Forecasting: Hindsight is better than foresight
*Peter Mansell explains why, when it comes to pharma forecasting, looking back is the new looking forward*
By on May 3, 2011Peter Mansell explains why, when it comes to pharma forecasting, looking back is the new looking forward
If youre an artist, according to Bob Dylan, you dont look back. If youre a forecaster, though, looking back is in the nature of looking forward.
Forecasts are a shot in the dark at best, while the conditions and assumptions that underline them are anything but static.
One survival strategy Dylan understood very well was the importance of maintaining a protean identity against a shifting historical, political, and social backdrop.
Forecasters need to be equally concerned with keeping their insights fresh and fluid.
The pace of change in pharmaceuticals has accelerated under influences such as new technology, scientific advances, pipeline exhaustion, patent losses, globalization, and economic contraction.
Product forecasts must constantly factor in these shifts both at the micro (e.g., phase of product development) and macro (e.g., price erosion, emerging markets) level if they are to stay relevant. (For more on price erosion, see Forecasting for generic erosion rates.)
Just recently its become a very big issue, says Nich Guthrie, head of forecasting at AstraZeneca in the UK.
He attributes this trend particularly to the shift in innovation in the pharmaceutical sector, with more products now licensed in rather than generated in-house.
We pay upfront, royalties, etc, based on a perception of value, he says.
So that perception of value is driven obviously by the forecast. Where we pay for our own innovation, its a hidden cost, if you like.
However, Guthrie adds, now that were bringing them in and you can see how much money youve paid out, there seems a greater desire to track the value over time.
More than just numbers
This calls for deeper insights than just a set of numbers.
I can tell you what we forecast for [AstraZenecas] Crestor in 2009 or in 1999, Guthrie comments.
What I cant tell you is why. I cant tell you what the assumptions were behind that.
Similar problems have arisen in some licensing deals.
We as an organization develop a forecast for the product, and maybe 12 or 18 months later we reforecast for the product and, of course, things have changed, Guthrie says.
That may be down to shifts in personnel or a different evaluation team.
Whatever the reasons, though, the organization says, Why is it worth more or less now than it was before? And you cant say why if you havent got the assumptions behind it.
AstraZeneca is trying to plug the gap by developing a forecasting database to capture the assumptions needed in these circumstances.
Its got to be flexible enough, Guthrie explains.
We have a core set of information that needs to go in there and additional stuff that allows you to adapt it to each situation.
This is something the forecasting team has wanted to do for a while, he notes, but now it has the organizational will to back the initiative up.
It should allow us to track how things change but also to begin to see whether were consistently wrong in certain areas, Guthrie explains, such as estimating the number of competitors, the product price, or the rate of uptake.
Once weve got a database, we should be able to begin to interrogate our track record and become more visible.
Decision points
How often AstraZeneca revisits its forecasting assumptions will depend on where a particular drug sits in the product lifecycle.
If the medicine is pre-launch, then the forecast will be updated once a year as part of the portfolio review.
Once the product is on the market, it will be looked at once a year as part of the business planning process.
But what were trying to do actually is less not more, Guthrie notes.
For example, the intention with pre-launch brands is to reforecast them now at major decision points but not in between.
This is partly economics, Guthrie says: No one is blessed with as many people as theyd like to do the necessary work.
But it is more about efficiency, he adds: trying to focus the activity where the activity is useful and to do a better job around big decisions.
It also reflects the stronger emphasis on value.
So if the team learns something new that is significant, it can look at what that means for valuation.
How closely an existing forecast is scrutinized will vary again according to the product lifecycle.
If its around a major decision point, then we tend to do a pretty rigorous and thorough rework, often because [the product] has moved from one stage of development to another and the degree of knowledge and precision has increased, according to Guthrie.
You may be just re-assuring yourself of the size of the opportunity.
The shift toward in-licensing
In the pre-launch phase, there is likely to be a lot more forecasting work around issues such as claims, market positioning, and their impact on potential usage.
The closer you get to launch, the more you know about the market youre going to be in, so the more precise you should be able to get about how the product will behave, Guthrie says.
The shift toward in-licensing has not so much increased the volume of forecasts as highlighted the need to be able to look back at past forecasts.
Weve always done it, but the desire to reconcile current view versus last view was not so great, Guthrie says.
In essence, Guthrie adds, the hurdles a bit higher to justify estimates both internally and externally. Were being asked to be more rigorous.
As industry attempts to replenish its R&D pipelines, everyone is more conscious that R&D investments must be justified.
But this more value-conscious environment has yet to have much impact on forecasting techniques.
We still tend to look for a trend, project it into the future and then apply events, Guthrie says. That hasnt really changed.
For some AstraZeneca products, weve looked more at patient dynamics, the patient journey.
But what has really altered, he notes, is the scrutiny things are being put under and the desire to be able to explain why the forecast is the way it is.
For more on forecasting, join the sectors other key players at Forecasting Europe on June 14-15 in Berlin.
For all the latest business analysis and insight for the pharma industry, sign up to eyeforpharmas newsletters.