What is the difference between strategic thinking in quantifying the future and Excel gymnastics? This was a distinction I needed to learn in my early years of marketing and, more specifically, market research.

Often, I possessed a rich bevy of insights from primary market research (think green and white-striped cross-tabulations on pin-feed computer paper), together with volumes of secondary data (and this was before online-accessible tools). But I was still looking at an empty spreadsheet.

How could I possibly answer the leadership team’s questions? Did they even have the right questions? Which analytical path should I choose to synthesize all of this learning? How could I tell a story that would enable them to make essential decisions for the company and, ultimately, for their novel discoveries that would treat the patients that needed them most? I felt a responsibility to those patients because I had “seen” them in the epidemiology numbers and heard about them from the HCPs I had spoken to.

Choosing the Best Path

It’s been said that “every forecast is wrong.” I confess that this old adage had a tendency to amp up Young Jim’s professional anxiety.

Wiser Jim of 2022, however, agrees — uncertainty is baked in from the start, and predictions of future events invariably differ from the real thing when they arrive. And yet, in the world of drug development, where investors are your lifeline and patients are your passion, you still need a forecast — every time.

A forecast model is an estimate of the future value of something. In biopharma, specifically, it’s a way of communicating that value — to investors, colleagues, regulators, and others who may (or may not) have an interest in supporting the work.

A forecast is not necessarily a full-blown model. As discussed below, quite often it is intentionally not a full-blown model. But whatever form it takes, it is a set of assumptions whose dual purpose is first, to tell a story that brings others along with you, and second, to establish priorities and guide decisions throughout a drug’s development.

Three Questions I Like to Ask

When developing a forecast, the question is not, “Do we need one?” Rather, “What’s the most appropriate model or level of precision needed given our current situation?” For that, we must start with the most basic of questions:

  1. In what stage of development is the drug?
  2. Is this an existing market or a new one that must be created?
  3. What decision(s) will this model inform, and how critical is that decision today?

These questions guide us in understanding the level of effort and complexity for the work that lies ahead:

  • A forecast for an early-stage drug in a well-understood market that is providing guidance for a multi-asset portfolio strategy can be largely directional.
  • A forecast for a drug in Phase 2, or mid-stage, that may be partnered or upon which the company needs to conduct public financing, requires a deeper analytical process.
  • The effort to understand the value of an early-stage drug with no precedents in a market that will need to be created will be more complex, more expensive, less precise, and will take longer.
  • With a high-stakes decision (e.g., the company is choosing between two assets to acquire), the confidence in the forecast needs to be high to justify any investment recommendation made.

Let’s consider an example. Imagine that you are working on a new treatment for diabetes, a disease for which there have been treatments available for decades. We know the number of cases, how long people live, and the types of therapies already available. We have historical evidence of physician practice, patient preferences, payer response, how value is defined, and what the market is willing to bear.

In cases like this, where there is already a lot of existing data available, the focus of our forecast will be less about validating the indication opportunity itself and more on differentiation and the share of the existing market that we believe we can capture.

On the other hand, if we are working towards a novel treatment for a disease that has never had an efficacious treatment beyond the provision of supportive care, the forecast will necessarily have a different focus. In this situation, you need to think about who will treat it (which may be diffuse), how and if it can be diagnosed, how awareness of the patient journey or treatment options may change over time, and the receptivity for treatment, among other questions. And, of course, what is the treatment’s value — to patients, physicians, health systems, and society at large. Overall, the newer the market, the harder it is to predict.

Both scenarios (and every variation in between) represent “forecasts.” But their focus and emphasis might be entirely different based on the information needed and the stakeholders that need persuading.

Additional, More Detailed Questions Are Needed

This foundational insight-gathering leads to a series of more granular questions that will guide the determination of which type of modeling approach the team will take. The answers to these questions are longer than we have space for in this e-conversation but should absolutely be part of your internal discussion before investing in a robust forecasting project.

  • Is the condition chronic?
  • If acute, is this an oncologic disease?
  • If not oncologic, can a patient have more than one acute episode?
  • How many patients are there with this condition?
  • Can the condition be diagnosed, at all or easily?
  • Are all/most patients with the condition diagnosed?
  • Does patient segmentation determine treatment candidacy?
  • How do patients flow through the diagnosis and treatment path, and where are the treatment intervention points?
  • Can patients be treated with NewDrug only once?
  • Is therapy self-administered by the patient or caregiver?
  • Does NewDrug patent expire during the forecast window?
  • What competitive events are expected to occur during the forecast window, and how might these influence the opportunity for NewDrug?
  • What are the pricing benchmarks in the space?

Conclusion

Every company needs to take the path with which it is most comfortable. That said, be careful not to burden the asset’s story with false precision for the sake of projecting a sense of “completeness.” The “best” forecast is not necessarily the most detailed. Rather, it is the one most appropriate for the specifics of the situation at hand.

Canadian investor and philanthropist Peter Cundill may have summed it up best when he said:

“I think that intelligent forecasting (company revenues, earnings, etc.) should not seek to predict what will in fact happen in the future. Its purpose ought to be to illuminate the road, to point out obstacles and potential pitfalls and so assist management to tailor events and to bend them in a desired direction. Forecasting should be used as a device to put both problems and opportunities into perspective. It is a management tool, but it can never be a substitute for strategy, nor should it ever be used as the primary basis for portfolio investment decisions.”