Forecasting for products in specialty market environments
Michael Turner, vice president, brand management practice, Campbell Alliance, on creating structured forecasts that account for unpredictability
By on Jul 26, 2011
Does the fall of the Roman Empire presage the end of America’s reign as a superpower?
Does the fact that Pol Pot was overthrown, arrested, and tried for his atrocities in Cambodia hint at what Hosni Mubarak has coming in Egypt?
Perhaps and perhaps not.
History certainly is cyclical, but when those cycles fall and to what degree is anyone’s guess.
“History repeats itself, but we can’t predict it,” Michael Turner, vice president, brand management practice, Campbell Alliance, said at eyeforpharma’s Forecasting USA summit in Boston.
“So what do we do to plan for it? When building a forecast, how do we predict for unpredictability but also understand what we can learn from things that did happen in the past?”
The best bet is to develop a structured approach, according to Turner, that combines insights from the past with primary market research and flexibility to account for the unpredictability that’s bound to come.
This is especially true in specialty markets.
“If we follow a very structured approach, we can at least minimize the impact of uncertainty,” he said.
“We can never truly predict uncertainty but let’s make sure everything is in place so that we can minimize it.” (For more on specialty markets, see ‘Forecasting for innovative products’ and ‘Patient compliance and specialty pharmacy products’.)
Mapping the market landscape
The starting point for forecasting any specialty product should be the market landscape.
Market landscapes are inherently dynamic, with new drugs gaining certification, others losing it, generics stepping into the fray.
In oncology, you have new tumor types, new pathways, and new therapies to consider.
You thus need a thorough understanding of the disease state from all angles, Turner said: “We can’t build a forecast until we really know where we’re playing. We want to make sure that [everyone with] input into that forecast—from clinical, from marketing, from finance, from management—has a clear, baseline understanding of where we’re playing, who else is playing there, and where some of those potential opportunities are.”
To accomplish this, Turner recommended two courses of action.
First, compile a disease and competitive dossier (DCD) that uses a variety of secondary market research sources to assess the disease state and therapy from clinical, commercial, and competitive perspectives.
This will help you identify unanswered questions and decide which can be handled through secondary sources and which should be addressed with primary market research.
Second, build a treatment flow model that breaks down how patients move through therapy and where your product will fit.
Primary research is important to understand unmet needs, treatment dynamics, current therapies, new therapies, and how physicians feel about all of the above.
“Most companies will start with in-depth interviews with physicians; you can do telephone interviews as well,” Turner said.
The method of interviewing is “not really critical. The key is making sure we’re asking the right questions.” (For an overview of patient flow modeling, see ‘Forecasting: How to get patient flow analysis right’, ‘Forecasting: Patient flow modeling comes of age’ and ‘Patient flow analysis and forecasting’.)
Finding the product fit
Once you understand the nuances of the market landscape, the question becomes where your product can play in that market and how.
For this, Turner recommends segmentation and profiling, methods that allow you to pinpoint and prioritize discrete groups of customers and opportunities for the brand.
“We want to make sure that we’re not going to treat everybody the same way,” Turner said.
“We’re not going to build one forecast for this market when we know that different patients and different physicians will respond differently.”
As you separate segments, delve into the subtleties of what motivates them and differentiates them, insights that will help you discover where you should play from the get-go, where you can play if have capacity to reach later, and where you shouldn’t go at all.
“Within each segment, we want a more complex understanding,” Turner said.
“What are the drivers, barriers, hurdles, and roadblocks set up by this group? Is this a set of physicians that are really eager and looking for something new, or are we the fifth to market in this class [without] any market differentiation?”
Turner recommends creating target product profiles to measure physicians’ future prescribing.
High-level assessments of existing and pipeline products are also good for determining how the target product can be positioned and for assessing the influence of future market entrants.
“You can’t build a forecast unless the fundamentals are there,” Turner said.
“I think the key to any forecast [is that] it’s only as good as the data that goes in.”
Estimating value
Finally, you need to assess value and pricing.
Here, Turner recommends in-depth global payer and reimbursement situational analysis, including payer, physician, and patient research findings.
This analysis should address questions like: What is the payer environment? Who are the key players? What are the drivers and barriers to reimbursement? And just as important, what will your value proposition be?
What are your pricing options—premium pricing, parity pricing, discount pricing, or risk share—and how will that influence patients?
“This is a huge component nowadays, especially in specialty markets,” Turner said.
“In oncology, there are some very, very expensive drugs out there and patients are paying a higher and higher percentage of the cost of that drug. The question is, at what point does it become too much? At what price do we lose compliance? Are patients going to push back against physicians and request something else?”
After the global payer and reimbursement situational analysis, it’s time for the forecast estimate.
A variety of forecast estimating approaches are out there, of course—an Rx trend model, market share model, impact analysis model, account-based model, patient-based model.
Each may be more or less appropriate given your functional area, available data, and intended use.
In general, Turner suggests applying a proprietary and highly flexible forecast estimate model that allows you to adjust your uptake curve profile and input overall marketing cost or to scale a portion of marketing costs to field-force size.
“Find something that’s flexible, that’s transparent, that’s adaptable to your needs,” he said.
Most importantly, choose a model that stays true to the type of research you’ve done and knowledge you’ve gathered.
“The last thing you want is someone saying, ‘That’s a share-based model but we have no way to anticipate what that’s going to be,’” Turner said.
“Let’s make sure we build a model based on the data that we have.”
For more on forecasting, join the sector’s other key players atForecasting Excellence USA on October 4-6 in Boston.
For an overview of eyeforpharma’s forecasting coverage, see ‘Highlights from eyeforpharma’s Forecasting coverage’.
For all the latest business analysis and insight for the pharma industry, sign up to eyeforpharma’s newsletters.