Basics for 
Mike Volker

The Sales Forecast

Contact: Mike Volker, Tel:(604)644-1926, Email: mike@volker.org

Planning vs Forecasting

In preparing a projection of a company's future revenues and expenses, a sales forecast must first be prepared. Whereas expenses can be controlled by management on a day by day basis, sales occur only when outside parties make a proactive purchase decision. Management, by devising a successful marketing strategy, influences the buying decisions of its customers but cannot make the customers buy. Therefore, management must somehow predict or forecast how many units will be sold and at what price and during what time frames.

Often, this process of forecasting is one of guesswork (wrong!). The marketing manager might say something like: "the total market for our products is 1,000,000 units annually and we should easily be able to capture 5% of this market or 50,000 units". This is a totally wrong approach, yet one which is frequently used. Why 5%? Why not 25%? For that matter why not 150%? (who says the market is limited to 1m units?) Outsiders, like stock analysts, might use this approach to predict a company's performance, but members of the management team should not. By the way, when entrepreneurs make a statement like "we're aiming for 2% of this billion dollar market" to prospective investors, they are likely to lose credibility instantly.

Forecasting, although it may be intelligently done, is just that: forecasting. The better approach is one that involves planning for sales. For example, the marketing manager might say, our plan is to sell 1k units in January, 2k units in February, 5k units in March, 10k units in April, and so on. We plan to achieve this number because we are recruiting 2 salespeople in November, then hiring additional ones in February and May. Also, we plan to attend 3 trade shows to launch the product and to sign up one major distributor and 40 retailers prior to the product launch date.

The Plan

The sales plan should show, usually on a month by month basis (with the months shown as columns on a spreadsheet), units shipped (this could be various models - itemized separately), and the average selling price for each unit. From this information, monthly sales revenues can be easily calculated. For example, the "forecast" would look like:

                    Jan       Feb        Mar  ..... FY Total

QP Product Line:   1000      2000       5000           50000

Average $Price:     113       105        100              85

Revenues:      $113,000  $210,500   $500,000      $4,250,000
The last column, FY Total, is the Fiscal Year total for the 12 months of the company's fiscal year. In this particular example, note that the average selling price declines somewhat over the first few months. This may reflect a price skimming strategy or simply reflect how the product is distributed, e.g. the ratio of distributor to dealer to end-user sales. The number of units must be justified. The planner should be able to support these numbers on the basis of human resources, competitive information, distribution plans, advertising programs, etc.

Start with the 4P's of Marketing

It is the four P's, i.e. product, price, place, promotion that comprise the firm's marketing mix. These ultimately determine the sales volumes. Think of this as a simple mathematical equation with the dependant variable, sales volume (or sales revenue), on the left side and the four independent variables (the 4 P's) on the right side. Although you don't know what the actual right side of this equation looks like, you may have a sense (or of you don't, you need to develop one) of the impact that the various decisions you make will have on sales. Focus groups, surveys, discussions, comparisons with like cases, etc may all provide some insight into this process. Remember - you cannot make a person pull out their wallet to buy your product and you cannot control your competition, but you can influence their buying decision by how you manage these variables. If your not sure, simply make some educated, reasoned guesses and then make adjustments to see how buyers react. For example, how important is pricing? If your product has a low manufactured cost, like software, you have a great deal of freedom in choosing a price. What if you reduce price by 50%? Or by 90? How will this impact your sales volume. Maybe it's better to sell 1000 units per day at $100/unit, than it is to sell 50 units at $1000/unit. Maybe a dedicated direct sales force will yield more predictable results rather than having independent dealers and agents. Perhaps a direct email promotion campaign will produce more  buyer interest than as advertisement in a trade publication. Trade shows may be very useful in promoting and selling one's products while at the same time providing a forum in which to obtain the answers to many of these questions, i.e. MBWA - manage by wandering around - ask a lot of questions and make a lot of deductions.

A good, practical way is to start by taking a look at where and how the sales are transacted, i.e. by looking at the channels through which the goods flow. Is there a distribution system in place through which the goods flow from the factory to the ultimate users? Where will these users go to make their purchases and how many sales will be made at these locations? How will pricing and promotion generate more buyer interest? 

In other words, start with the users of the product. How many potential users are there and how are you going to make them aware of the product (determined by your promotion plan) and then buy the product (determined by the product's features and benefits and price) at which location or through which distribution channel(s) (i.e. place)? 

Here's a simple illustration. You are selling a golf ball finder. In the first year, you decide to sell this in Canada and the USA. Through market studies, you can find out how many golfers there are in this market. These golfers read golf magazines and watch golf programs but because of your limited capital, you cannot advertise heavily. Instead, you decide to partner with a distributor of similar products that already has established golf retailers and golf courses as its clients. You can then determine how many retail outlets might carry your product (get your experienced distributor's insight on this). Based on some focus group discussions (i.e. how will golfers be made aware of the product and what percentage are likely to buy one at your target price?), and the size of these outlets (e.g. average sales per location), what's a reasonable number of units that you can move through the average retailer? Of course, this assumes that the distributor will push it (he might brand it and advertise it for you) and most importantly, can you produce and deliver the product at a profit - for the distributor and the retailer and yourself? It's not easy, but you've got to start somewhere. Start small and build up! 

How Much Detail?

So, how much detail should be in the plan? The more, the better. However, in terms of a plan which might be shown to investors or outsiders, too much detail may not only be unwieldy, but may reveal too much proprietary (or competitive) information. Certainly, as a planner, it is essential to have done the detailed work to support the summary numbers. The above sales plan may have been distilled from substantially more detailed work, for example:

QP Product Line (monthly shipments):
  Model QP100       200     400    1000
  Model QP200       300     700    1500
  Model QP300       500     900    2500

QP 100 Price        100      95      92
QP 200 Price        110     105      99
QP 300 Price        120     110     103
This level of detail may be further supported by showing how the selling prices have been determined based on the distribution channel mix. For example:
 QP100 (unit sales for January):
   Dealer Sales@$112:   80
   Distr. Sales@$92:   120
This shows that of the 1000 units in total to be shipped during January, 200 units consist of the QP100 model. Of these 200 QP100 units, 80 are sold via dealers at $112 per unit and 120 are sold via distributors at $92 per unit. This procedure is repeated for the other two models, the QP200 and QP300. This yields the aggregate numbers of 1000 units in total at an average selling price of $113 for total revenue in January of $113,000.

This approach to "forecasting" sales is more likely to produce actual results that are close to forecast. To achieve 120 units in sales through distributors, the company must have salespeople on board who can in fact sell 120 units to distributors and the company must have the collateral materials (brochures, price lists, web site, etc) to support this effort.

Planning is Iterative

At the outset, management devises its overall plan based on information available with respect to the market, competition, and so forth. If, after the first month, only 600 units have been shipped as compared to the plan to ship 1000 units, what should be done? Well, usually, excuses are made such as: "there were engineering glitches", or "the salespeople promised too much", or "those darn competitors beat us again". Regardless of the reasons, the plan (or the execution of the plan) needs to be reworked. Is our plan for the next few months still reasonable. If so, why? Can we "catch up"? How?

There is often a tendency to adjust the totals (usually downward). This may be somewhat defeatist in nature. Maybe the strategy should be revisited. For example, maybe a price adjustment will produce the desired volumes. Maybe a minor engineering change will increase demand. Maybe more distributors (instead of dealers) should be signed up. Perhaps our salespeople are just order-takers and don't really know how to sell! Isn't management fun? Is the plan wrong or are we having problems with the execution of our plan? Maybe its a bit of both.

By constantly monitoring, adjusting, and revising a plan in concert with the resources employed to execute the plan, a company will become very adept with this process. In my own experience, I recall being off target by a wide margin in the early stages of the company's development. However, after a few years, we could forecast to within a 5% accuracy for one to two years in the future.

It's always better to have a plan and then work that plan with perspicacity rather than leaving the results up to serendipity!

Copyright 1997-2003 Michael C. Volker
Comments and suggestions will be appreciated!
Updated: 030521

Back to Main Page