Coronavirus Forecasting: The New Normal

Written in the first few weeks of the 2020 coronavirus lockdown, this article has a few thoughts on how to regain some form of accuracy in your organisation's forecasts, even whilst we remain in lockdown.

Introduction

At the time of writing (late March 2020), the UK feels like it's in a surreal freefall: deserted streets; police drones following dog walkers; breweries making sanitiser instead of beer. Every day brings new restrictions, budgets or branding on the front of the Downing Street lectern - and sometimes all three.

It's not just wishful thinking to say that this can't go on forever. However crazy it might feel now, humans are exceptionally good at adjusting to a new status quo. We might not like it, but we will get used to it.

The latest we're hearing is that the current restrictions - one form of exercise a day, no unnecessary travel, non-essential retail & leisure businesses closed - are likely to last for 12 weeks, followed by another 12 weeks of partial easing.

The New Normal

This will surely change as things evolve, but either way the restrictions are here for some time to come. Long enough that this will become the 'new normal': whilst many elements of daily life will remain surreal, a degree of consistency and predictability will return to people's behaviour.

For those of us interested in forecasting, the question is "how consistent and how predictable"?

The answer will of course depend on the sector you work in. Restaurants, theatres, bars? Not so much. But for other organisations trying to predict demand, it can help to ask a few questions of your data:

  1. Are all parts of your customer base / product or service range behaving the same? When you dig into the data, are some components unchanged?

  2. Even amongst those areas that are behaving differently, are any of the old underlying patterns still in evidence?
    • Seasonality (by hour, day, week, month)
    • Weather
    • Product availability
    • Marketing activity
    • Geographic variation
    • etc.

A Prediction

We're going to stick our necks out and say that even in these craziest of times, some of the old truths will persist. Sausages will sell when the sun shines, A&E admissions will peak on a Monday morning and the returns on the same old marketing promotion will continue to diminish.

The challenge is finding those signals amongst the noise. But if you can do so, it gives you and your teams something familiar to work with. It's all too easy to get into the mentality of 'there's nothing I can do, we're at the mercy of coronavirus', which can in turn harm your performance and your ability to come out of other other side firing on all cylinders. Being clear on what you do and don't know can focus minds and help you make the best of what is undoubtedly a horrible situation.


If you'd like to learn more, these articles might be of interest:

Thanks for reading.

What we do

Skarp uses machine learning-powered predictive analytics to generate accurate, automated demand forecasts - and an explanation of what is actually driving performance.

By removing uncertainty and quantifying the impact of factors affecting performance, Skarp can reduce costs and improve customer satisfaction.

We offer a fully-managed service, designed for organisations with limited in-house data science resources.

IDENTIFY PERFORMANCE DRIVERS
OPTIMISE INVENTORY & STAFFING
MEASURE MARKETING ROI
Find out more

Get started for free

There is no setup fee or minimum contract term with Skarp, and we offer all new clients a proof of concept free of charge. We believe the accuracy of our forecasts will speak for itself.

Book your demo