Should you answer strategic questions directly or decompose them? | Milner Strategic Marketing

Should you answer strategic questions directly or decompose them?

Building a wall from blocks

Many strategic marketing decisions require answers to challenging questions which often relate to the fundamental growth of our businesses – new market entry, product launches, pricing decisions etc.  In this blog post I want to explain why decomposing these questions to understand the individual drivers within the market will lead to better decision making as well as transparency and trust in the underlying intelligence.

Challenging questions
I started thinking about the tough strategic marketing decisions that companies have to make as I watched Google announce its entry into the nascent home assistant market last month.  The announcement was a competitive response to Amazon’s Echo which was launched last year and in just 4 months achieved over 3m sales.  The home assistant market appears to be opening up a new high-growth category, so for Google, the questions it must have been considering were: should we respond, how big an opportunity might this be, what sales volumes and market share could we garner and what would be the main barriers to adoption.

I had similar conversations with colleagues in London earlier this year at a Forecasting Practitioners’ Workshop put on by Lancaster Business School, where we were looking at new product forecasting.  One of the things that I took away from the event was the power of decomposition.

How does decomposition work?
Decomposition is a judgemental forecasting technique which involves breaking down a seemingly simple question such as ‘how many units will we sell?’ into smaller questions about underlying drivers.  Each driver is identified and forecast, before being aggregated together to provide a more sophisticated and accurate forecast.  The logic behind decomposition is that it is more manageable to forecast the individual components of an opportunity than the whole thing and this results in more accurate findings. For example, the whole market can be decomposed into a set of geographies/sectors/segments/SKUs and then aggregated up to provide information about the total market opportunity.

Benefits of decomposition forecasting
Companies that adopt this forecasting approach have access to a more granular understanding of the market opportunity, with its underlying drivers.  This provides three key benefits:

  • Greater accuracy – it’s easier to be specific about smaller areas than about the total. This prevents trends being over extrapolated by understanding the fundamental drivers
  • Greater transparency – explicit assumptions and drivers behind each component forecast can be revealed, documented and understood.  This provides an audit trail and a defensible rationale for the forecast
  • Greater flexibility – the underlying drivers can be manipulated to understand the impact these have on the overall forecast.  This sensitivity analysis is particularly useful when exploring different market scenarios

Access to this sort of rigorous market intelligence should result in better decisions being made.  However, forecasting at this granular level is challenging and time-consuming. Do you think it’s better to forecast at a whole-market level or decompose the forecast into its separate components?

About the author:
Jonathan Davenport is Head of Market Analysis and has particular experience of using market and competitor analysis to support strategy development. He works for Milner Strategic Marketing, a business consultancy that helps companies grow their value by delivering a range of services from strategic management consultancy through to specific marketing programmes.

< Back