This month we’re going to look at the principles, stepwise approach and top tips for creating a robust forecast for your brand for the year ahead.

So how good are you and your team (and even your organisation) at forecasting? It’s worth thinking about this: How accurate were your assumptions (quantified and documented of course) for last year’s forecast? How accurate and easy to use is your patient-based forecasting model?

How many times did you need to reforecast? What was the balance of top-down and bottom-up forecasting? (ie, were you empowered to create and recommend a forecast as a team or did you have a target handed down to you from ‘on high’?) How well did you create alternative forecast scenarios to manage risk and uncertainty?

Over the past few years, I have written and run a number of pharma forecasting training programmes and am continually surprised at the complexity, time and angst that is associated with forecasting. I trust I can show you that forecasting can be fun! (Or at least a manageable exercise...)

Let’s start with a few principles for creating a good forecast:

• Master your inputs: quantified assumptions, sources of information on patients, the competitors, the healthcare environment, the internal product profile and approximate resources available
• Master your forecasting environment: know who can help you with the forecast, the model, the inputs and who you need to engage and present the forecast to and in what organisational context (high growth, profit focus, portfolio considerations)
• Keep it simple, logical and patient-based: apparently 80% of complex Excel business models have been shown to contain logic or numerical errors that would fundamentally affect senior management decision-making. So, I’d recommend creating a straightforward, patient-based forecasting model without complex macros and hundreds of input boxes that can confuse the user and approver
• Use different techniques: ideally use different sources and methods for creating your forecast and see the different results – this can help you identify any areas of bias in your input research or in your models
• Keep it flexible: document your base case assumptions and then create at least two other scenarios with assumptions and with the resulting impact on your sales forecast so that you can understand and communicate the risk profile of your brand or portfolio in your market
• Check your figures: don’t forget to check your forecasting models, graphs and presentations again and again – one little numerical error can completely derail a team or senior management presentation and undermine confidence in all your good work.