Three lessons in managing supply-chain data

On the surface, it seems like an easy recipe for success. Step 1: Sell a product in a growing segment. Step 2: Sell that product in fast-growing economies. Repeat.

And, on the surface, that’s how L’Occitane en Provence has done it. Its skin care, body and bath products — particularly those that feature natural ingredients — are riding a wave of demand for natural and organic consumer goods. The company’s focus on emerging markets such as China and Russia helped boost global sales to €1 billion (about $1.3 billion) for the 12-month period ending March 31st, up 14.2% from the previous year.

But beneath the “where” of this success story is a more complex tale of “how” — that is, how a massive technological update is helping L’Occitane manage that rapid growth, and how it may help the company meet an ambitious goal: doubling its revenues within five years.

“Our investments today are not only helping us to manage the demand today, but to manage tomorrow’s business, which is going to be much bigger,” explained Aidan Goddard, FCMA, CGMA, who is the CFO and COO of L’Occitane’s Asia-Pacific operations.

During the past three years, Goddard has helped implement technology that improves the flow of data that help the company more accurately gauge demand — allowing for improved coordination up and down the supply chain, including the French farmers who grow lavender used in lotions and other products.

The upgrade follows years of unwieldy growth by L’Occitane in the 1990s. During the ramp-up, different divisions used different systems to track point-of-sale data to guide forecasts. Incompatibility bred inefficiency. “As the company was getting bigger and bigger and bigger, it was just getting more difficult to pull everything together without having a unified solution,” said Goddard, whose division contributed about half of L’Occitane’s global sales during the last nine months of 2012.

L’Occitane’s shift to a more unified system offers several lessons on how Big Data, when harnessed, can help a company cut costs and inform sales and marketing strategies, and how the role of finance is evolving into a key strategic partner.

1  Quality data are good, but it’s better when the numbers talk to each other.

Unlike many of its competitors, L’Occitane manages its own manufacturing, product development, distribution and sales to about 1,200 company-managed stores, which make up about 75% of group revenues.

One of the company’s major challenges had been getting information from its stores to its factory. But a new system — in which forecasting software, regularly updated with actual sales data, produces better future projections — is strengthening an important information cycle.

Armed with a more accurate, real-time understanding of what its customers want, when they want it and where, L’Occitane can more easily tune its manufacturing, shipping and retail processes to make them more efficient. The data loop works like this:

1. L’Occitane uses point-of-sale software to collect stock-keeping unit (SKU) data on every purchase made at each of the retail stores it manages.

2. Every night, the point-of-sale data are transferred from each store to a central business intelligence data warehouse in France. The data are processed and sales reports are generated, allowing the company to review the daily performance of every store, and to measure growth versus last year and against the budget.

3. Local marketing teams use the sales data and forecasting software to create monthly sales forecasts at the individual-country level. The sales forecasts are based on years of actual sales data. The forecasts are also based on the selection, at an individual SKU level, of the most appropriate statistical model, including “level shift”, “double smoothing exponential”, “polynomial regression” and “linear regression”, among others. The models are designed to cope with the differences that may arise due to the actual product lifecycle stage of any SKU, or due to seasonality impacts. Because the sales data are updated daily, the forecasts are dynamic. So the company uses a predetermined date each month to capture the forecast.

4. Local supply-chain teams then turn the sales data into purchase forecasts on a per-item basis, looking ahead 12 months. This demand resource planning (DRP) process also results in the routine creation of purchase orders at a subsidiary-company level. The DRP process also captures details of actual inventory on hand as well as inventory in transit to produce the purchase forecast. If inventory runs below certain “safety” levels, it will add to the sales forecast. And if it is above safety levels, it will deduct from the sales forecast.

5. Consolidated purchase forecasts are then fed into the factory manufacturing planning systems, allowing the factory to determine how best to schedule production, delivery and procurement of raw materials from suppliers for the next 12 months.

6. Finished products are transferred to a central warehouse in France. Upon receipt of actual purchase orders, which should correspond with purchase forecasts, the products are shipped to local-country warehouses. From there, they are sent to the local stores. Details entered into the local point-of-sale system can then show real-time availability at the store level. The loop is closed when the products are on the shelf for final sale. The process begins again when a product is sold to a customer.

Throughout the process, L’Occitane is closely monitoring accuracy. Sales and purchase forecast data are compared monthly with actual sales and purchase data. The company aims for a minimum 75% sales forecast accuracy (weighted), and a 90% purchase forecast accuracy. The feedback focuses on variances and enhances the learning process while promoting responsibility.

All of this informs the inventory equation, which is critical given the relatively short shelf life of the company’s products, particularly those that include perishable natural ingredients.

The process has helped the company improve product availability while decreasing inventory levels — without significant shortages.

2  Software is the tool, but people drive the process.

Much of the system is automated, but finance professionals are critical to ensuring its accuracy and making the data work within the company’s strategy.

“Software is a tool, but it’s the people who are actually using the software that bring the results,” Goddard said. “The software, if it’s not properly calibrated, not properly used, will not necessarily give a good result.”

When it comes to sales forecasting software, for instance, forecasters are responsible for selecting the appropriate statistical models to achieve the most accurate forecasts for up to 18 months. Every month, the finance team compares actual sales with the forecasts of the previous six months, three months and one month to gauge accuracy. The reports help the company decide which products to develop and which to discontinue.

Finance can also drill down to work out what the data mean for inventory at the local level, enabling the company to make sure it has enough capacity in warehousing or in production to manage scenarios, such as spikes in demand around holidays.

As with sales forecasts, the accuracy of purchase forecasts is measured each month, and a root-cause analysis is performed to determine the cause of variances and create remedial action plans.

“Warehousing is expensive. Over-production is wasteful and expensive and leads to write-offs, which we want to avoid,” Goddard said. “At the same time, we want to make sure that we don’t run out of stock and that we always have enough product to sell. So that’s where finance brings a discipline, brings a control model to that whole process.”

It’s increasingly important for people on the finance team to communicate with other divisions what the forecasts and budgets say, what’s most likely to happen, and to question things when they don’t look right.

Finance is also expected to think more analytically and be able to explain and interpret key numbers for senior management, Goddard said. Many finance executives tend to focus too much on managing the budget and costs, and not enough on honing their financial analysis skills.

“A key ability is to interpret financial information and tell the Finance is also expected to think more analytically and be able to explain and interpret key numbers for senior management, Goddard said. Many finance executives tend to focus too much on managing the budget and costs, and not enough on honing their financial analysis skills.

“A key ability is to interpret financial information and tell the story, and to have a degree of courage to make predictions,” Goddard said. “Especially at the board level, finance executives should be very strategic.”

He added: “We need people who can go and talk to people in a sales function, in a marketing function, in a supply-chain function to understand how their work is evolving and what that means in terms of future revenues and future costs for the business. … Analysis begins with asking yourself questions. And if you can’t answer them, it means going and talking to someone.”

3  Keep it fresh and let the customer be your guide.

L’Occitane normally carries about 500 regular products at any given time. Each year, the group introduces about 100 new products and discontinues about 100 existing products. The goal is to improve or remove every SKU over a five-year period. In addition, the company produces special sample sizes, the sales of which must be co-ordinated with the sales of regular products.

“If people pass our stores every day and see the same products in their window, they won’t get excited,” Goddard said. “They won’t get interested. So, we have to change and change and change. Even when we have very good products, we want to continue developing those products and offer better versions to customers. It also gives us an opportunity to trade up in terms of price and margins as well, and grow into higher-value segments without actually taking away existing segments.”

Customer point-of-sale data help L’Occitane determine which products to develop and which ones to discontinue. The data also help the company make decisions about the design of storefronts and how the company sells to customers who don’t live near a bricks-and-mortar location. “Most countries offer the ability to purchase products over the internet, and we want to understand these customers as well by using similar processes,” Goddard said.

The company can also better understand customer preferences through data collected from customer relationship management (CRM) programmes. Take L’Occitane’s VIP programme. Every time a VIP customer makes a purchase, he or she collects points on a loyalty card. The points can be redeemed for a free product at a later stage. Such programmes are popular with the company’s Asian consumers, many of whom are keen to know first about new product launches. That helps the company tailor how it reaches customers and how it advertises new products and promotions.

“The data can tell you whether the demand is growing for a product, whether it’s static, or whether it’s falling,” Goddard said. “So, if demand is growing, that basically informs the whole supply chain and factory production. … It also helps marketing to make decisions about that range. ‘Should we add to that range by extending the line?’ ‘Do we need to make a change now?’ ‘Should we start thinking about phasing out?’ ‘If we’re phasing out, what’s going to replace that?’ So, a lot of life-cycle decisions are made based on the information.”

Implementation tips

Aidan Goddard occupies two roles in his region’s C-suite: CFO and COO. The data-integration project he has overseen at the Asia-Pacific division of L’Occitane en Provence shows how the roles are increasingly inseparable.

“Supply chain, IT and finance are very closely linked,” said Goddard, FCMA, CGMA. “You can’t have a supply chain without good IT systems, and the IT systems must be appropriate. You cannot design IT systems or information solutions until you actually have a model of how your supply chain should work in terms of data requirements, data processing and information generation. … [And] there’s lots of information that finance can generate to steer people.”

Goddard offered these tips on how to manage a successful implementation:

• Understand your business model and ask yourself if that’s the model you want, you need or makes sense for you. Build your strategies on a solid business model and think five years ahead.
• Determine what information you need to help you achieve your strategy. Then you can start to talk about how you build systems to process data.
• Determine how much you want to spend — and don’t just buy a product because competitors have it. “It’s very easy to spend a lot of money on IT,” Goddard said. “Once you make a commitment, once you make a decision to go and start, it’s hard to unravel.”
• Make sure your finance team is integrated into all of these important decisions, and ensure that they can help calculate the impact on the business of the investments, both in terms of benefits and the cost going forward.
• CFOs and other finance team leaders need to be hands-on during an implementation and operations. Check on the progress every day. Consider having formal reviews each week.
• Get the best people around you when you’ve got difficult projects, and make sure that the target is clear. A manager is only as strong as the team members.