Demand Planning and Forecasting – WHY LESS IS MORE WHEN DRIVING TO BEST RESULTS (get to 90%+ outcome with 25% of efforts and costs)

What is demand planning and forecasting?

Demand planning and forecasting is the art and science of predicting customer demand for products and services, and ensuring the right product is available at the right time and right location to meet demand. A smart demand plan considers historical data and trends, while also feeding in variables related to changing consumer preferences, trends related to market and macroeconomics, supply chain disruptions, sales and marketing spend for the company, new products introduced, growth plans, and more. Ensuring the right product is stocked in a timely fashion prevents loss of sales opportunities. Reciprocally it ensures that items are not overstocked, a scenario that can increase costs and hurt margins.


Why is demand planning and forecasting so essential for organisations?

Anything that has an impact on any one of revenue, customer retention, and margins assumes huge importance to an organization. A demand forecast influences all those three important financial and company health indicators. A lack of a smart demand plan and not having enough inventory can mean losing sales and customers, while overstocked situations can have disastrous impacts on cost management. The difference between having an intelligent demand plan and not having one can mean the difference between success and death.


Why is it difficult implementing the right demand planning and forecasting tool?

  1. Costs: Demand planning tools can be quite expensive, running from a few thousand dollars a user annually to tens of thousands of dollars per user. Added across users, and building in custom integration and development costs, companies can spend hundreds of thousands to millions of dollars to build a smart demand plan and forecast.
  2. Understanding and listing the variables that drive an accurate forecast: Supply chains are complex, and every company’s supply chain is unique. Building a smart demand plan is not just about the system chosen but also the ability to understand and list the historical and macroeconomic variables that would impact the demand plan. This requires subject matter experts from within the company to work closely with the software vendor.
  3. Implementing integrations and ensuring quality of integrations: A demand plan needs to read in historical and real time sales data across selling channels, inventory data, purchase orders placed, purchase orders in transit, lead time information for production by item type, and more. Assembling and aggregating these varying data points involves ensuring successful integrations with different systems and data sources. This can be time-consuming and expensive.
  4. Predictive algorithms used in demand plans: Most demand planning tools use complex statistical algorithms, and a combination of them to inform a company’s demand forecast. Selecting the right model or combination of models can take time, since the validity of the models needs to be ascertained against historical data. The added burden to consider is that the choice of one model cannot be constant. The models need to be constantly reevaluated as historical data changes with new products and market trends.
  5. Building in change and improvements: Adding to point (4) above, all assumptions and variables about the business, the economy and the demand plan must always be questioned and challenged. Without these edits, a smart demand plan can quickly devolve to a plan that misleads and that can result in very costly mistakes.


How to get “best results” from your demand plan and forecast and how to keep it simple for best results. Less is more.

With demand planning and forecasting I am huge proponent of less is more. There is of course a value and premium to varied and complex statistical models, but they take time and money to build and more time and money to reassess. The principal of applying 25% of the average effort and cost towards getting 90% of the outcome is the one to practice and celebrate. In fact, I would argue that even with those that have big budgets, it makes more sense to start simple. The layers of complexity for the incremental gains in accuracy should be phased in later.

  1. Simplicity of implementation, a Step 1 plan to get to 90% of outcome with 25% of time and costs.
    Many businesses run their demand plans and forecasts on excel sheets. An intelligent step one to aim for is to grade up from excel sheet work and digitize the demand plan. Instead of spending months or even years integrating different systems, a step one here would be to import data sets at daily frequencies. Import routines are faster to implement and far less expensive, though not as real time as an integration. Once data sets are imported into the demand planning platform, rule engines can be built in in the place of statistical models. This combination of aiming to digitize workflows, importing data, and finally building rule engines on the aggregated and connected data sets delivers a Step one demand plan and forecast that can often deliver 90% of the output at 25% of the efforts and costs.
  2. Number of variables used.
    Some of the most detailed statistical models use hundreds and even thousands of variables to inform the demand forecast. While the added variables potentially provide incremental intelligence to the forecast, often that added value can be an illusion or worse counterproductive. Not only does it take more time to align on this long list of determining factors, but the upkeep and revision of such a complicated model could end up becoming a never-ending cost and time investment. Using the same 90%/25% principle, identifying the most important vectors that influence a forecast should be sufficient to get to a reliable level of intelligence. Perhaps that is 10 or 20 most important variables and not a 1000, but such a list has the highest influence on the plan, and much less expensive to maintain. Keep in mind that that much of the data required to populate the additional variables used in a demand plan might need to purchased from a syndicated data source, further driving up costs.
  3. Number of integrations and imported data sources.
    Related to the above points and in the spirit of simplification, there is also value in starting with fewer and the most important external data sources. Most critical historical data sources are sales across channels, inventory, production data, lead time information, supplier data, and purchase orders. Starting with integrating and important data sets that drive 90% of the plan should be the step 1 focus.


In conclusion, boiling the ocean and spending millions at the start when implementing a demand plan is not required, and is often counterproductive. Less is more, get to 90% outcome with 25% invested, and then increment up. Good luck implementing your demand plan.


Suuchi Ramesh
Founder, CEO
Suuchi Inc

To find out more about our integrations please visit or write us at


Explore More

Blue Line Icon on