Four Core Ingredients Of Successful Forecast Model Design

A forecast model should make your life easier, not harder. It should allow people with varying capabilities, responsibilities and experiences to input their ideas and rationale, and understand what will happen to a product if the market changes.

Therefore, a successful forecast model gives structure to thinking, facilitates discussion and decision-making, saves time within the organisation while reducing questions and increasing confidence in the outputs. But what does success look like in forecast model design? What are the key elements of a robust, accurate and easy-to-use model?

We’ve broken it down into four core ingredients, to shape your thinking when designing and refining your forecast model:




When developing forecasts, it can be difficult to achieve transparency due to the number of stakeholders involved. What’s more, while certain functions will strive for visibility, this might face resistance from other stakeholders. Poor forecast models can lead to misleading outputs; for example, low targets set by inaccurate assumptions.

So how do you overcome this problem? One of the key approaches is to break down your forecast model in more detail and provide space for rationale, giving an insight into the logic driving the numbers. For instance, if your product is going to be prescribed for a specific group of patients, say severe asthma patients with a specific biomarker, it would be important to determine the size of those specific patients with the biomarker as opposed to defining the patient population at a broader level.

Successful forecast model design increases accountability, by making it easier to assess and challenge assumptions. This transparency in turn eliminates guesswork and the risk of making poor decisions based on poor data.



   In an effort to reduce costs, pharmaceutical companies may be drawn to off-the-shelf forecasting solutions. While this can be appropriate for certain situations it is important to ensure any forecasting tool can be tailored to meet your needs. It might be cheaper in the short run, but it’s dangerous to cut corners when embarking on such a business-critical process.

Effective models are almost always tailored – to the brand, to the product, to the company. Each product will be targeting a particular group of patients, and the model should reflect this. Forecasting also varies significantly from one therapy area to the next, with different market forces affecting the brands in those spaces.

Meanwhile, a forecast should be created with other processes in mind. How, for instance, will it feed into profit and loss calculations? You need to understand what inputs go into your P+L, and tailor the model accordingly. This ensures the outputs are generated in the right format for seamless integration into other systems / processes.



Tailored models also give you flexibility. Updating and replacing forecast models is a costly and time-consuming process, so when investing in design you should focus on longevity. However, as markets evolve and your product lifecycle matures, it’s inevitable that the needs and purpose of your forecast model will change.

This is where the benefits of working with a pharma-specific forecasting company really come into play. Industry experts can support you in deciding which functions your model needs, which ones aren’t necessary, and, crucially, which will become essential as you move forward.

It is important to understand the market in question and look ahead in the product lifecycle, ensuring you are prepared for what’s coming. Functions that fall into the last category – essential in the future, but not now – can be built into the model with an on/off feature, meaning your model isn’t cluttered with functionality you don’t yet need.



The benefits of having a transparent, tailored and flexible forecast model are extensive, but the challenge lies in the design and implementation. While some stakeholders might strive for this level of detail, they need to be sensitive to the needs and desires of all the stakeholders who will be interacting with the model.

A model that is too complex and demanding is likely to face resistance – and even resentment – from those responsible for inputting the data. Simplicity is important if you want to increase buy-in to the forecasting process.

This, of course, goes against the first three ingredients of a successful model, you should bear this balancing act in mind when pitching our fourth ingredient against one, two and three… We’ll be exploring complexity versus simplicity further in the near future.


Summing up

Designing a forecast model takes time and thought, but get it right and you stand to reduce costs, increase accuracy, improve buy-in and enhance decision-making. The resulting forecasts will make discussions easier and more comprehensive, with confidence in the logic behind the data and assumptions.

By partnering with a pharma-focused forecasting consultancy, like J + D, you can design a robust, long-lasting forecast model that will deliver successful outcomes for your company and its products. 


David James, Founder and CEO

0161 486 5005

Contact details