Reshaping Basic Research at Merck

Late in October, Merck topped off its announcement of third-quarter financial results with a long list of organizational changes. The ones that made headlines had to do with cost-cutting-elimination of 7,200 positions, including one-quarter of all executives, plus the shuttering of research facilities in Tsukuba, Japan; Pomezia, Italy; and Seattle.

The stories had the facts straight, but from the perspective of Kathleen Metters, they got the emphasis wrong. Metters, senior vice-president of basic research for the company, says that the coverage tended to overlook important, far-reaching changes that lay behind the layoffs and cuts-including some profound changes to the way Merck handles the process of basic research.

We caught up with Metters at Merck’s Rahway research campus early in November. What follows is a transcript of the conversation, edited for length and clarity.

Perhaps you could start by giving me the general outlines of how Merck is reorganizing its R&D.
Basic research at Merck has a proud history. We just celebrated the seventy-fifth anniversary of the Rahway site, and, although not all of the sites have been in our global network for that long, every site has a proud tradition here. What was also true was that research facilities came into the Merck network through various mechanisms. Some grew de novo from scientific efforts, and Rahway is a good example of that. Others were acquired as parts of strategic plans for various regions and the facility-certainly Merck Frosst, would be a good example of that.

What happened from an organizational perspective, is that every site grew up to be very, very efficient in terms of its local organization, but what we had was best described as a sort of early-stage European Union. It was a sort of confederation of states, as opposed to anything else.

We had a very decentralized operating model based on local best practices. That model served the company for many, many years and capitalized on a tremendous amount of expertise across the company-expertise in chemistry, development, and pharmaceutical research, for example. But it didn’t profoundly capitalize on therapeutic area expertise. And we realized that we were not playing to the overall strengths of the lab by being in a structure that was skewed toward subject matter expertise, as opposed to the therapeutic direction. So we introduced what we call a balanced matrix structure. In layman’s parlance it’s therapeutic-area expertise on one side, which we call franchise, and domain-area expertise, which we call function; and the matrix encapsulates both in a very even-handed way. We now have a research and development organization where everybody’s subject-matter expertise, wherever it lies, is brought to bear.

And this had some profound consequences. The huge benefit was that what we were doing in the Merck research labs division, was mirrored in what was happening in the commercial end of the business. We ended up with an end-to-end process, where therapeutic-area expertise runs across the entire spectrum of activities that make up Merck, so we can take an idea from its earliest conception in the labs right through to launch and even post patent expiry and manage the lifecycle of those opportunities.

So, you can take into consideration things like reimbursement and the financial implications of a drug?
Everything, from the earliest point in time. And so, what becomes very important in this new model is that the therapeutic areas can think very proactively about the strategic implications of every decision, from the earliest to, as you point out, way beyond launch, to extract the best value for patients, and the business. But we truly believe that if you are producing novel therapies that add value to patients and are differentiated from other therapies that are already on the market, then the business support from a revenue perspective, will be there.

It’s important to understand that this was a fundamental shift for the whole corporation, and so, it has everybody lined up, facing the same direction, much more explicitly than ever before.

How how does it play out in practice, though? How do you implement a matrix approach?
Well, when you have a matrix that has an explicit leadership structure that reflects therapeutic-area and domain expertise, you cannot operationalize it in a world where you have an independent coalition of labs. So when the matrix was launched, two and a half years ago, we put in the top layer of leadership first, then went into this long, thoughtful process to develop an operating model.

So, we took that as a first step. We have this matrix structure. We understand what we want to do. We want to bring differentiated medicines to people. High value opportunities. We want to introduce new therapeutic modalities into the business, everything from biologics through vaccines. We want to be able to introduce biomarkers and experimental medicine paradigms, but have them integrated into the drug discovery process. We want mechanisms by which we can fail early. And, for all of these reasons for all of those goals, we needed a new way of operating across the labs.

Because of the historic interests of these independently-constructed laboratories, each therapeutic area was split over many labs. Rahway, for example, grew up thinking about diabetes projects, so there were projects here. Montreal Merck Frosst, grew up thinking about diabetes projects, so there were projects there. The labs in Japan also had an interest in diabetes, so there were projects there. Some therapeutic areas were split over multiple laboratories.

So, after about six months of living with the matrix structure in 2006, it became self-evident that the old way of doing business in basic research was not going to realize the benefits of the new model. The sites were being asked to work together in a cohesive unit. But the organization wasn’t structured to do that.

What did you do at that point?
We did an assessment of life in the labs, working over several continents, with sites that employ from 1,000 people to 60 people, from drug discovery sites to sites that are just prosecuting platform technologies. We talked to approximately 25 percent of the employees. And the optimistic way of describing the feedback was that there was a huge opportunity to improve.

The not-so-optimistic way to describe it is that the way we were organized was severely limiting the ability of the laboratories to innovate. What was particularly sobering was that we knew that the way we were organized was an impediment to accessing external science, but what came back loud and clear was that it was an impediment to how we organize internally. So, it was every flavor of what you don’t want. As a global network, it was dysfunctional, and what we want to do is prosecute science in a global network and not as a series of small and large sites.

The other thing that was important about this project: It was transparent to the employees and staff within basic research.

So, right up front, you told people which sites would close?
No. We did not go into this with any idea that we would close sites. From the beginning, I said to the employees, this is not about site closures, though as a result of the process we may close facilities. This is not about centralization though, we may centralize certain functions. But at the moment, this is a agnostic way to look at the labs.

As a next step is, we did a scenario planning exercise. We basically said, “Okay, what does the future look like in 10 years?” And we created a whole range of futures. The underlying premise was that the future is not under your control, and the present is. And that’s how you move forward. And so, the futures we looked at were things like the regulatory environment, access to external science, the global availability of talent and how that might change with the developing world- everything from very, very sobering, to slightly less sobering, through to optimistic. In that exercise, we came up with a whole series of core strategies for what the operating model should look like. They were all important, but a clear driver for what subsequently happened in terms of the decision made, was the decision that a franchise should be located as a single physical site.

What was the reasoning behind that?
It was around the complexity of operating a franchise across multiple sites. The fact that it fragments the franchise strategies, that it silos franchise resources, that it means that you limit your core location with the clinical groups that are prosecuting the same area. Those were really the main rationale.

So you can move toward face-to-face interactions instead of doing everything through meetings?
We’re not naïve enough to think that we’re going to walk away from the meeting culture entirely, but certainly, the idea is, yes, that there’s much more agile interfaces. Now, do you give up the theoretical advantage of having, and I’ll exaggerate for effect, all six franchises at all of the sites so you can capitalize on theoretical synergies across franchises? Sure, but frankly, experience tells you that the advantages of the core location outweigh the advantages of decentralization.

So end-to-end control is more important than the possibilities cross platform discoveries.
The possibilities of cross-platform are much more theoretical than actual, and you can actually mitigate that risk using a successful portfolio management approach.

So anyway, we ended up with a series of core principles for the operating strategy. One was around having the franchises consolidated. Another was be absolutely sure that we started to, much more proactively and explicitly, access external science through partnerships. The other strong principle was that we had to have a pipeline floor model for each area that we’re prosecuting. It’s very important that you think about having sufficient resources across all of the discovery phases, in order to actually make sure that there’s an equal sampling of opportunities across all areas of discovery for each disease area.

Another important concept was that if you really prioritized programs, you make active choices about what you move forward, and about what you stop. Not all things advance. It was very much about building a sustainable pipeline for the company, based on pro-active management of the pipeline for each disease area, understanding that not all disease areas have the same requirements. In some areas there’s more opportunity to differentiate early because the animal models are more relevant. In some areas for example, in oncology, our competitive advantage is a profound understanding of the disease pathways.

We also wanted to take down the barriers to working externally. Frankly, the complexity of the internal labs meant that a lot of the researchers were expending so much energy collaborating with themselves that the energy to collaborate with the external world was reduced.

How will you work with external science?
We have created a virtual research site to access that science and there’s a head of that site, there’s a head of chemistry for that site, there’s a head of therapeutic area for each of our franchises that live within that site. Their only job is to prosecute external science. They have no internal lab space. Their whole mandate is to do it through partnerships.

Before, the internal folks did their own projects internally and they would try and collaborate externally. Frankly, the pull is always to do the internal work, so external science became a very small component of everybody’s job instead of 100 percent component of a specific group of our researchers.

Will the external science groups be forging partnerships between internal groups and external groups or will they just be filtering results in?
A bit of both. There’s every sort of variation on the theme. It’s around thinking about how to work more effectively in early discovery with academic groups, how to access preclinical testing programs through biotech, how to work out shared risk deals. Again, the idea is that there’s a proactive strategy for working with every sector as opposed to, frankly at its worst case, just seeing what turns up. So when we bring in opportunities, it’s opportunities that we want-not that we’re not open to good ideas wherever they turn up, but it is done within the express strategic framework for each franchise.

I’m trying to imagine how it works.
In practice we have several functions working in the external world. One is a scouting function, and those folks spend time in the field visiting universities, visiting small companies, visiting consortia, talking, looking to see what’s out there. They understand the internal strategies, so they’re looking for a fit. They may come up with an interesting target -say in diabetes there’s an insulin secretagogue and they know that we’re interested in that kind of thing. They would come in and talk at the franchise level. The franchise is linked through an explicit structure with external and internal labs and they would sit down and say, “Does this fit with our strategy? Is this an interesting target in terms of the science behind it? Has it been done well? Is it rigorous? Is this somebody that we feel that we could work with?"

What used to happen is that the partnerships would come into Merck. We have 3,000 people in basic research. You could have 3,000 people operating to bring in external ideas. In the end it becomes nobody’s job. Calls aren’t returned; things are passed around. Nobody makes a decision. Nobody has the funding. That’s actually what we’re trying to avoid by having a specific portal.

The system has been going for about a year. I would say that- And it’s sort of still very new, but it has resulted both in a growth in our partners and those partnerships being much more strategically aligned with the overall franchise strategies.

That sounds like it would enable you to do things that on the sales side would be routine, using CRM software to track everyone so you don’t lose contact. The old way you did things that would’ve been impossible.
A lot of those mechanisms were in place through the licensing function, but yes, we steal shamelessly from wherever there are good ideas. And, yes, it’s much more proactive in terms of tracking and follow up. Yeah.

Ninety-nine percent of science sits outside Merck walls. We want to be actively involved with those folks as opposed to thinking that we can rely always on our own expertise.

Where are you at in terms of implementation?
We just kicked this off on the 22nd of October. At the moment we’re putting together the implementation teams. It’s very important that we have subject matter experts who are very involved in implementation; but the implementation really comes in three phases. One is the program transitions, because we will be moving programs around. That’s a great opportunity to stop things that are not high value and to make sure that things are high value obviously are moved efficiently. Another one is around the people transition. There are certain capabilities that are moving. Not all, but some. And finally in that first phase is site transitions. We’ll be closing out three of the research facilities as part of this.

So for me this is very much an organizational problem. It’s got a lot of moving pieces and it’s got a lot of logistics but it’s a planning problem so when you think about how you move programs, you do it on franchise priority, you do it on complexity, you do it on maturity and we have a very proactive view into what our portfolio looks like and the milestones for delivery, so using that information that we can plan out the transitions in a very transparent and pragmatic manner. So we’re in that kind of implementation, that first phase of planning. We think probably somewhere between three and five years for it to be at steady state.

And at what point do employees kind of know what their fate is within the plan?
That’s a difficult question to answer because all employees are impacted somewhat differently. But if I sort of took that very generally, the way we think about this in terms of a new organizational design is that we will fill out that design basically in a top-down way, confirming the layers of leadership down but also in a sort of a vertical way in thinking about our capabilities and what we need to do for each run. I’m struggling with that question because there’s going to be large parts of the lab that are simply unaffected by this so we have a high-throughput screening facility in North Wales. Not going to change. So those employees, they’re not anticipating an impact.

Right. But then you have the genomic unit in Seattle.
Yes. So we were very aggressive with them. Those employees know their fate now. Many of them have offers in hand. Many people are already moving so that’s well on its way. So it depends on where you sit with the organization. We’re thinking about it in terms of a capability shift and a leader shift. We would imagine that everybody will know by end of first quarter next year. It’s not going to take that long.

And are you finding that the, especially the scientists who you want to keep, the people who are repositories of memory and knowledge, are you finding that they’re willing to make these moves?
What I can tell you is that every single key leader that we’ve offered a job to, to date, has accepted it. Every single one, 100 percent.

That’s impressive.
Yeah.