What would you do with $1 trillion dollars? That’s 1 million times 1 million. It’s also how much money companies are losing due to out of stocks and overstocks according to research from IHL Group.
Sears, Kmart, Gander Mountain and more are joining the ranks of fallen companies unable to adapt to a new customer dynamic. They’ve failed to provide unique customer experiences, delivering the right products at the right time, without bankrupting the company.
Consumers have access to endless information, make decisions in moments, and act in seconds. Companies today must change their approach if they want to keep up and avoid becoming another victim of the Amazon effect. $800 billion of that $1 trillion of out-of-stocks and overstock losses are due to data, process or personnel issues.
Companies have been lulled into the mindset that inventory planning and forecasting will always be imprecise and not to be trusted, and with good reason. Traditional forecasting has not been very accurate. Due to historical limitations in computing power and data availability, inventory planners have been forced to rely on basic statistical methods – think, “What happened last year about this time? Something similar should occur.”
Customer behavior is far from traditional today. Market dynamics are forcing companies to take a new approach – understanding the customer, what influences them to buy, proactively planning and responding to changes rapidly.
Customer Transactions: The Key to Forecast Accuracy
Today, with the wealth of big data, including transactions and customer influence factors, your company can know buying signals and look for or even create factors.
Big data analysis has revolutionized the understanding of customer demand, providing visibility and precision for inventory planning on a whole new level. The days of death by averages, estimates and assumptions are over. Now you can not only better understand what influences buying behavior but begin predicting what and when customers will buy with unprecedented accuracy.
You can also determine the impact of various influencing factors, including advertising, price, promotion, social media, SKU affinity, display impact and even the weather.
The Old Way Doesn’t Work Anymore
To date much of the struggle to achieve highly accurate inventory planning and demand forecasts can be attributed to an over-reliance on statistical forecasting models. New challenges in the market today are compounding the risks associated with purely statistical methods:
- Multichannel environments and the vast array of customer buying options
- Demand-shaping activities that entice customers to buy, buy now or buy more
- The relationships of SKUs to one another, and how one drives demand for the other
- Multi-echelon demand environments which are more deterministic in nature
Today, demand is changing constantly. What drove customers to buy last month, may not even be a consideration by next month. Relational and causal factors can’t be accounted for by purely statistical forecasts that try to predict demand by analyzing data from the past.
A Better Approach: Demand Segmentation
Companies seeking to keep up in an increasingly digital age need a new approach and likely new technology. To serve today’s customer, you must analyze every transaction individually and segment demand based on causal factors, deterministic considerations and on stochastic (or statistical) demand across every channel, location and item. This is all possible with today’s technology.
Demand segmentation refers to creating groups of demand types based on common characteristics. Demand segments could represent seasonal, promotional, loyalty program, social or other influences.
By capturing customer transactions and the influencing factors around them, it is possible to get the best of both a statistical or stochastic understanding of demand, and a causal understanding. To the extent that causal factors cannot be identified, or don’t show substantial influence on demand, that portion of data is relegated to a statistical forecast. Demand is segmented to apply the most effective technique, assuring that the combined forecast is as accurate as the data will allow.
Today’s technology can enable in-depth segmentation for inventory planning. This can help you reduce the variability of demand forecasts, gain insight into demand-driving factors and provide exactly what customers want and when they want it, at lower costs for your company.
What Should You Do?
To create the most cost-effective plan to meet demand, you need a precise prediction of what that demand will be. Take these steps to improve forecast accuracy:
- Analyze every single customer transaction individually
- Understand the factors that influenced customer behavior
- Attribute causes to demand and create a separate segment of the forecast for causal demand going forward
- Create segmented forecasts for every single item, location and channel
- Revise your forecasts daily (or even in real time) to account for trends and disruptions
Understanding your customer better and precisely predicting future needs is the first step towards greater accuracy in inventory planning and forecasting, keeping customers satisfied and shareholders happy.
Rod Daugherty is the Vice President of Product Strategy for Blue Ridge