Retailers are operating in an era when margins are under pressure, and consumers – and investors – expect strict adherence to robust sustainability practices. So, it follows that accurate demand forecasting can be an indispensable tool in reducing waste, optimising sales, and managing the selling price throughout a product’s life cycle.
Take multinational FMCG manufacturer Grupo Bimbo as an example: The parent of premium brands including Sara Lee, Entenmann’s and Thomas operates 59 bakeries in the US alone, employing more than 20,000 associates and distributing products through 11,000 routes that service the majority of store outlets offering baked goods in the US.
The perishable nature of baked goods demands rigorous stock management to ensure the freshness that shoppers expect. That means manufacturers like Grupo Bimbo walk a fine line between waste due to overstocks and lost sales from out-of-stock.
After deploying Zebra’s Work Cloud Demand Intelligence, part of its Product Lifecycle Management (PLM) suite of solutions, Grupo Bimbo managed to reduce product demand forecasting errors by 30 per cent – even throughout Covid-19, an era when no one had any historical reference points to predict demand given the pandemic was a once in a 100-year event and demand for baked goods was volatile.
“That is a big reduction,” Darren Bretherton, regional sales lead software solutions for ANZ at Zebra Technologies, tells Inside Retail. “In their industry, where food wastage is a problem, it has had a massive impact – both on their own output waste and the effect on their suppliers.
“That’s got to have a pretty dramatic effect on the bottom line if you cut out 30 per cent of product that is made and not sold.”
Every retail manufacturer is focused on environmental, social, and governance (ESG) right now, and Grupo Bimbo aims to reach net zero by 2050. “If they are able to reduce overproduction, they are not wasting energy to make products that aren’t going to be sold, and it can minimise its food waste. That is going to have a massive flow and effect for them, their customers, their suppliers, the entire supply chain.”
Zebra’s demand intelligence helps companies project within 80 per cent accuracy how many products are likely to sell next month based on past sales history and factors that impacted sales, such as holidays, climate events, social media trends, and even geopolitical issues that may affect supply chain routes – right down to every SKU. The technology can incorporate localised events along with seasonality, pricing, promotions, and product lifecycles by combining projections from the supply chain, finance, sales and marketing teams into what Zebra calls a ‘unified demand signal’.
For Grupo Bimbo, deploying Zebra’s Work Cloud Demand Intelligence resolved a major organisational obstacle: coordinating a dozen fragmented regional bakeries and better connecting with shoppers who expected the last two slices of their loaf of bread to be as fresh as the first. Outdated operations that relied on desktop calculators and manual spreadsheets needed modernising, and the company’s front-line workforce needed to be empowered with tools to maintain the highest quality, from the production line to store delivery.
The forecasting methodology considered more inputs – many captured directly from retailers and route drivers – and went beyond the historically-driven base and promotion demand to include real- or near-real-time inputs, including weather, local events, store stocking constraints and actual point of sale (POS) data. The AI-powered demand forecasting and predictive ordering technology helped right-size production and localise delivery down to an SKU/store/week level by factoring in such granular-level data.
From baked goods to fashion to cosmetics
Bretherton says the tool is helpful for any companies bringing new products onto the market, especially where there may not be any relevant sales history.
“For example, an apparel company may have a new fashion item but is not sure if it will resonate with consumers. With no historical data to go on, the brand can use Zebra’s Work Cloud Demand Intelligence to analyse a lot of resources, such as influencers and what is trending on social media, to help give them some level of determination around what future volumes might look like.
Estee Lauder is another manufacturer that uses Zebra Work Cloud Demand Intelligence globally to predict demand for its products, and what the next season’s hottest colours and makeup shades might be.
They look for insights into what Kim Kardashian or Blackpink might be talking up as the next big thing, so the brand can adapt and ensure it has its supply chain geared up to get that product in the right volume in the right stores, and thus in the hands of as many consumers as possible at the peak of the interest.
Estee Lauder is doing this at a global level, then by country, and then within a country down to individual stores in order to get maximum sales.
“You are making sure that any data you’re pulling into the solution has been cleansed of any anomalies that are going to skew the outcomes. So it can ignore any outliers where there may be just some outright errors that can happen in recording sales transactions, for example.”
Diving deep
For a bakery, demand forecasting can accurately predict how many loaves of bread you will need next week, while with a makeup brand, you can dive deep down into how many different types of makeup, or how many different shades – all related to demographics – by country, city, suburb and store, says Bretherton.
The technology works equally well in predicting and optimising sales online, in stores, on third-party marketplaces – and even direct-to-consumer.
“Each market is unique, even perhaps in the same country. In the winter months here in Melbourne, maybe consumers prefer darker shades, but in Queensland, where it’s sunnier, consumers might want something different. Again, the software allows them to have those understandings, and it takes brands to a very micro level, enabling them to optimise their entire sales forecasts and volumes by fine-tuning their products to meet the specific market they’re trying to deal with.
Where does a company start?
Creating accurate demand forecasting using the Zebra Work Cloud Demand Intelligence solution begins with data. Clean, accurate data is essential.
Bretherton also counsels that manufacturers and retailers have an understanding of what they are trying to achieve, knowing upfront what their end goal is and the challenges they are trying to overcome.
“All of the steps to achieving that goal and understanding the customer, their marketplace, where they sit in that market, what issues or external factors influence any future volumes, what their competitors are doing, how the weather impacts sales, and so on – all in their space,” he explains.
“So the company needs to articulate clearly where the data could come from, and then we’ll find a way to bring that into the solution so that it is factored into the results.”
