Market forecasting has tended to rely on expert opinion rather than iteration of relevant factual data and formal modeling. The art of forecasting has been far from perfect in terms of accuracy, detail, and currency.

The McIlvaine approach has evolved over 25 years. A major step forward has been the combination of the Internet for easy access to millions of specific forecasts and the streamlining and structuring of the methods for determining the forecasts. This approach improves accuracy, provides greater details, and makes it possible to continually revise data


Here is an example of the system and its value. We have selected polymer sales in secondary municipal wastewater treatment plants in Taiwan. Polymerco is a hypothetical polymer supplier who sells some polymer in Taiwan but the amount is small. First of all Polymerco wants to know their percentage of the existing market. Next they would like to determine the longer term potential. Based on this knowledge they will then project sales for this country. They would continually evaluate revisions of this data and if changes exceed specified limits an “action trigger” would force re-evaluation of the campaign.

The McIlvaine online report Water and Wastewater Chemicals World Market not only provides the forecasts but also the factors behind the forecasts. Polymerco can select Taiwan, municipal wastewater, and organic flocculants. As seen in Figure 1, Water and Wastewater Chemicals: Industry, Product and Country, the sales of flocculants will grow from $2.39 million in 2002 to over $4 million in 2005.

($ Millions)

          YOU SELECTED:

($ Millions)

2000 2001 2002 2003 2004 2005
 $1.66  $1.99  $2.39  $2.86  $3.44  $4.12

The report also provides the relationships between the forecast and the causative factors. In Figure 2 Polymer Sales Taiwan Secondary Municipal Wastewater Treatment Plants, the top line is “polymer sales based on sewage transport mgd. Since much of the present wastewater does not pass through secondary treatment plants, the quantity of polymer which would be needed if all sewage received secondary treatment is much larger than the amount which is based on actual and predicted secondary treatment capacity. This difference is the world penetration factor.  McIlvaine is continually revising forecasts of sewage transport mgd, primary treatment mgd, and secondary treatment mgd for each country of the world.

Figure 2.

The second line is “theoretical polymer sales based on secondary mgd.” This is a ratio of world polymer sales /world secondary mgd. However, in the base year (2000) actual polymer sales in Taiwan are well below the world wide average. This difference is the geographical penetration ratio.

Line 3 is the projected polymer sales. This is the forecast for Taiwan for polymers but is actually derived from line 4.

Line 4 is the actual polymer sales in the base year growing at the industry growth rate.  But we predict that polymer sales will grow at a greater rate than the mgd of secondary treatment. The difference is the excess growth rate.

Line 5 represents Polymerco theoretical polymer sales in Taiwan growing at a rate to maintain a constant market share. This line is based on the ratio of worldwide Polymerco polymer sales divided by worldwide secondary mgd. This line is an indicator of sales that would result if the same level of penetration was made in Taiwan as in the rest of the world.

Line 6 represents Polymerco theoretical polymer sales in Taiwan but growing at the secondary mgd growth rate or industry growth rate

Line 7 is Polymerco minimum potential (explanation at end)

Line 8 is Polymerco projected sales. These are the sales revenues that Polymerco actually plans to obtain and could be characterized as the quota or target etc.

Line 9 is Polymerco sales if they grew only at the same rate as total polymer sales, in other words, if the market share is held constant. The difference between line 8 and 9 is the excess growth rate

Line 7 is the Polymerco minimum potential. This line represents Polymerco sales if Polymerco achieved the same rate of success in Taiwan as in the world as a whole (polymer/secondary mgd) and the company maintained its market share at this higher rate.

The value of this information is greatly enhanced by the ease with which the numbers can be updated. Since McIlvaine is continually changing projected secondary mgd numbers and since it is easy to link all the numbers, these forecasts can be kept current

The value of making forecasts based on the broader understanding of the market is considerable. If, for example, Polymerco has a polymer which will make belt filter presses achieve 40% dry solids instead of 20%, then the market for polymers in general and even secondary treatment would change. Therefore, knowledge of the line 1 numbers gives this broader perspective.

Most sales forecasts are based on perceived industry growth rates without regard for the potential of the “excess growth rate.” For example let’s say the industry is growing at 10% and Polymerco’s sales grow at 11%. Management could view this as a successful effort. But what if polymer sales are actually growing at 12%? Polymerco is actually losing rather than gaining market share.

One legitimate question is whether the cost of all these computations is so high as to be greater than the value. The answer is that the money has already been spent to track the industry data and work out the programs. The value is considerable so there is a high rate of return on the investment.

Purchaser Determines the Future Market

The total sales of a product equals the total of the purchases by the individual industry segments. The 2005 market for ball valves will equal the 2005 purchases by the power, wastewater, chemical, and other industries. These purchases in turn will be affected by the production growth rates in these industries between the present and 2005.

However, the purchase of a product is also affected by other factors which cause product sales to rise or fall at a different rate than the purchasing industries’ own production.

All these factors can be identified and quantified and integrated into the following formulas:

Present sales of a product in an industry in a county = WPF x GIC or GIP x GPR. Future sales = present sales x the industry growth rate x EGR.

These terms are defined as follows:

World Penetration Factor (WPF)

This factor is the ratio of specific product revenues in the base year divided by the potential revenues if all potential purchasers utilized this product rather than an alternative. Membrane systems are starting to compete with sand filters in municipal drinking water plants. The ratio of membrane system sales for drinking water in the base year divided by the sales of all filters for drinking water plants provides the WPF. The potential is measured in product revenues per unit of production. In semiconductors this could be wafer starts, in power it could be megawatts, and in pharmaceutical it could be production revenues. In wastewater it is millions of gallons per day. If there are 1,000 mgd of new secondary treatment plants but only 200 mgd will use belt filter presses, then the penetration rate is only 20 percent of belt filter presses.

Geographical Industry Capital Factor (GIC)

This factor is the ratio of the individual country capital investment in an industry in a present or future year divided by the world industry capital investment in the base year. So if the capital investment in the pharmaceutical industry in Taiwan in 2005 is $10 billion and the world
investment in 2000 is $200 billion then the GICIF is 0.05. This ratio is often based on some value other than revenue. In the power industry this is usually megawatts of new construction. In fact it is typically megawatts of coal-fired boilers or megawatts of combined cycle gas turbines. There are even values which are synthesized to provide maximum utility. Wet path megawatts are an example. It combines all the megawatts of new power plant types in such a manner that the resultant values are directly proportional to revenues for valve, pump and certain other products used in the power industry.

Examples in other industries are wafer starts per month (semiconductor industry), tons per day of refuse burning capacity (municipal incinerators) and barrels per day of new capacity (oil refining industry). Municipal wastewater treatment plant expenditures can be based on additional numbers of people served. However, providing separate values in millions of gallons per day for sewage transport, primary treatment, and secondary treatment provides much more useful tools.

Geographical Industry Production Factor (GIP)

This factor is the ratio of the individual country production revenues in an industry in a present or future year divided by the world industry production revenues in the base year. So if the semiconductor production revenues are $43.3 billion in Taiwan in 2005 and world semiconductor revenues in 2000 were $204 billion, then the GIP is 0.21. This ratio is often based on some value other than production revenues.  Megawatts, wafer starts, millions of gallons per day are all utilized as per the explanation above under GIC.

Geographical Penetration Ratio (GPR)

The market for a product may not be distributed geographically in the same percentages as the end user industry. But For many products there is no deviation between the industry activity and the revenue for the product being forecasted. If 30% of the production is one country then 30% of the specific product revenue will also be in that country. Any potential for deviation is lessened by the criteria used to determine the industry shares. For example instead of dividing the world power industry by  the simple megawatts of new construction or installed capacity, more specific divisions are used such as combined cycle gas turbines, or an even more specific "wet path megawatts".

In some cases the end user segmentation does not precisely fit the distribution of product revenues. The two most common reasons are regulations and product bias. Pollution control equipment revenue is typically driven by regulations. In one country 100% of the new equipment
will be fitted with back end air pollution control equipment and in another none of the equipment will be required. An example of product bias is in minienvironments for semiconductor plants. The Taiwanese jumped on this trend while the U.S. and others stayed with the traditional ballroom cleanrooms.

The penetration ratio is actual product distribution divided by the industry distribution. So if 30% of the pharmaceutical production revenue is generated in the U.S. but 60% of the product revenue in the base year is generated in the U.S., then the U.S. penetration ratio is 2.0. If the
pharmaceutical production revenue in a small country is 5% but only 3% of the product was sold there in the base year then the penetration ratio is 0.6. The penetration ratio is only useful in determining the base year product revenues is specific countries or regions. It is obviously 1.0 for the world. Also, the market in future years beyond the base year is determined by the combination of industry and product growth rates.

The penetration ratio concept can also be used to determine the future market for a specific proprietary product. The same ratio of present product sales in each country to industry revenues provides the proprietary product penetration ratio. This is a measure of success for the proprietary product in each country.

Excess Growth Rate (EGR)

This is the growth rate for the specific product revenue (SPR) in excess of the GIC or GIP. If the industry is growing at 6% and the product sales to that industry will grow at 10%/yr then the excess growth rate is 4% per year.

Product Use – Equipment

A product can be used as part of a new capital investment. This in turn can be new capacity or replacement capacity. Where capital investment is used directly as the unit of measure both new capacity and replacement capacity are included. But where units such as installed capacity and incremental increases are used it becomes more complicated. For example, mgd might be increasing at 5% per year but if the product life is only ten years then replacement is 10% of installed mgd. Total equipment purchases for 15% of the installed mgd are needed. (Note that this value representing 15% of installed mgd will increase 5% per year.) New capacity is the incremental increase in mgd each year. But some plants are being retired. So specific product revenue must take into account both new capacity and replacements.

McIlvaine industry forecast date is generally provided in two forms – new capacity and total capacity. Capital equipment purchases are a combination of incremental capacity additions and capacity additions that are replacements (equal to retirements). In some cases, capital investment forecasts are available. This includes both new and replacement capacity additions. For example, McIlvaine is continually updating forecasts of semiconductor capital equipment orders. Sales of cleanrooms, ultrapure water systems, and etching scrubbers will rise and fall proportionally to semi equipment orders.

In the pharmaceutical industry, capital investment forecasts are not readily available and equipment forecasts will need to be derived from production revenue forecasts. The incremental capacity increase equals the production increase. The replacement capacity value is derived as a percentage of total production.

So for every application in McIlvaine forecasts there is a capital equipment chart in percentages with the base year equaling 100%. There is also a consumables/replacement chart with the base year equaling 100%. If the capital equipment chart is based on total capital investment, then only this one chart is needed for equipment forecasts. If the chart is based on incremental capacity increases, then both charts are needed to determine equipment purchases.  

  Base Year Year 1  Year 2
Example 1      
Capital Investment $ 1000 1100 1200
Equipment % 100 110 120
Example 2        
Production Increase $  30 50 70
Equipment (new capacity) $ 3 5 7
Total production $ 250 300 370
Equipment (replacement) $ 5 6 7.4
Total Equipment $   8 11 14.4

In Example 2 the equipment for replacement includes new capacity offset by retirements as well as equipment replacements of existing plants.

Equipment forecasts are based on orders not startups. Some equipment is ordered three years before startup and some closer to startup. Equipment order forecasts are different where the interval between orders and startup is different. Therefore, multiple templates (application charts are necessary). For example, if coal-fired boiler installations are forecast at 4,000 MW in 2004 and 10,000 MW in 2005, 2002 FGD system orders would be factor x 10,000 whereas CEM orders would be factor x 4,000.


Purchases of consumables are generally proportional to installed capacity or more accurately capacity in use. Production revenues are equivalent to capacity in use. Therefore, most consumable forecasts are derived from industry capacity in use or production revenues.

There are occasions where the consumable purchases associated with capacity additions are significant enough to warrant inclusion. This only occurs when products with a two or three year life are considered consumables and when additions are very large compared to installed capacity.

Bags for coal-fired boilers are an example. Bag life is three years with x representing value of all installed bags. 0.33 x is the yearly replacement value. But if in one year capacity increased by 33% the value of bags for new plants would equal all the replacement bags.

Where consumables life is one week, it makes little difference whether capacity additions are 10% or 33%; yearly expenditures are still a function of the total capacity.