Forecasting: If You Make It, Will They Come?

At Iron City Distilling, our sister company is the historic Pittsburgh Brewing Company, which has been in business since 1861. Best-known for producing some classic (and delicious) American lagers such as Iron City Beer and IC Light, they keep themselves pretty busy day-in and day-out.

Their beer is a part of the cultural fabric of western Pennsylvania and the kooky, fun-loving “yinzers” that inhabit the region. Theirs is the beer of coal miners and steel workers as well as the well-coifed hipster bartenders taking pride in the brewing history of the City of Bridges. Basically, there are always orders for beer, and business is good.

One day about two years ago, I was walking down the hallway and passed my buddy Brandon Mayes, PBC’s director of brewing operations. He looked flustered. When I asked him what the problem was, he replied, “Nothing. Just the damn production schedule is messed up again.”

To which I replied, “What the fuck is a schedule?”

To be clear, I was having a bit of fun. Of course, I know what a schedule is—it’s just that I view production scheduling very differently from my brewery brothers and sisters. For them, production is based on orders. In fact, the vast majority of production taking place at PBC is only scheduled if an order comes in. (Not a problem, since PBC is one of the few major breweries currently in growth mode with the majority of their brands.)

As a distiller, I can’t work like that. Sure, we get orders. But the orders themselves are only a small component in how we decide how much to make and when to make it. Whiskey, with its long maturation cycles, simply cannot be “made to order.” Even our gin takes around six weeks to produce, making the order-to-production decision tree more than a little complicated to execute.

For all our talk as distillers about the elevated form of liquid art that we hand-forge in our distilleries and warehouses, it can be easy to forget that making liquor—no matter how high-concept or artisanal—is still an actual business. To ensure there’s a cash flow to support the continued production of our boozy masterpieces, we need to make sure there’s some semblance of a plan.

[PAYWALL]

Why Forecasting Matters

There’s no “one size fits all” approach to production planning and forecasting. Each company is going to be different, but hopefully we can offer a few bits and bobs of wisdom to illuminate what is typically a dimly lit path for most distillers.

A few years ago, I was giving a talk at the Alcohol School in Montréal. Speaking after me was one of my friends and mentors, Douglas Murray, who at the time was nearing retirement with Diageo after being with the company for nearly 50 years. (He predated the formation of Diageo by a few decades when he worked for Distillers Company.)

After his talk, he opened the floor to questions. I remember one person asking him how Diageo—which at the time controlled about 25 percent of Scotland’s whiskey distilleries—determined how much whiskey to make. His answer is one that I’ll never forget. He said—and I’m paraphrasing here—that most people assume Diageo has some set of complicated calculations on which to base their future needs and production decisions. However, he explained, the truth is far simpler: When sales are good, they make as much whiskey as possible; when sales go down, so does their production.

Honestly, I’m sure it’s more complicated than that, but perhaps not by as much as we would assume. With a large multinational corporation at the helm, managing a wide array of distilleries and who knows how many SKUs, I could see a true irony coalescing: More complicated supply chains and production systems very likely defy more complex predictive forecasting models and thus things need to be kept simple.

Diageo has the benefit of enormous size and expansive financial resources. When times get tough, they can simply slow their roll, so to speak, and sit on extra inventory until the next boom cycle inevitably comes along. (Think of all the ghost whiskeys that Diageo regularly releases from distilleries such as Brora and Port Ellen.)

Well, that’s all fine and good for a company as large as Diageo, but what about the rest of us? Is there another approach that benefits the craft distiller?

As it turns out: Yes, there is.

When I need advice on production and sales forecasting, I turn to my friend and colleague, David Large. David is the director of finance and strategy at Thoroughbred Spirits Group, a prominent industry consulting group. (Disclosure: I also work for Thoroughbred as a production consultant.) I’ve known David for years, and he has consulted on a few of my projects, often working up highly detailed forecasting units that serve as a roadmap for my distillery’s success.

As David tells me, a lot of folks take the “If we’re gonna make it, we’re gonna sell it” approach to forecasting. That is to say, they often just jump into the production calendar going full speed without realizing that this tack is more akin to being a bull in your own china shop. A little thought-out strategizing can go a long way—and it all starts with one simple question:

“What Is Your Definition of Success?”

That question effectively defines everything that happens in your distillery.

Do you want to focus only on your hometown market and primarily sell out of your tasting room? Do you want to become a 100,000-case brand in five years? What about 1,000,000 cases?

It’s a simple question, but that doesn’t mean it’s simple to answer. You need to be realistic and honest with yourself about your strengths, weaknesses, and resources. David says that pretty much all conversations about production levels and associated forecasting begin with an understanding of the distillery’s capabilities to handle sales. With the client, he’ll typically discuss how many salespeople they have, brand ambassador programs, and more. He’ll also reach out to contacts in the distributor tier to get a finger on the industry’s pulse for various categories. Information in hand, he then sets himself to the task of forecasting production needs for sales outlooks of up to 10 years.

When I ask about his calculation method for these things, he says there’s generally a growth curve where the producer starts somewhat small the first year and steadily builds in subsequent years. Typically, you would want to grow production 30 to 40 percent after year one, with similar increases in years two and three. After that you should begin scaling the growth back five to 10 percent every year until you reach a comfortable production plateau.

This can be confusing to many folks, so let’s dive into a hypothetical example.

Let’s say there’s a distillery that’s making its own bourbon and expects it to be ready in four years. After internal deliberations, they’ve decided that they expect to open their future market at 5,000 six-pack cases in the first year on the shelf. With that number, they now have a good idea of the minimum amount of liquid they need to have ready in four years.

After doing the math on potential losses and angel’s share, they decide that to reach 5,000 cases they need to produce 150 barrels of bourbon in their first year. If we assume a 30 percent increase on the first year, then the following years’ projections would look something like this:

• Year 2: 195 barrels
• Year 3: 254 barrels
• Year 4: 305 barrels
• Year 5: 336 barrels
• Year 6: 336 barrels

What we’ve done here is quite simple. From our humble year-one production level of 150 barrels, we’ve added 30 percent to year two’s production. Year three gets another 30 percent boost, but year four sees only a 20 percent increase. Year five is a mere 10 percent bump over year four, and year six sees us steadying producting at 336 barrels.

David is quick to point out that forecasting models are never perfect and that they should be thought of as living documents. “There’s always going to be mistakes,” he says. So, it’s worth revisiting, reviewing, and updating internal forecasting models regularly. Adjust as needed—you’ll need to.

Maybe you have a product that takes off with a greater sales velocity than you expected. Or perhaps a distributor from a different market comes knocking on your door with a good opportunity. There are dozens of possible scenarios we could list as justification for messing with your forecast numbers.

Feed the Forecast

Decisions on forecasting shouldn’t be made in a vacuum. To have even a remote chance of success at the predictive endeavors, you’re going to need data.

If you’ve been in business for a few years, then you already have sales trends from previous financial quarters that will give you valuable insights into how individual products are trending. But it also helps to have data on other companies, products, and categories. At a minimum, you should sign up for a few industry newsletters. The Spirits Business and Shanken News Daily are both well worth the time it takes to fill out their subscription forms. Mark Brown’s daily Industry News Update is probably one of the best time investments you can make at the start of your day. All three of these news sources offer varying levels of detail and insight into industry trends and volumes.

If you want more granular data—and you probably will—then you’ll need to fork over a little cash. Starting around $1,500 per year for distilleries, you can sign up to become a member of the National Alcohol Beverage Control Association (NABCA), which can provide a wealth of information, including how specific SKUs are performing in individual control state markets.

The Distilled Spirits Council of the United States (DISCUS) is another valuable resource. While not always the most up-to-date source of info, the DISCUS website often contains a ton of sales data and trend reviews for previous years, helping you make better informed decisions on which way the industry winds are blowing.

Data are great and can fuel some mile-high ambitions—but unless you’re realistic about your distillery’s capabilities, you might be in for some rough surprises.

Jenna Diubaldo, partner and blender at Sons of Vancouver Distillery in British Columbia, describes the challenges of forecasting and keeping up with demand when you produce multiple product categories.

“We’ve been at max production capacity for a few years now, so our approach to balancing whiskey versus other products basically always leans toward, ‘How do we make more whiskey?’” she says. “Every mash spent making neutral spirit for liqueurs is one less whiskey mash on the books. Our demand for whiskey is far higher than any of our other products, so we try to make the least amount of neutral spirit possible while keeping all our products in stock. There’s no in-depth forecasting method for this other than looking at our sales from years past, calculating how much alcohol goes into each product, and then sorting out how much neutral spirit is needed based on where sales are headed for the year and the overall state of the industry.”

How Much Can You Make?

This is a key point to understand before you can truly go down the forecasting rabbit hole: You need to have a thorough understanding of your production capacity.

Essentially, the definition of production capacity in this context is the highest amount of liquid you can produce over a given timespan. This is typically viewed in terms of proof gallons or cases per year—you can decide exactly which metric works best for you.

There are a few steps to calculating production capacity. First, map out your production for every product you currently make. Be detailed. Which pieces of equipment are involved? Different products often require different equipment. For instance, let’s say you make bourbon from your own mash, but you also make gin using neutral grain spirit that you purchase. The bourbon needs a cereal cooker, fermentation vessels, and your still. Meanwhile, the gin needs only the still.

Next, figure out how long it takes to make the product. That time may be four years for the bourbon but only a few weeks for the gin. This is where you need to dive deeper and figure out whether there are any bottlenecks in your processes—very few craft distilleries are engineered to be perfectly balanced production houses. There’s usually at least one piece of equipment that slows the whole process down. (In my own distillery, there are fewer fermentors than the theoretical number of mashes I could produce in a given week.) Once you’ve figure that out, your production capacity is usually pretty simple to calculate: It’s nothing more than the max throughput on your slowest piece of equipment.

For instance, let’s say my distillery has a cooker that we could theoretically mash four times per day, producing 6,000 gallons of mash in a 24-hour period. But my fermentation capacity can handle only 3,000 gallons per day. In this case, my max production is 3,000 gallons of mash, not 6,000. That 3,000 gallons will need to be distilled, which leads to a certain amount of product that can easily be calculated if you know how much your recipes typically yield.

At Iron City Distilling, we have a theoretical maximum annual whiskey capacity of 6,000 American Standard Barrels (ASBs). That assumes we move production to seven days a week, 24-hours a day. From the metrics we’ve measured in our warehouse, we know that we wind up with approximately 42-proof gallons of liquid per barrel at the end of our maturation term. That means that our maximum annual whiskey production capacity is just over 250,000 proof gallons of matured liquid.

Knowing that maximum, there’s no point in allowing a forecasting unit to get into numbers over that level. Sure, my salespeople might feel like they could move four times that much liquid—but I can’t physically make it with the current equipment, so there’s no use in forecasting at that level. (Of course, this assumes we don’t purchase any more equipment to do the job.)

There are usually ways to improve the value of maximum production capacity, and that is something to have in the back of your mind as you move through these exercises. Could you increase the specific gravity of your wash? Could you reduce your fermentation time from five days to four? No matter what, you need to be sure that any improvements for the sake of production capacity don’t negatively affect your product quality. For instance, I could definitely increase my fermentation throughput here at Iron City by reducing my fermentation time from 96 to 72-hours. But I believe the resulting distillate wouldn’t be of the same quality, so we stick with the longer cycle.

Do Your Homework

Forecasting can be a headache, but it doesn’t have to be. All it takes is an honest understanding of your facility’s capacity and team’s capability to make your product. Sprinkle in a basic grasp of what’s happening in the industry, and you should be able to make sense of it all.

And if you’re still stressing over it, do some of these exercises with a glass of your favorite spirit in hand. It’s all good. You’ve earned it.