Here's a simple but powerful insight: a WFM forecast based on a four-week moving average can sometimes outperform complex models simply because when patterns don't change much, history really does repeat itself.
In case you missed it Philip Stubbs's full episode can be found:
This hits at the heart of why we sometimes overcomplicate workforce planning:
Complex models can fail not because of their design, but because they depend on other forecasts (like sales projections) that may themselves be inaccurate
The most sophisticated model in the world won't help if your input data is unreliable.
What's missing from many forecasting discussions:
The operational cost of complexity... simpler models are easier to maintain, explain and adapt when things inevitably change
Your workforce planning approach needs built-in flexibility, because perfect forecasting doesn't exist... "the best contingency plan is the one you never need, but always have"
The irony of my 20+ years in WFM is that while I've built some incredibly sophisticated forecasting engines, I've learned that simplicity often wins the day. The real art isn't just building accurate models but knowing when complexity adds value and when it just adds noise.
What's your experience? Have you found simple models outperforming complex ones in your environment?
Share this post