Why ERPs Provide Mediocre Levels of Demand Forecasting Accuracy
When it comes to ERPs, they are built to provide companies a standard solution for inventory management, sales forecasting, procurement, etc. For ERPs, they make money when companies want a customized solution so on top of the standard solution, if you want more customization, you will be charged separately for it. Over the course of a few years, ERP solution providers can collect in the hundreds of thousands of revenue from just 1 customer.
When it comes to demand forecasting, ERPs generally only take into account a few sources of data: 1) Historical Sales Data & 2) Inventory Data. Based off of these 2 sources, a demand forecaster/planner than also uses his/her gut intuition to decide how much existing product to replenish, when to replenish them, and when to buy new products.
The problem here is that ERPs aren’t built to factor in the 20–30 sources of data that influences demand forecasting. They factor in only a few sources and get you mediocre results. In this day and age, getting more precise accuracy around demand forecasting revolves studying 10–30 sources of data all in REAL-TIME including your historical sales data, inventory data, vendor data, seasonal/promotional data, marketing data, customer demand forecast data (if applicable), previous stockout/excess inventory data, transportation cost data, fulfillment center data, warehouse management data, accounting data, e-commerce data, retail location P&L data, and so much more.
At the moment, many demand planners are taking data from ERPs and manually building forecasting models (which is also influenced by their intuition) to get anywhere between 50–80% accuracy. That 20–50% improvement in accuracy can help companies exceed sales targets, avoid excess inventory, improve their customer service/lead times, better manage P&L, expand to new product lines, and so much more. Unfortunately, even though most companies aren’t happy with their demand forecasting efforts, they feel they have no other option and they don’t want to switch to another ERP solution because of switching costs, training team(s) on a new system, etc. The reality is that companies can improve their demand forecasting accuracy without replacing their ERP solution (I’ll get to that below).
Many companies are also using different ERPs for different purposes. One single company can be using 3–5 different systems that aren’t speaking to each other. The problem here is that when a problem or bottleneck takes place, it’s hard to identify the root cause of it. Just understanding the root cause and proactively tackling it is a nightmare that wastes a lot of employee time and effort.
Also, ERPs, especially those which are cloud-based tend to take a lot of time to record real time data into their database if the internet connection is not stable which delays data syncing. So sometimes getting real-time reporting is also painful. Overall, existing ERPs do a good job of presenting you hard/raw data but they don’t do a good job of understanding trends and recommending the steps you need to take to hit KPIs. This is where data scientists come in. They can analyze the raw data in your ERPs and understand trends much more accurately, especially those that are influencing your business outcomes.
What is PredictiveDemand90 all about?
At Lovers of Data, we have a system called PredictiveDemand90 which studies any and all data sources that impact demand in real-time. In 90 days or less, we leverage predictive analytics on top of your existing ERP solution to identify the optimal levels of inventory to meet demand with a high degree of accuracy. You’ll be able to exceed sales targets, reduce excess inventory/holding costs, and increase customer satisfaction without worrying about switching costs or investing in new software. If you’d like to chat about this, feel free to book a call here: https://bit.ly/3k6FSdk