R&D outsourcing is becoming increasingly popular as companies learn to let go. Sarah Houlton reports
R&D outsourcing is becoming increasingly popular as companies learn to let go. Sarah Houlton reports
Research-led companies are breaking away from holding on to in-house R&D as their ’crown jewels’ and entering the world of outsourcing. In manufacturing, outsourcing has long been an established way of controlling costs in business; for commodities, labour costs are a substantial part of the overall expense, and it is frequently more cost-effective to manufacture in a cheap economy, despite the subsequent shipping and import costs. Now R&D is catching up.
As Jim Darwent, vice-president, corporate research at Unilever, explains, much of the chemistry-based industry has been moving closer to the concept of ’open innovation’ pioneered by US academic Henry Chesbrough. Chesbrough defines open innovation as any external collaboration whose purpose is to obtain or codevelop a technology, material, product, process or brand on either an exclusive or preferred basis.
’In a closed system, the smart people in our field work for us,’ Darwent says. ’To profit from R&D, we must discover something, develop it and make it ourselves. If we discover it first, we will get to market first, and we should control our intellectual property to ensure our competitors don’t profit from our ideas. But in an open system, we don’t have to originate the research to profit from it. Building a better business model is better than getting to the market first. We should use our IP and buy others’ IP whenever it helps our business.’
He points out that the mindset was very different when he started in the industry, and companies jealously guarded their IP. Sharing anything with competitors was anathema.
Open to ideas
Open innovation can take many forms, whether simple R&D, co-branding and marketing, or even joint distribution. It can also happen in all areas of the business, whether it’s discovery at the more academic end, through more established technology to the finished product itself.
But why do it? ’The world is much richer outside your organisation than within it,’ Darwent says. ’There are a massive number of potential collaborators out there. By far the highest critical success factor in innovation is understanding your market, but the most important lever is collaboration.’
He cites several examples of open innovation from Unilever’s business. With its Lynx (Axe in Europe) male toiletries, Darwent says that although they have the top three brands in the sector and a large number of in-house experts who understand malodour, they still work with about 200 people in 20 outside organisations to develop new Lynx products. Another example is the recently introduced Comfort water for steam irons. It was not compatible with Philips’ new irons, so the two companies worked together to develop a Comfort dosing system.
One company that has embraced the open innovation model enthusiastically is BP, which shut its in-house research centre in Sunbury in the mid-1990s. When Martin Atkins, now director of technology for its Chinese operations in Dalian, joined the company in 1980, he says the company’s technology development was very closed. ’Sunbury was a BP fortress!’ he says. ’The only external things we did were small university collaborations.’ He claims it has been quite a journey from closed to open, and some drawbacks are worth mentioning. ’We lost experience [with the closure of the Sunbury site], and we are still suffering from that today,’ he says. ’It has taken a decade to recover from the loss, not the two or three years we expected. And as the organisation evolved by acquisition, we found we had a shortage of specialists, especially chemical engineers. But in China, for example, we find we have a much larger pool to recruit from.’
The pharmaceutical sector in recent years has used significant amounts of outsourcing in manufacturing and, increasingly, the development stage, but hit generation and lead optimisation were considered a company’s core competence. ’In drug discovery, we made compounds in house, whether they were drug candidates or designed to test a biological hypothesis,’ explains Andy Merritt, director, outsourcing collection enhancement at GlaxoSmithKline. ’We hired chemists, and did it ourselves.’
He cites one project where computational chemists had used earlier drugs for a particular target to design a new pharmacophore. ’The diamine cores they found looked simple but were not commercially available, so we didn’t have ready access to them. We could have made them ourselves, but we chose to outsource them. We could also outsource for substituent groups to attach to the cores, and then do chemistry internally with these fragments to get our lead compounds.’
Increasing amounts of routine chemistry involved in the drug discovery process are now being outsourced. ’Standard peptide synthesis, for example, is considered a commodity,’ Merritt says. ’We used to have a large peptide group internally, but over time this has been shifted out.’ For target validation, moieties such as fluorous tags typically need to be added to the peptides. GSK scientists will now decide which tag is best, and then work with partners to teach them how to add it. ’It’s good for them because they learn a new skill, and good for us because it saves money. But it’s important that we still retain internal capability for "difficult" synthesis.’
’The real novelty comes from how we design what we want to make, and then we will commission how we want to get it made. Will we do it in-house, or will we outsource it? How much is commercially available? We will do it in-house if it’s too risky to go elsewhere, or if it’s something we are better at doing, such as solid phase or microwave chemistry - the things we know we can do. Building blocks are ideal for outsourcing, as they tend to be reusable in many different projects.’ This sort of thing really plays to the strengths of companies that offer bespoke services - the custom houses. ’There has been an exponential growth in these companies - there were fewer than 20 in 1994, and probably nearer 500 now around the world,’ says Merritt.
Merritt adds that, although there is some potential for simply going out and buying research chemicals as building blocks, this is limited and they now focus more on what’s not available off-the-shelf. This is largely in two areas: virtual libraries and arrays of structurally diverse compounds. Buying in virtual libraries allows them to explore someone else’s chemistry capability, and numerous companies have sprung up in this area, particularly in former Soviet states such as Russia, Ukraine and Georgia. ’They may, for example, have heterocyclic cores or scaffolds we’re interested in, and we could get them to add a couple of our building blocks, too,’ he says.
With so-called diversity arrays, it is more a case of buying in skilled labour to carry out chemistry that GSK has already worked out. ’It gives us access to more people with technical skills,’ he says. ’This is a less established market for partners, especially overseas. They may need help with technical capabilities such as purification, but they can make arrays for us.’
There are a minimum set of requirements that are essential if a collaboration is to go ahead, he claims. ’If they don’t have good safety, facilities, staff, track record and quality of service, we won’t go there.’
And where next in pharma outsourcing? ’There will be an even greater appetite for building blocks, screening compounds and prescriptive arrays [outsourcing arrays that have been designed in-house],’ Merritt predicts. ’Low-cost countries are now the main providers, but there are niche suppliers in western Europe and the former Soviet states. Cost and service are the main drivers, and we may eventually have the complete supply chain from building block to validation to production in one country to cut shipping issues.’
In the chemicals sector, there is a much lower tendency to outsource R&D, with companies more likely to rely on their in-house capabilities. According to consultant Gunter Festel of Festel Capital, Germany, the big exceptions are in catalyst development and industrial enzymes, two fields where specialist experience is key.
’In catalytic services, companies are very open to working with specialised R&D service providers in areas such as high-throughput screening because you need very detailed knowledge and equipment. And in industrial biotechnology, there is good R&D cooperation between industry and service providers. But in the traditional chemical area, I think the industry is very conservative and they have enough in-house resources. For example, if you take a look at specialty chemicals companies like Degussa, Ciba and Clariant, they normally only work with external companies in these two fields.’
Festel claims the main driver here is cost. ’One strength of the pharma industry is that it is very open to external collaborations, and this has improved R&D efficiency. But the point is always money, and the chemical industry does not have as much money as pharma, so there are some hurdles in the way of working with external service providers. In chemicals, cost is one of the main drivers.’
One of the biggest risks in R&D outsourcing, regardless of sector, is the potential loss of control of IP, and BP’s Atkins claims one must be mindful of which controls are put in place. ’There is a great risk in outsourcing without control,’ he says. ’We once had an informal chat with some academics, and they had seven projects under way in the area within a month. It caused us to look very closely about how we do it.’ He adds that it’s essential to have a consortium agreement in place at the start. ’It must cover IP, the exploitation plan, options to finish early, termination, and rights for the partners after completion. It may all be hunky-dory to start with, but you never know what will go wrong in the future.
’In China, relationships are probably more important than technology - get the relationship right and you can do business, get it wrong and it’s very different. It took three or four years to sort IP out with one big institute we know very well, and in the end it was a commercial problem. We learnt that you have to recruit locals into your organisation. IP was a big challenge, and we got it wrong at the beginning.’
It’s also important to be clear about your strategy, Unilever’s Darwent says. ’There are many examples where people have tried to do innovation in-house that overlaps with collaborators. Don’t duplicate your efforts - it’s very bad news!’ He cites the example of an enzyme company Unilever worked with for a while. ’They felt as if they were in competition with our scientists in-house,’ he says. ’Don’t try and supersede the collaboration with your own business - be clear about the skills you have inside, and what you don’t.’
GSK’s Merritt is clear that collaboration has to be a two-way process, and has to work for both parties. ’Full-time equivalent (FTE) rates tend to work best for both if there is to be some continuation, but if it will be, say, 90 per cent of their work that’s an issue,’ he says. ’If we pulled out, that company would be gone. We’ll try never to take more than about a quarter of their FTE base so they don’t rely on us too much.’
Contractors are clear that good communication is essential if a collaboration is to succeed. ’It’s key to get scientists to talk to scientists, and not involve people like me,’ says Ed Dutton, business development executive at Peakdale Molecular. ’That way there are no Chinese whispers.’
Mark Bell, manager at pesticide formulation specialist Battelle, agrees. ’I appoint a project leader and ask a client to appoint a technical guy as one too,’ he says. ’That way, the technical people talk to each other and cut out the managers. Tech-to-tech conversations help to build trust - and to get repeat business.’
As commercial pressures to save time and money and get new products to the market increase ever more quickly, open innovation is sure to grow in importance. The closed model, BP’s Atkins concludes, may have advantages in terms of confidentiality, IP and the structured organisation that comes from having one team. But, he says the disadvantages are increasingly coming to the fore, and in a highly competitive worldwide market, it is becoming difficult to sustain organic growth.
Atkins claims that the open model has been very successful at BP, while among its competitors, Exxon-Mobil remains very closed and Shell is at an intermediate stage. ’It will be very interesting to look at it in five years to see how many technologies have been commercialised in the different models over a realistic timescale.’
Much of this article is based on the RSC conference: Creating business through outsourcing R&D, held on 12 March 2007 at the Patent Office in London, UK.
Sarah Houlton is a freelance science writer based in London, UK.