One of the key levers in building market analytic models is the use of ratios, of which there are many types. For example, the core model here at MDM Analytics uses a per-employee product consumption ratio. Once we define a total annualized market size in dollars at the end-user level for a product category nationally, we apply our estimated consumption by industry sector, down to the 6-digit NAICS level. We source the number of employees in each sector through Dun & Bradstreet to determine a per-employee consumption dollar amount.

That per-employee consumption ratio is what we use to profile specific geographic territories, from ZIP code on up. The power in this type of ratio profile is that it is based on the specific type of industry. So a profile of Detroit will reflect the heavy metal-cutting orientation, while a profile of Houston will reflect the strong petrochemical “DNA” of that defined territory.

This type of ratio works great in fairly predictable and steady product consumption environments, such as MRO and production. But you can’t apply it to project work and expect a reliable model. If you sell into hospitals, bed counts are often used to estimate ongoing product consumption. Fleet size and type is used for transportation customers; rig counts in oil and gas.

Construction markets are a little different animal. Employee ratios work well; so do building permits issued and other publicly available data that reflect construction activity in close to real-time. Employee ratios have been used for many years to estimate potential into contractor segments. For electrical products, a common ratio used is that contractors buy $40,000 to $50,000 of electrical products per employee annually. HVAC product marketers use a slightly higher ratio.

If there is a database or resource to get these types of counts, you’re almost there! Use your sales history to start building these ratios, with your high-penetration, best customers as a template. The first pass will be crude, but it’s possible to refine and get a close enough benchmark to start profiling the broader opportunity in your markets.

Sometimes in working with models, Winston Churchill’s quote comes to mind: “Democracy is the worst form of government, except for all those other forms that have been tried from time to time.” A lot of models fit that description. They may contain some margin of error that could be up to 20 percent or more, but directionally they provide perspective and insight to make better decisions. And as you refine them, they get even better. Ratios are an essential part of any market analytics toolkit.