Combine different data models and sets to accurately forecast demands amid COVID-19

New Ideas in MarketingEssential news for marketers, summarised by YouGov
November 27, 2020, 2:25 AM UTC

Analysing social-media sentiment can help companies gather insights on consumer momentum.

The article states as consumers’ attitudes and behaviours change amid the pandemic, brands would do well in understanding these new patterns for better demand forecasting. Finding alternative and similar data models from past analogies like the dotcom crash or natural disasters could help brands forecast demands to a certain extent.

Alternatively, brands can create “ensemble models”, wherein predictions from different models can be combined for more accurate estimates. A/B testing the collected forecasts can help brands optimise their marketing channels, compare performances and test new simulations.

Tracking consumer data “near-real-time” can help brands assess consumers’ current behaviours and attitudes. For better accuracy and reduced store returns, retailers must ensure their data is infused with local knowledge. Seek advice from epidemiologists and trade associations for industry perspectives.

Read the original article

[7 minute read]