Research Spotlight: Fearless forecasting30.07.2015
For many, the term ‘Forecasting’ can sound like an enormous quantitative and analytical undertaking. The type of project for which a joint mathematics and statistical PhD is required. Perhaps even a theoretical-physics degree, àla Sheldon from the Big Bang Theory. Basically, extreme statistical expertise that not all of us are fortunate enough to possess. However, after recently conducting a Forecasting project which we at first regarded as particularly challenging, I can reassure you that you no longer need to fear having to interpret reams of data, or coming face to face with a debrief containing a wall of numbers. You can have fun with forecasting.
To clarify, by Forecasting we mean capturing future use of a product in the context of market changes. This recent project was for a product soon to be joining acompetitive therapy area, for which research was required to help feed into the 5 year forecast. This research was further complicated by the fact there were new competitor products joining the area within a similar timeframe for which we also needed to assess uptake. The challenge here then was making the findings as accurate as possible, to closely reflect future reality; because I don’t know about you, but I barely know what I’m having for lunch, let alone what products I’ll be using in 5 years!
To ensure we were able capture future uptake of a product which had multiple products in the competitive set, we decided to keep things as straightforward as possible for the respondents by starting with a ‘back to basics’ approach, i.e. gathering background data for current context, before considering the future setting. We then assessed how the market might look once new competitor entrants have had time to bed in. Once we had an understanding of the impact of these products on the market, we introduced the product of interest to determine what impact it will have once launched.
Using one of HRW’s statistical proprietary techniques we were able to calculate potential share of the product of interest, whilst also assessing which products it will steal share from. At HRW we feel strongly that it’s important to stress test the data we collect, and so we used a combination of approaches to recalibrate the data collected, to give a feel for the upper and lower limits of uptake. Ultimately, we were able to measure uptake, cannibalisation and steal in a realistic setting.
Another key challenge for us in survey design for Forecasting research is that HCPs don’t always find prescribing-type questions the most engaging. How to challenge this, I hear you cry? Why, by using gamification techniques of course! We incorporated a number of visual techniques which resulted in extremely positive respondent feedback, such as “A great improvement on most of these surveys – actually enjoyed it and felt engaged” and “a nice and stimulating type of survey- with little need to type”. This type of increased engagement is extremely useful in generating higher-quality and more accurate responses- which is exactly what we are after in this type of research.
Being acutely aware that debrief audiences don’t always appreciate being presented with a deck filled with a forest of numbers, our final challenge was to make the debrief as interesting as possible. For this, we used data visualization techniques, keeping the slides as clean as possible – yet insightful, by focusing on providing commercially meaningful direction and relevant recommendations. Finally, to make the feedback actionable and clear, we incorporated a SWOT analysis (not as exciting as ‘SWAT’ teams, sadly) for the brand.
So, I have to admit that I actually had FUN with Forecasting. I’m happy to say I am now a forecasting convert, and you can be too – we don’t need to get Sheldon involved just yet.
By Lucy Wates