against price. If I’m accountable, deal with issues quickly, save time, and can mitigate most of the buyer’s risk, then I expect they buyer will realize a significant benefit in our interaction during discovery. When the buyer realizes the value, the probability of us doing business together will be more likely. Inquiries to make regarding the sales person’s forecast might be:

  • What are the compelling business reasons behind this prospect’s purchase order?
  • Is the buyer already working with another vendor?
  • How will buying the product from us resolve the customer’s need?
  • What does the customer like, or dislike about our products?

Timing – Timing can be examined from many different angles, but the obvious one concerns the buyer’s purchasing timeline. While it is risky for a sales person to push a buyer, they obviously cannot let the sale linger forever. For this reason, it is important to set milestones and goal posts in a sales cycle. Attaching a time expectation to the purchasing steps, will help the buyer manage the purchasing process better. In addition, a clear timeline clarifies to the sales person how far in the future they can expect an order. A sales person can also use timing to gauge the potential in the business interaction by monitoring how well the prospect adheres to the schedule of events. Things like internal planning, budget approval, product sampling, demonstrations, contract review, Executive bridging discussions, price negotiation, expectation alignment, legal adjustments, and other essential steps can highlight the purchase timing and clarify the buyer’s intention. Questions to ask when further qualifying the timing of the forecast might be:

    While it is risky for a sales person to push a buyer, they obviously cannot let the sale linger forever.

Execution – While the previous four points are useful in helping to remove ambiguity from the sales person’s forecast, it should not be overlooked that closing new business is often a team sport. It is not an option to turn away new business, therefore, forecasting it should be taken seriously by the demand manager. Any sales person would prefer to under promise and over deliver. At this point, qualification of the forecast should involve at least a cursory review of the company’s ability to execute on the new order. There is nothing more embarrassing and frustrating to a sales person than making a commitment to a customer and then finding out that it cannot be delivered as promised. Furthermore, there is nothing more frustrating to a planner than unexpectedly having to jump through hoops in order to keep the unrealistic promises made by a sales person.

Determining the ability to execute on the forecast requires a review of available inventory, the current rate of demand, resupply lead time, manufacturing capacity, and what is already outstanding in purchasing or plant schedules. Other customer’s priorities may have to be shifted when the sales person’s forecast is late. If the order incurs expedited delivery cost, or production needs to stop and retool to get the order out on time, margin is eroded. Depending on the competitive pressure to quote aggressively, reporting a late forecast puts unnecessary pressure on profits and hinders a company’s ability to operate efficiently.

It should be duly noted by sales people that supply planning is not an easy task, and the implications of reporting a forecast to supply chain late may be detrimental to their relationship with the customer. In the best interest of the business, it is vitally important that forecast expectations be openly communicated between the sales person and the forecast manager. The following questions should be addressed following questions should be addressed with planning.

  • Will there be enough inventory to fill the customer’s order without shorting others?
  • How long will it take to get more product if there is not enough to fill all the orders?
  • What will be the impact on operations to fill the order on time if supply is short?
  • Is there any risk that the order might not get filled on time?

Requirements – This should be self-explanatory - what does the customer want? How a customer thinks and feels about a company and their products is a key aspect of trade. A sales person’s interaction with a customer will shape the perception of value. Dealing with the sales person’s company will further develop the client’s perception too. Therefore, it is everyone’s task to meet the needs of the customer more precisely than the competitors, without exceeding the limitation of an acceptable margin of course.

Although it may be assumed that customers usually think carefully about the business purchases they make, it would be impossible for them to know all of their choices without speaking to a salesperson first. Wherever there are options involved in a purchase, the “art of possible” also needs to be thoroughly explored with the customer. A customer’s complete understanding of what options are available may change the forecasted requirement. Before making commitments with a customer, establishing requirements with the internal supply planning team, ensures that the right products will be available when the customer orders. Questions to explore regarding requirements might be:

  • What has the customer asked for and why?
  • What is the best configuration of options and price to meet the customer requirement?
  • Is there anything else the customer might be interested in that wasn’t discussed?

Contributions from sales and marketing are essential to improving forecast accuracy. Companies that produce the most accurate forecast, manage variability by discussing upside potential and downside risk not represented in their statistical analysis.

    Wherever there are options involved in a purchase, the “art of possible” also needs to be thoroughly explored with the customer.

Measurement – Although, measurement is not part of the BANTER acronym, it is an important element to the topic and should not be overlooked. The famous statistician George Box is quoted as saying “All models are wrong; some are useful.” In demand forecasting, accuracy is a relative term without the accompaniment of a measurement. All too many organizations I’ve helped were initially confident their forecast was near perfect, yet rarely were they able to produce the statistical evidence to back up that claim. The two most important indicators in measuring forecast accuracy are how inaccurate the forecast is and which direction it needs to go to improve. Forecast performance can be validated by comparing the expired forecast to actual sales. Measuring the forecast will help build organizational confidence in the numbers and inherently help improve the quality.

Obtaining an excellent reliable forecast requires a good set of tools and a deep understanding of the marketplace demand. Awareness of the customer’s business needs requires open communication about demand creation activity. The ability to detect bias in both the quantitative and qualitative inputs is essential to gaining improvements in forecast accuracy.

Once an accurate representation of future demand is available, an organization will be more capable of determining a supply strategy. They will be better equipped at optimizing inventory and creating the right mix of products to elevate customer order fill rates. Finally, a high quality forecast will help a company improve awareness of customer behavior and market trends, allow them to operate more efficiently, drive profitable business growth, and increase their strategic performance.

Published by Slimstock, market leader in forecasting, demand planning, and inventory optimization
Since 1993, Slimstock has been synonymous with better demand forecasting, effective inventory optimization, clearer inventory analysis and continuous improvement of inventory reliability. Our customer base consists of over 650 companies worldwide, across a diverse range of industries, covering large, medium and small enterprises.