In the Part 1 of this article, we have already explored the causes of unbalanced inventory and the problems it leads to. Let's consider the most effective way to address this issue.
Balancing inventory: a simple solution to a complex issue
The solution to the "sales vs economy" dilemma often arises in a company with a new and ambitious purchasing and supply chain director. This person has most likely already worked in systems businesses and has learned that inventory optimization and balancing lost sales and surpluses is easily achievable only when supply chains are fundamentally transformed and automated using advanced solutions.
Let's imagine a hypothetical grocery supermarket chain, Cranberry Market, with annual sales of $500 million and an inventory value of $179.8 million (20% of the company's assets). EBITDA for last year was $30 million or 6% of revenues. Currently, the company has 20 people working on order generation, and they spend 50% of their time on it. They are assisted by an ERP system that builds forecasts based on historical data, while procurement specialists make manual adjustments based on upcoming promotions and seasonal trends.
Furthermore, let’s imagine that Linda, who previously worked for a supermarket chain as a purchasing director, takes over the Cranberry Market's supply chain department. Linda’s trial period strategy will include implementing an automated inventory management system because the previous company used such software, and she has seen all its benefits. In addition, Linda is confident that the new system will allow her to show excellent results during the six-month trial period already. Still, it is now essential to convince the management of Cranberry Market of these benefits.
What opportunities will come with the new system?
- Cloud solution
Linda will undoubtedly choose a SaaS solution, as a cloud-based monthly subscription solution does not require costly servers and can be used as long as they are satisfied with the results. This, in turn, guarantees the vendor's continued involvement in maximizing the short- and long-term benefits for the customer.
- AI-based demand forecasting
The new system will provide automated short-, medium- and long-term sales forecasting for up to a year (which is especially important for products with long delivery times, such as those from China or India). Sophisticated AI algorithms will consider most factors affecting customer demand at the SKU-store/channel-day level: sales history, seasonality, days of the week, trends, regional characteristics, upcoming holidays, promotions, etc. In addition, thanks to machine learning, such a system will continue to improve forecasts' accuracy over time.
- Fully automated replenishment system
An automated replenishment system will allow the company to generate and send orders to external suppliers or the warehouse centrally without the intervention of demand planners or store employees. The system will consider current balances, delivery lead time, in-transit quantities, product expiration dates, supplier terms, and schedules so that each item arrives at the right destination and in the correct quantity without creating overstock and ensuring optimal availability. In addition, it will avoid human error and free up the staff involved in order generation for more critical strategic tasks by letting the system deal with routine tasks.
- Efficient management of fresh stock
Unique algorithms will consider fluctuations in demand by day of the week, supplier conditions, packaging, and delivery times for perishable products. This will ensure maximum availability of fresh products daily, with minimal write-offs and markdowns at the end of the day/expiration date. In a large chain environment, this is an essential step towards lean and reducing waste and, therefore, improving operational efficiency.
- Multi-echelon inventory allocation
The algorithms of the new system will also take into account all the peculiarities of goods movement and generate highly accurate forecasts for inventory management at both regional and central warehouses. Demand for goods will be estimated for each store replenished through the DC. This will include the number of deliveries of goods by an external supplier to each store based on the DC delivery schedule, seasonality, promotions, and other factors, which will allow optimal and timely distribution of goods, regardless of the number of logistics links.
- Automatic support for assortment rotation
Linda knows from previous experience that assortment rotation, even at 30% per year, is more effective when done automatically. When new SKUs are introduced, the system automatically predicts the first order depending on its specifics. The new SKU is a replacement or an analog of an existing one, or it is entirely new for the network. This is the only way costly errors in assortment expansion and replacement can be avoided. In case of product withdrawal, the program will consider the supply chain structure to track inventory at all levels for efficient product withdrawal.
- Informative dashboard and deep analytics module
It is critical to keep a finger on the pulse of all key metrics to manage inventory on an ongoing basis effectively. Linda knows that, and she also knows that this requires having an informative dashboard at your fingertips, where you can see both operational and strategic KPIs in a few clicks. It makes it possible to notice and eliminate deviations on time as well as to build an effective strategy for the future. The new system should also contain a powerful analytics module and a comprehensive reporting system for maximum visibility and problem analysis at all levels.
What effects can we expect from the implementation of an inventory optimization system?
Referring to the example of a hypothetical supermarket chain, Cranberry Market, we think it is fair to consider two variants of events and, accordingly, calculations.
Option 1
Linda joined the company at a time when the "sales first" approach had long been the basis of the business strategy. As a result, product availability averaged 96% with a turnover rate of 35 days (compared to 28 days for a direct competitor).
While in Linda’s previous experience, the solution implementation resulted in a 3% increase in sales, she realized that even with already high availability, a 1% increase in sales could be expected through better forecasting and distribution. She set a turnover target of 30 days rather than 28 days because the system is "inertial," and it will take time to eliminate excess inventory without incurring additional costs. Calculations showed this would yield an annualized EBITDA gain of $1,914 million (7.7%).
In fact, the expected annual increase in net income in this case would be about $1.55 million or 17%.
Option 2
If Linda arrived when the company was in the “economy” mode and product availability was 92% with a turnover rate of 30 days, she expects to increase availability and thereby increase sales by 2% while maintaining and even slightly improving turnover by 5% to 28.5 days through the implementation of the new inventory management system.
Expected economic results:
EBITDA growth: $2.41 million (+9.6%)
Net profit growth: $1.75 million (+16%)
Bonus effect
In both cases, significant labor savings were expected due to the centralization and automation of replenishment processes. As we know, the company had 20 people working on orders, half of their working time when Linda arrived. At an average salary of $90,000 per year, this amounted to 900,000 annual costs.
After implementing automation, only 5 people spent 100% of their time dealing with orders. Assuming salaries remained the same, the resulting costs were only $450k annually. Naturally, these savings are conditional, as we are not discussing staff dismissal. The freed-up time can be devoted to strategic tasks and projects and establishing more effective interactions with other departments and suppliers, which will increase the overall efficiency of business processes.
Using these figures as a basis, Linda met with the system vendor to make a preliminary, albeit not exact, estimate of the amount of investment in the project. Afterwards, having calculated the ROI of the project, she was happy to see 500–600% and started to prepare a presentation for the company's management. There is no doubt that the calculations will convince Linda’s management to invest in the new inventory management system..
In conclusion
As we can see, the financial and operational efficiency of a retail business can suffer from many and often non-obvious factors, such as unbalanced inventory.
The world of retail is too dynamic and the amount of data that needs to be analyzed is so large that the experience, intuition of employees, and relatively simple models in Excel or ERP are not enough to make optimal decisions. It is very difficult to find a balance without good methodology and sophisticated mathematical models in the age-old battle between the strategies of "salespeople" and "economists,". Fortunately, today the role of the great equalizer in this matter is firmly occupied by AI or rather modern technological solutions based on artificial intelligence and machine learning. Such systems impartially make calculations and tactical decisions for people on the basis of complex algorithms, and also help managers to make important strategic decisions. These systems can simply “see” the key clues and patterns in the endless array of data that are ordinarily inaccessible to human eyes.
In addition, simple and approximate calculations show that even minimal effects of such a system give a significant increase in profit and allow you to recoup your investments in the first year of implementation. And in subsequent years, it provides opportunities for stable growth and expansion of the business, freeing up enormous amounts of money and fixing inefficient business processes.
