Essential Reading for Retail CFOs: The Three Most Common Mistakes in Forecasting

Editor's Note:"CFO in the Know" is an editorial series exploring the role of CFOs across industries. The full series can beviewed here。

Last year was a tough one for Barneys. The luxury retailerclosed 15 of its 22 storesand turned the 2019 holiday season into a clearance sale asit prepared toshut its flagship store in Manhattan. Other major retailers also struggled,with at least 16filing for Chapter 11 bankruptcy protection.

Growth in online sales (which now account formore than 11%of U.S. commerce) was a factor, and so were tariffs. Some of last year's most popular items, including electronics, shoes, and hats,were affected bythe trade standoff with China.

However, retailers' forecasting missteps are also to blame. Shares of Macy's, the largest U.S. department store, fell13% in the second quarter of last yearto a nine-year low of $15.82, partly due to weak forecasts.

"Our fashion misses in key private brands in women's sportswear, as well as slow sell-through of warm weather apparel," forced the company to cut prices to clear excess inventory, Chairman and CEO Jeff Gennettesaid。

Forecasting demand is always a guessing game, but many retailers commit unforced errors by clinging to flawed forecasting processes, financial planning and analysis (FP&A) advisors told CFO Dive.

"Retail forecasting is a little bit of science and a little bit of art," Carlos Castelán, founder and managing director of retail consulting firm Navio Group, told CFO Dive. "We think of sales as the result of a transaction, but there are many steps before that. Understanding the sales funnel more broadly can greatly improve your forecast."

Here are the three biggest forecasting mistakes retailers make.

Hitting the forecast wall

Many retailers believe they are doing dynamic forecasting when they build in a way to update throughout the year, but in reality their process is static, FP&A advisors said, because what makes a forecast dynamic is not the update process but the time horizon.

"One of the mistakes CFOs often make is what we call 'forecasting to the wall,'" Steve Player, managing director of The Player Group, told CFO Dive. "A lot of times large companies will prepare a forecast, but if you look at what they're really doing, they're preparing a declining forecast."

A declining forecast, or what planning experts call "forecasting to the wall," is when a CFO creates a six-, 12-, or 18-month plan and follows up with updates as new data comes in, but the time horizon stays the same.

In a common scenario, a CFO starts a 12-month forecast in January, updates it quarterly, and keeps the time horizon fixed at Dec. 31. As a result, the company forecasts over increasingly shorter time increments.

"The first one is 12 months, the second is nine months, the third is six months, the fourth is three months, and then they jump out another year," Player said. "In that situation, all they're doing is updating their forecast to validate their budget targets."

If your sales compensation is based on hitting a target on Dec. 31, your behavior will be very different than if you ignore that deadline.

Brian Kalish, advisor

A better approach is to use a rolling forecast, where the time horizon is pushed back the same number of weeks or months with each update. If you use a 12-month forecast, in addition to updating your estimates with new data quarterly or monthly, you also push the time horizon out another month or quarter and remove the corresponding tail period. That way, the forecast stays 12 months, but you never reach the end, or the "wall."

FP&A advisor Brian Kalish, whose clients include retail companies, said a rolling time horizon enables executives to make decisions based on market dynamics rather than artificial targets.

"If your sales compensation is based on hitting a target on Dec. 31, your behavior will be very different than if you ignore that deadline," Kalish told CFO Dive. "The business is going to continue to exist on Jan. 1, so what happens is you start making economically suboptimal decisions to hit an artificial target."

Over the past year, Kalish has been helping a spirits company move from static to rolling forecasts to recover sales lost due to chronic inventory shortages.

"They were basically looking at what happened over the trailing 12 months, making a best-guess growth estimate, and assuming that's what they were doing," Kalish said. "It's not terrible. But the problem was out-of-stocks. All of a sudden, orders came in and they couldn't fill them efficiently."

By moving to a rolling forecast, he said, "it gave them better insight, not only because they could see internal information, but because we built the structure so they could start bringing in third-party information. All of a sudden, you can start to see growth."

Conflict of interest

Whether you do static or rolling forecasts, unless you change the incentive structure in the annual budgeting process, you are almost guaranteed inaccuracies in inventory levels, which in turn affect sales, advisors said. That's because most incentive structures are built around sales targets rather than accuracy.

"Budgeting is a little bit like a negotiation process," Karen Sedatole, professor of accounting at Emory University, told CFO Dive.

"The sales team negotiates a budget that is lower than what they think they can achieve because they want to make sure they hit the budget at the end of the year because they want to get their bonus. If I think I can sell 100 units, when I negotiate my budget target with higher-ups in the organization, I'm going to negotiate a target of 90 because I think I can sell 100. That way I make sure I hit the budget at the end of the year."

Player called the practice a "huge conflict of interest" because it almost guarantees sales estimates are understated in the initial forecast.

"The budget manager is not going to negotiate a stretch target or superior performance because he's going to be judged—his bonus is going to be based on it—so now his incentive is to give a conservative budget with minimally acceptable returns," he said. "If you were designing an internal control system, the first thing you would avoid is a conflict of interest, and here the budget puts it right in the middle of the mechanism."

That conservatism is a source of inventory shortages and lost sales because once a budget is approved and the forecast for the coming year is set, production planners start a process of underproducing inventory, which can lead to shortages at critical points in the market cycle.

A rolling forecast can solve some of the conservatism problem, but not the root cause. It can help mitigate inventory mismatches by giving executives the opportunity to periodically adjust demand estimates when data shows sales exceeding or falling short of forecasts. But changing production capacity mid-year is an expensive way to solve the problem.

Tying bonuses to forecast accuracy sounds like a good idea, but perhaps fewer than 20% of companies do it.

Karen Sedatole, Emory University professor

"Do we need to make some capital expenditure to get another production line running to meet demand?" Kalish said. "Do we think we really need to build a new plant? Do we go out and borrow money? Is that going to affect our credit rating? Are we going to spend corporate resources in those ways, or is there something better than building another plant? In the long run, that's not the most efficient."

The fundamental solution is to eliminate the conflict of interest in the first place by aligning incentive targets with forecast accuracy rather than sales targets, but advisors said most companies are unlikely to change their practices anytime soon.

"Tying bonuses to forecast accuracy sounds like a good idea, but perhaps fewer than 20% of companies do it," Sedatole said. "The reason is, we want salespeople to sell. So, we want them to have very strong incentives to sell."

Sedatole said she is working with a company that sells chemical products that has incorporated accuracy into its incentive structure, but it's only a small part of the sales bonus.

"When I talk to the sales managers, they say, 'We don't want to focus on that,'" she said. "'We just want to focus on hitting the sales target exactly.' So, even though they have it in principle, it's not really effective in practice."

Real-time data sharing

One of the most direct ways to improve forecast accuracy is to communicate sales data to executives as close to real-time as possible. By doing so, finance, sales, production, marketing, and other business operations can track performance in real time and make adjustments in real time. The more a forecast is updated with data with minimal lag, the more accurate the forecast becomes over time.

"Communication is one of the biggest forecasting problems companies face," Kalish said. "What happens is, information exists in silos. You're just not leveraging it."

The most practical way to accelerate communication is to move from on-premise software to a cloud environment, because cloud environments allow data to come in as it's generated and give executives access from anywhere.

"If you're using spreadsheets, get off of them," Kalish said. "That's killing people. There's so much technology that can really support the business."

Navio Group's Castelán pointed to Lululemon's success in leveraging real-time data. It was one of the first retailers to use radio frequency identification (RFID) technology to feed sales and inventory data from all locations into a central hub, allowing it to manage inventory across its operations in the most efficient way. If a consumer in Chicago buys an item online, that item might come from a store in South Carolina with excess inventory rather than a central warehouse.

"That's helpful from a forecasting perspective because you're able to use all the inventory in the network, and it also increases your inventory turnover," he said.

If you're using spreadsheets, get off of them.

Brian Kalish, advisor

Real-time sales communication among executives is especially important in retail sectors that change too quickly for annual sales forecasts. Kalish pointed to the cosmetics industry as a good example.

"A fragrance company might have a one-year horizon because [product designers] go to shows in Paris, learn about trends, and then manufacture products," he said. "There's no way to predict what that's going to be, so they have to be very good at reacting."

Another example is smartphone cases. "The sizes are changing all the time, and they can't have any variance," Kalish said. "They have to be perfect. So, they don't try to predict how Apple or Samsung will change the width and length of the phone; they just want to be very good at quickly changing their production lines." And that requires immediate access to real-time information.

Better forecasting

There's a saying among FP&A experts: All forecasting models are wrong, but some are useful. If you want to move your model into the useful category, here's how to start: transition to rolling forecasts if you don't already use them; eliminate bias in budgeting by aligning incentives with forecast accuracy rather than sales targets; and migrate to a cloud-based budgeting environment to generate and apply actual sales data as close to real-time as possible. That way, forecasts can be adjusted before inventory problems affect sales.

If you make these changes, your forecasts may still be wrong, but they might steer your company toward better performance.

This series is brought to you by BMO Financial Group, one of the largest diversified financial services providers in North America, with leadership in commercial, corporate, and investment banking. To learn more about their financial expertise, visit their websitehere. BMO has no influence over CFO Dive's coverage.