Why your forecasting should be able to cope with COVID-19

Willem Torfs
December 5, 2020

Why your forecasting should be able to cope with COVID-19

Bottom-up forecasts allow you to gain insight into your portfolio. In times of COVID-19, agile forecasts are key to adapt to change.
Willem Torfs
December 5, 2020

Why your forecasting should be able to cope with COVID-19

Willem Torfs
December 5, 2020

2020, started with a bang. COVID-19 introduced challenges on a scale humankind never experienced before.

Today, May 12th, almost a 3rd of the world population lives in quarantine. The energy market is turned upside down, with abrupt changes in consumption profiles and prices. Many energy retailers are left wondering how their portfolio will evolve in the next years. The time has come to truly leverage forecasting tools and smart meter data to map portfolio performance. Traditional top-down forecasts don’t provide the insights needed to assess individual impact of large accounts on the overall business. In this blog we’ll dive deeper in what every utility retailer should prioritize when adopting a bottom-up demand forecasting solution.


Traditional top-down forecasting techniques roll forward the historical consumption of a retailer’s portfolio to estimate future demand. Industry and market development variables are then applied to adjust for expected events. More advanced applications incorporate geographic information into their forecasts to identify location-based trends. But, even at this level, data is aggregated, flattened and removed of any detail. Top-down forecasts work well for stable markets and limited portfolio diversification such as B2C portfolios, at least for now. With a highly volatile market introduced by electrification of the fuel mix, more complex consumer profiles and the introduction of digital meters, utilities are challenged on using more detailed forecasting tools throughout their portfolio, and starting with B2B.

Many forecast specialists don’t have the time to do predictions on energy behavior or wait for full blown impact evaluations to measure change and incorporate these in their forecast models. The only way to accurately map impact of COVID-19 on the current energy landscape is by focusing on building up the forecast from the bottom-up. Meter by meter, client by client. The question is if COVID-19 is now the final push to look into having bottom-up forecasts upgraded to use them as a tool to actively assess your portfolio and look for opportunity.


Bottom-up forecasts truly leverage the advantages of interval meters and in combination with industry knowledge and customer input it can provide deep insights in site-level performance of each client. When done right, the insights obtained is only limited by the data available. In times of peace, an accurate forecast will result in significant savings in your hedging department, while keeping risk under control. But what happens if unforeseen events drastically shake up the market? Forecast methods, previously so accurate, now suddenly need a complete overhaul to cope with unprecedented change. While strong at running proven models (albeit often slow), existing solutions have one thing in common; they lack agility to cope with events like COVID-19.  

At Gorilla, we believe bottom-up forecasting hasn’t been used up to its full potential. Many large utility players have adopted external agencies to perform these data heavy calculations for them or have built something in-house. Forecasting a portfolio of 100.000 meters takes hours or days, resulting in long wait times and reduced forecast agility, and applying new insights every hour or day to fuel that forecasting algorithm is cumbersome at best, and often not even possible.


Agility is the key today, when improving long-term forecasting. Imagine being able to rerun your entire portfolio forecast instantly, assessing any new update from an individual customer and applying industry forecasts as they roll in.

At Gorilla, we believe agility in a forecasting solution can only be achieved through a combination of Configuration, Speed and Accuracy.


A utility retailer should be in charge of their bottom-up forecasts. Any parameters requiring deep industry knowledge or customer insight should be exposed to the retailer so that these can be tweaked on the fly. Certainly, in times of COVID-19, the knowledge extracted from your conversations with high-end customers will heavily impact the accuracy of your bottom-up forecast. It would be a loss if this information is not easily inserted in your risk management tools.

2.     SPEED

Forecasting should not be a static aspect of a utility forecast strategy; the technology is present to run forecasts for 100 000 meters in less than 5 minutes. Enabling retailers to run, test and adapt to achieve the most optimal and accurate result. A true understanding of the exposure to risk is to be able to compare tens of scenarios every day, updated with the latest information within minutes after the information comes in.


The best forecast is the one that is tested, validated and adapted the most. With fast calculation times and the means to change parameters based on business-critical knowledge, a retailer can assure higher accuracies than any other external party. When not limited by time, bottom up forecasts can be used as a powerful segmentation tool.  They will help you gain insight in your customers and allow you how to create win-win situations based on data.

The impact of this global phenomenon will only get clear in the next couple of months, many think the world economies will change forever. Now is the time to evaluate your portfolio and assess the individual contract performance of key accounts. Tomorrow, you’ll think about how to arm your company better to be more resilient for future events like this. One of those weapons is a better and more agile forecasting tool.

Learn more about what Gorilla can do for you here or fill out the form below.

Share this post