The Forecaster recently had a chance to chat with Graham Clarke, CEO of ImmunoBiology, about the progression of his career, which has often centered on forecasting. Like many who spend their days trying to predict what’s to come, Clarke has followed an interesting path to where he is today.
The forecaster: Can you tell us a little about the steps you’ve taken that brought you to this point in your career?
Clarke: Forecasting is not the only job I do now, but in my history it has been a major component. I’m like a lot of people; I did a science degree to start with— a BSc and MSc in biophysics. I started life as a science guy, and then ended up doing something different. I started with a project management role in R&D, followed by a product management job which took me more toward the commercial side. Then I started asking myself if, with a science background, I was really well-equipped to understand the work I was doing, so I went to business school for an MBA. Then the question was, “what do I do?”
In my case, I became a management consultant in a pharma and biotech practice—due diligence, valuations, that sort of thing. Inevitably, when you stare at markets, write business plans for clients or read their business plans because you’re trying to help people form a view of what they should do, you have to think carefully about the future. You have to figure out what that future might be like, and do it in a robust way that maximises the value of the company. You’re looking at other people’s forecasts and asking, “Do I believe this?” or of course, coming up with forecasts yourself.
I worked with a variety of companies and investment banks, did a couple of secondments, and then I got approached by SmithKline Beecham. They happened to be looking at the time for that sort of consulting background. A lot of valuation/business-planning forecasting groups sit in the center, but you don’t have direct line management over what goes on. A lot of the job depends on influencing skills, teamwork, matrix management, trying to form an objective position with all the power plays you’ve got going on around you. Forecasts are very political—a lot of decisions flow from them. You want the project team to be very positive about problem solving and behind the forecast, but they can become unquestioning champions. You want them to be champions, but it becomes a question of how do I objectively look at whether what they’re advocating makes sense from an investment standpoint? So SmithKline was looking for relevant consulting skills to bring in as these are very applicable in a forecasting role. While most of my colleagues went off to biotechs, I went back to “big pharma”. I originally ran the disease area management program, evaluating the areas SB should invest in and then I got more involved in portfolio management. I ended up as the Vice President of the Strategic Product Management global operations that included all of the worldwide forecasts, market research, business intelligence etc. Essentially, forecasting was at the heart of it. We did a lot beyond product forecasts—we looked at inward licence opportunities, healthcare environment scenarios, the portfolio evaluation & prioritization system, assisted investor relations in presentations to analysts, all sorts of things.
We had always looked at competitors and whenever mergers loomed in the industry, we ran the numbers to look at concentration, combined market share, development portfolios and so on. As a result, I ran the commercial side of the antitrust clearance for the creation of GSK (GlaxoSmithKline). I then ended up as the head of the executive decision support group for GSK R&D for a period, before moving on.
Actually, it was major move, from a very large company to a very small company. I’m now the CEO of a biotech company—so the picture changes again. To us, understanding market needs becomes really personal. Our investors and stakeholders are hanging on it. And you can’t hide anywhere in a small company; if you make the wrong decisions, you don’t have a big portfolio of other stuff to fall back on. So making the right decisions becomes much more critical. As an equity holder of the company and its head, the buck definitely stops here!
The forecaster: What is a specific challenge you feel professional forecasters typically face today and how do you feel it can be dealt with?
Clarke: At the business level, I think it’s around what’s killing the industry: attrition rates from R&D project failures. Whether you believe the specific numbers thrown about or not, for sure the cost per NCE is going up dramatically because the success rate is going down. We’re tackling harder issues, chronic diseases, unclear/unaccepted end points, poor animal models and longer, more expensive clinical studies. Increased average prices, especially in the US, faster take-up curves and good volume demand have been good at holding up the sales line, but with relentless, growing pressures, including from generics, you need new products. Ultimately, investment in R&D has to pay off. Times are getting tougher; I know people always say that, but I think it’s true. Forecasts have to better assess likely success rates.
Simple extrapolation is pretty darn unreliable. You need navigation aids, so the forecasting role is important and getting more so. A lot of emphasis has been put on tools, but I think issues still remain around the people: the right competencies; knowing who’s judgement is best when making decisions and in formulating assumptions; establishing the right, productive relationships across the organisation and whether or not you’re listened to. No amount of process design can obviate the importance of people.
And you need the right experience. Forecasting groups are great places for new people coming into that part of the business. Classically, it’s an early stage career step. You see a lot new MBA types – not heavyweight in terms of experience – so they become technicians of assembling forecasts. They do it smartly—it takes research, synthesis and all that—but it’s the underpinning assumptions that are really important. How good an assumption have you put forward, and on what basis is that judgment formulated?
We try to be analytical about it, but at the end of the day, you have to make a decision about an uncertain future, so judgement is really critical. As I’ve just said, project success rate have not been brilliant—sometimes they’ve been OK, sometimes poor, and sometimes horribly poor. But what do you do, hold up your hands and say, I can’t do this, it’s too difficult? Or do you say, well, we’ve got to improve—what do we have to do?
Often at conferences people get very excited about techniques and tools and that sort of stuff, but they miss the elephant in the grass. There’s this giant thing we’ve got to get right, but we pretend that finessing option theory or something similar is going to give us the answer, and it doesn’t: you need the right people with good judgement, notably about technical do-ability of R&D projects.
I find that often at conferences, we’re all broadly like-minded people: analytical creatures who feel that logic should win the day and we’re mystified at why organizations make different decisions.
So what do you learn over time ? You get guys who’ve left various companies and become advisors from afar. You think they’re going to provide sage-like wisdom, but actually they’ve often become disconnected as things move on, so I’m always cautious of their views. But what you do see with hindsight comes back to the attrition issue: we (the industry) have been systematically over-optimistic of technical do-ability. You actually don’t need to spend a ton of money on early market research to know fundamentally whether the market opportunity is big, small or in the middle. Yet, we tend to put a lot of emphasis on that side of the equation.
What we often observe is that it’s events that change actual versus the forecast, and often these are technical. For example Avandia was in its very early stages when I joined SB. People looked at the diabetes market and built forecasts and we had epidemiology information, identified unmet needs, pencilled in prices etc. But basically we figured out that if we really had the profile of the product that we were aiming for, it would be a valuable asset. Of course, a more demanding profile leads a higher sales stream, but achieving it is much harder. Attrition rates are worse, so matching a target product profile against the realistic likelihood of achieving that is what counts. The higher you set your target, the more unlikely you are to achieve it.
For Avandia, this did indeed become a valuable asset, apparently meeting its target product profile, until reports started on its safety profile as usage increased. If you read the GSK press releases, they feel some of the negative media coverage has been disproportionate, but nevertheless an event occurred which changed its profile. It’s not the commercial piece that changed Avandia’s prospects.
One thing is very clear: big pharma still has to deliver against the premise of why company size is good. When we did the GSK merger, we had to think carefully about why size was good. Part of our thinking was around scale of the portfolio, taking the lumps and bumps out of the top line, despite individual product lifecycles of its constituents, together with a more efficient and effective R&D. We were expecting to industrialize the research end, so that it was more predictable and effective as life science technologies matured. That was the model.
However, if you look at the big pharmas now, you can’t honestly say they’ve outperformed in terms of sales, profitability or share price. They still have blockbusters that leave a giant hole when they go off-patent. There aren’t enough products going through development to have even out product lifecycles. And a despite a lot of the technologies people thought would improve attrition rates, they have actually gone the other way.
There are a whole bunch of challenges out there. Forecasting is an important part of the tools, but in context with all the rest of the stuff we have to do. At the end of the day, you either decide to invest in a product or you don’t. The question is, how well have you thought through the issues and made your assessments? You need superior decision-making; you need people with insights who understand the risks. Even then, you may be lucky or you may not.
The forecaster: Can you share an experience that helped you excel in your forecasting career?
Clarke: People often search for the holy grail of how to do things. Conferences and courses are great, there’s some good stuff out there but I go back to the softer aspects. Persuasion, communication, articulation, engagement, teamwork: those are the sorts of competencies you really need. So consultancy skills are good, but it the experience has drawbacks too. Usually you don’t see things through to market or live with the consequences of your recommendations. You move between clients, you see a lot; you have to learn a lot and pick up the skills pretty quickly. You have to be compelling and authoritative, you have to think about time management and prioritization—for me, consulting was a very valuable period.
Now I’m running a vaccine development company. We’ve got supportive investors, including Nestlé’s venture fund and the Bill & Melinda Gates Foundation. But we’re a modest-sized company, and anti-infectives are particularly hard to forecast. Epidemiology can change dramatically—the classic is pandemic flu—how do you decide if you want to invest in pandemic flu or not?
