Demand forecasting in uncertain times — time for a new technical solution
DataSeries highlight:
- With the pressure of the coronaviurs pandemic, it is becoming harder to make accurate demand forecasts — we need new methods that track the uncertainty that the pandemic brings.
Whilst the spectacular dynamics of the last month’s consumer stockpiling may look to have calmed down, there is no doubt that the technology and processes for forecasting retail demand and supply need to change for meeting needs in both ongoing crisis and longer-term. The effects of the COVID-19 pandemic will have differential medium and long-term consequences on consumer CPG demand and product preferences and there will many different phases during this global crisis. Consumers food choices are changing (e.g. chilled to ambient/frozen), brand choice is changing (e.g. entrepreneurial to established, trusted brands), pack choice is changing (e.g. convenient to bulk) and value choice is changing (e.g. growth in the private label). CPG companies and retailers will need to adapt their forecasting and planning models to manage efficiently across this dynamic journey.
While the pandemic unfolds and impacts society and economy, new diverse data at different frequencies are generated and the relative impact of contextual factors becomes more than historical, established patterns making existing forecasting methods unreliable. So, it is important to analyse the problem from the perspective of data, methods and the required actionable insight for timely, quality decision making in the new scenario.
A capability that dynamically learns and predicts using all market and consumer data including COVID-19 related data and its impact on consumers
The current situation requires
- A much-improved ability to anticipate consumer preferences and buying behaviour at local and granular product segment levels in the rapidly evolving phases of the crisis.
- The utilisation of this dynamic insight to create predictive demand scenarios and localised forecasts
- Evolving effective replenishment plans within the constraints of the current supply chain including potential collaborative interventions within suppliers and retailers
The solution lies in an automated augmented capability that learns from all relevant data including COVID-19 related data and its impact on consumers, generates a comprehensive granular forecast and enables retailers, consumer goods companies and delivery partners to simulate various scenarios for most effective collaborative solutions for getting the product in the hands of the consumer.
A comprehensive approach to AI-augmented decision making in the new world
It is clear that the ability to use all data to derive future scenarios of changing demand and evolving driving factors is important. It is also important to be able to use these in actionable decisions and optimise business processes in time. This is where a comprehensive approach as outlines below is important for transformative impact now and in new normal:
- Use all relevant market and consumer data including the COVID-19 specific constraints, the system will ensure the mitigating impact of any future anomalies as well.
- Use predictive analysis and scenario simulation to identify the most reasonable and risk optimised decisions
- Use AI solutions which can help make sense of chaotic and diverse multi-level data
- Augment human judgement and automate collaborative decision making to make it faster and scalable.
The need is business model agility enabled by a robust AI & data core.
It is not a new business model, rather the agility and innovation in the business model that will be important to survive and thrive. The new retail would be founded with a much more robust competence to understand, anticipate, adapt and respond to evolving market events and consumer behaviour by using AI-augmented, automated data analysis capabilities.
Singular Intelligence Product
Singular Intelligence innovation lies in a predictive simulation of consumer behaviour with the application of novel AI modelling techniques harmonising diverse COVID-19, market and consumer information and using the outputs in our existing Singular Intelligence AI-based platform for more accurate, granular, rich demand forecasts than are possible using prevailing techniques. The product modules, designed for direct business use, in small to large Consumer goods and retail companies, have a framework of Predictive market exploration, Alerts and Recommendations and Scenario Simulations to facilitate agile, augmented, automated decision making.
Contact us to learn more, discuss your challenges, potential solutions and plan a trial.
Sarvesh Kumar, Founder & CEO, Singular Intelligence
Email: sarvesh@singularintelligence.com