How Big, How Long, How Much? – Providing Quick Answers by Forecasting with Data

Main Room November 7, 2017 8:30 am - 9:30 am

Bookmark and Share

Troy Magennis


About this session:

The core of this session is how to answer the three most ubiquitous forecasting questions without detailed estimates (and often without estimates at all):

  • How big (is the feature without doing detailed analysis)
  • How long (will it take if we just delivered this feature)
  • How much (will I get)


Understanding why forecasts fail is also important. We will discuss the top three reasons forecasts fail to match reality from my painful experiences and will challenge the assumption that work complexity and effort correlates with delivery time.

By attending this talk you will be well-placed to use the free spreadsheet tools shown or build your own tools to perform probabilistic forecasting. It really is easy, much easier and less stressful than detailed estimates and often more accurate because it avoids individual cognitive biases (wishful thinking).

You will discover:
  • The goals of forecasting and what makes a Good versus Poor forecast
  • How to answer the three biggest forecasting questions
    • How Big – determining how big is the feature or project with less effort
    • How Long – determining how long it would take to deliver a project or feature
    • How Much – OK, not everything will fit, what will and is that the right value
  • The three biggest reasons for forecasts to fail and what to do about them
  • How to use the data you have to forecast better and what to do if you have no historical data at all