How not to deliver a forecast: a worst practices guide:
Richard Murgatroyd, forecasting manager, Roche.
Richard talks about how to deliver a forecast, convincing people of its value. He sees a lot of forecasts- he’s sat 180 forecast presentations! And he’s going to discuss what he’s learnt from them. Giving a presentation should involve telling a story, rather than just going into boring analytics.
Richard emphasises: strategy comes first before the forecast… The forecast needs to be supported by strategy and value proposition.
He then discusses some pitfalls he’s observed in forecast presentation:
- Unclear or weak value proposition. He uses the analogy of blocks stacked on top of each other, making a tower. The first block represents a clear understanding of market needs; this has to come first. Then comes a development strategy; the Target Product Profile; the Clear and Differentiated Value Proposition; and the cash flow forecast
- Lack of evidence or lack of evidence of looking for evidence: If you make up a number, say that you’ve made it up!
- No, poor or unclear expressions of uncertainty: People sometimes confuse lack of certainty with lack of confidence- but don’t. There are inevitable things we can’t know. All forecasts need to have a ‘what if’ mentality acknowledging the uncertainties around key assumptions. Be clear and honest- and don’t pretend the number’s real! Knowing the downside is critical- it gives us an idea of whether the product will be commercially viable.
He then discusses things that Roche do well:
- Don’t misunderstand roles. Make sure someone owns assumptions and outputs. One named person should have responsibility for the forecast figure.
- Don’t talk about the science. It’s there to support the figures, but don’t go into detail during the presentation. Only mention it when consistent with the ‘story’.
- Don’t pass buck. Everyone should have clear responsibilities.
- Do your pre-meeting preparation.
- Don’t use templates as a substitute for thinking. They are a guideline, that’s all! Don’t slavishly fill them in!
Summary:
- Consider and explain clearly the purpose of every forecast/ valuation (ask, ‘why are we doing this?’)
- Look at the past: be informed by it, but don’t be constrained by it.
- Explain assumptions: but also be honest about caveats, uncertainties and alternatives. If there’s no clear evidence- say so (even to the point of saying you’ve made something up!)
- Distinguish most likely from average (important technical point)
