John Whatmore's blog

Can accelerators work for big companies where innovation times are inherently long?

John Whatmore - 26.03.2012

Where accelerators (intensive hothouses designed to generate rapid development) focus on innovation-to-market, do they have anything to offer in those sectors of the economy whose innovation times are inherently longer?

Pharma has its own special reasons for wanting to accelerate the generation and marketing of new drugs, among them expiring patents, declining pipelines, increasing costs of development and harsher regulatory regimes; and its development times for new drugs are enormously long.

GSK introduced its new approach to drug development three years ago - for that is the length of its new stage-gate cycles, with smaller multi-disciplinary teams, each focused more clearly on 'discovery performance', 'with scientists back at the centre of the process'; and these new drug development teams have just faced their first confrontation with a senior management Dragons Dens-like meeting to assess progress and continuation of funding.

Clearly this new approach requires managers who are less operators of big company systems and more focused on specific outcomes; its multi-disciplinary teams will make for more effective team working; and it exerts strong pressure on the teams and the development process. But, there is the danger that within the company, competition will erode the potential value of collaboration, and the single three-year stage-gate cycle is probably not appropriate for all such teams. And while this draws on recognisable models of development, has GSK recognised that good research can probably be found in quite different kinds of places? Time will tell! 

GSK's new structure and stage-gating focuses more clearly on leadership, involvement and output, but raises questions about scientific collaboration and project evaluation; and it both helps and hinders morale. The company is changing the way it looks for new medicines to give it 'the focus of a biotech within the resources of a large company', says the head of Medicines Discovery and Development. It has broken up research into competitive teams and 'put scientists back at the centre of the process'. But researchers who do not adapt must go, says an article in Bloomberg, (though admittedly some have already been poached by rival drug makers).

GSK's CEO who took over in 2008 divided the six disease-focused research centres into 38 smaller units, to be known as Discovery Performance Units, (and an invitation to make suggestions for new DPUs received more than 50 responses good enough to be considered); and these DPUs were to compete for funding every three years. The Head of Research and Development says that under the old system, talent was buried in the ocean. Some scientists have gone from working on ten projects to just managing one or two (one scientist went from a position of overseeing 120 employees to just 35). Most 'grew up as leaders in a remarkable way, while some struggled', and scientists now live or die with their project.

One challenge of the new regime is to get the right balance between competition and collaboration. 'We need cross-pollination', said one team leader; '38 silos would be a disaster' commented the Head of Medicines Discovery and Development.

Insiders say that some things have clearly changed: with less time and fewer resources at hand, research teams have cut the time between when a drug candidate is picked and the first clinical tests. They also rely less on animal studies, opting to use human tissue wherever possible to avoid the surprises that sometimes come with differing species. 

Two other indicators are mentioned: people seem to be making tough decisions - about terminating projects - earlier in the process, and an internal survey shows a drop in the number of researchers seeking jobs elsewhere.

In November 2011 GSK completed the first appraisal of its new model: DPU team leaders faced a panel to which they had to defend their achievements, and then at a second round, set targets for which they sought future funding. And the company's senior research managers are now deciding which teams deserve more funding and which ones do not; and its conclusions are just due. 'But,' said one commentator, 'you are worried; you own it more, so you are worried about the future of your science.'

Opponents say, given that new drugs take about ten years to develop, that researchers deserve to be evaluated when their product reaches a strategic milestone, rather than at arbitrary times. The CSO of a big competitor says that R&D doesn't work this way; you have to make long-haul choices on products and strategy, not abrupt decisions on how to allocate funds.

'Individual intellects can really make a difference, instead of just the big engine cranking out stuff,' said another commentator.