Five Lessons Immunex Taught Me
What happens when a biotech company disappears? We think that the Web remembers everything. Do a search on Wikipedia for Immunex and you get Amgen, with little mention of Immunex at all. The results of over 500 publications and several thousand employees over 20 years are pretty much invisible on the largest collection of facts on the Internet.
One of the most successful stories in Seattle history and it is mostly gone from view. Ten years ago, Amgen announced it was buying Immunex. Xconomy will be commemorating this historic event in December.
Immunex may be gone from the Web but the lessons it taught still have impacts today. Before Immunex completely disappears from view, I wanted to discuss Five Lessons Immunex Taught Me.
• Get to decision points rapidly.
• Live in the gray.
• Size matters.
• Leverage insurrection.
• Vet in the open. And then vet again.
I was a researcher at Immunex for 16 years as it grew from a small community where everyone knew each other to a large, global undertaking with people spread out in multiple locations. While some of these lessons may be part of any good MBA program, there is nothing like dealing with the day-to-day reality to drive the lessons home.
I’ve used these lessons over the 10 years since leaving Immunex – as a Vice President in charge of research at a biotech startup, as a founder of my own organizations, as well as in collaborations with other organizations such as the Sustainable Path Foundation, the WBBA, Global Health Nexus and the Washington Global Health Alliance.
Caveat lector: These are my recollections about events – some of which are over 25 years old—so forgive me if I do not get the events exactly right. After all, these are the lessons Immunex taught me. Others may have learned something different.
1) Get to decision points rapidly. Immunex always worked towards a decision point, an action that would permit us to either kill the project, move forward or change direction. Simply delaying a decision was almost never a consideration.
Shortly after joining Immunex, I was a member of a presentation we gave to one of our collaborators – a company with lots of proprietary chemicals looking to find pharmacologically important ones. Steve Gillis was there along with several other Immunex scientists.
In contrast to the Immunex crew, the people we met seemed more interested in finding reasons to delay a decision for six months or so than in what we had to present. They kept nitpicking rather than moving forward. Any excuse to delay a decision was good enough.
For example, we produced a standard operating procedure for performing a biological assay we had developed to use with their chemicals. It described how to grow the particular cells at several specific temperatures such as 35°C, 37°C and 42°C.
The response from our collaborators? “What about 40°C or 39°C? Or 22°C? We want those in the SOP also. The entire thing needs to be redone to include these.”
Gillis exemplified the Immunex response to this – he simply told them not to run the procedure at those other temperatures. Decision made and we just moved on.
The large company seemed to be afraid to accept our protocol at all, perhaps because they might make a wrong choice. Being wrong was a greater transgression than delaying a decision.
Saying “yes” could be a mistake and so could saying “no.” Delay meant “maybe.” Who could make a mistake with “maybe?”
But delay itself is a decision and usually a poor one in a rapidly moving field, making it very hard to adapt as circumstances change.
That may be why our collaborator never became a big leader in the biotechnology arena. We created billion dollar therapies.
Get to those decision points as efficiently and as effectively as possible, then actually make a decision. This is critical for an innovative company.
Move forward, kill it or change directions. Delay actually can be the same as moving backwards.
2) Live in the gray. While seemingly contradicting the previous point, this actually follows from a key aspect of decision points – the need for as complete a set of data as possible.
Good data is critical for making a decision. Incomplete data can destroy any decision.
Like the Fog of War, the gray of the Fog of Science that often surrounds biomedical research can punish decisions made without proper supporting data. Do we have enough data to let us move forward, kill it or change directions? If not, can getting more data help us get to a decision point?
I learned the importance of the gray from our work on Interleukin-1. We now know there are two forms of IL-1, each with a different protein sequence and produced by different genes, although they have similar biological activities.
Of course, we did not know that when we were first trying to isolate the molecule. We thought there was only one.
In the primitive days of the early ’80s, we initially only had a biological assay to work with and thought that only one protein or gene was involved. To enhance our chances of isolating the correct protein and gene responsible for the activity, we tried two different approaches – one group attacked the problem from the protein side and one attacked it from the gene side.
Both approaches successfully identified molecules with the correct biological activity but it quickly became clear that the two approaches could not have isolated the same molecule – two different molecules were being examined. Since we expected only one, which one was the correct molecule?
There were a lot of conversations about the results. Perhaps one approach was wrong and had identified the wrong molecule? Perhaps there was a contaminating protein messing up the assays? Maybe the purification process was incomplete or the DNA sequencing data was wrong? Maybe someone made a mistake? Had the reagents been made properly? Should we decide that one of the possible molecules was wrong and only concentrate on the other? Could we continue doing two separate approaches?
We realized that we did not have enough data to produce a final decision point. It simply was not possible yet to remove the Fog of Science and get clarity on the real science of the system.
So we continued moving forward by taking both approaches to their end points – getting the genes cloned for both molecules and making sure the proteins produced by these gene sequences actually had the correct activities.
That would get us to a true decision point. We only had to live in a few months of uncertainty to get clarity. A very unforeseen clarity.
It turned out that each approach was actually successful. We unexpectedly isolated two different molecules, each having the activity we were looking for. We now call those proteins IL-1 alpha and IL-1 beta today.
Knowing when you do not know enough is an important part of any process. Just as there can be pressures to not make a decision, there can be strong pressures to decide now – one way or another. Every day costs more money and uses more resources. But making a decision based on incomplete data can bring a company to its knees.
Data is king in biotechnology.
By waiting until we had all the data we needed, we ended up knowing more than if we had decided too early. This information allowed us to have a strong lead over others who only isolated one molecule.
Living in the gray can be as important as making a decision. Too many companies think that delaying a decision – staying in one spot – is the same as living in the gray.
But standing still is not progress. Learning to be comfortable in the Fog of Science lets one move forward, even if you can not easily see your steps.
3) Size matters. Immunex was pretty small when I joined – about 50 people located on 2-3 floors of one building. When I left 16 years later, there were almost 2,000 people found at more locations than most of us knew about. We went through lots of changes as we grew but the largest one for me became clear when we grew through 150 to over 200.
Dunbar’s number has been much discussed. About 150 people are all that our brains can hold and still intuitively know how everyone fits. I learned this first hand.
When I joined Immunex, it was easy to know what everyone was working on, when they failed and when they succeeded. If we cloned something in the morning, Gillis would be breaking out the champagne in the afternoon. We’d all be drinking beers down the street at the Mark Tobey, the nearest pub – another important Seattle institution lost on the Web.
But, and this was long before I ever heard of Dunbar’s number, when we made a large expansion through 200 people or so, I noticed something happen – all of a sudden there were people walking around that I had never seen before. I had no idea who they were, what they were working on or what their favorite beer was.
There were all these strangers, looking busy but doing who knew what.
Communication became much harder because the default communication system no longer worked well for effectively moving information around. We no longer could fit everyone into the Mark Tobey or actually any one room in the company. In the hierarchical system that started developing, there were now too many steps between people for information to traverse as easily as before.
For example, one of Gillis’ incredible talents that I greatly appreciated was his ability to walk in cold to any meeting and ask the one question that cut to the heart of the matter under discussion, no matter the topic or how much everyone else was dancing around it.
But as his time became more constrained by other needs, he could not make as many research meetings. As we got bigger, it became harder for any of us to really know what everyone else was doing.
We started dancing around making decisions. We made unnerving moves towards the delaying approaches of other companies.
And it became harder to reach decision points because the right data were not getting to the right people. We had a harder time moving out of the gray because we were not getting the information we needed.
This is something many companies have to face. Just as the oxygen-carrying systems of organisms get more complex as they get larger, so too must the information-carrying systems in a company.
Plan for dealing with communication inefficiency as the organization grows. Simply increasing spots on a hierarchical chart does not work.
4) Leverage insurrection. Creative minds need obstacles to overcome if they are to produce innovation.
Immunex always lacked money so finding innovative ways to surmount research obstacles was a very important skill. Sometimes Immunex creatively found ways to use these abilities in other areas than research.
When Immunex was small, all scientists had offices with doors that closed. These were useful for holding impromptu meetings with collaborators or for finding a quiet place to read a journal article.
Around the same time we were beginning to realize that growth was hurting information flow, we ran out of office space. So upper management decided we would move into another building next door and provide cubicles for the scientific staff, instead of offices.
Unfortunately, they failed to effectively communicate this decision to the scientists before it was made. When the researchers heard that some scientists would still have offices while others would have small, almost public cubicles, there was a lot of emotion. ‘Others’ were making decisions that affected ‘our’ work without talking with us. It seemed as though the egalitarian atmosphere of the company was disappearing.
The researchers got together to decide what to do. Gillis was present and, demonstrating his ability to focus on the important things, stated that the decision had been made and Immunex was moving forward. Delay was never an option.
And kill it was not going to be an option here. But a change of direction was still possible – maybe we could find a creative way to produce usable cubicles instead of the generic ones planned.
If the scientists could come up with something that satisfied us and also satisfied the corporate needs, go ahead. He gave us about a week.
It actually became a great experience, with several of us working hard to find a workable compromise between space and privacy.
We added a few extra walls to provide some privacy. But they were still too low, allowing anyone standing up to see everyone else.
We brought in an acoustic specialist who told us how sound works and how to muffle it with foam. We came up with the novel idea of putting colored plexiglass extensions on top of the walls with a tube of foam on top. It not only blocked everyone’s view but also provided really good soundproofing.
We ended up creating just as many cubicles as the company needed but with much more privacy.
Instead of telling us that they had made a decision and that was that, live with it, the management of Immunex wanted us to adapt their decision so that it would work for all.
Management wanted our creativity to be used on this problem. And they supported that creativity.
It was one of the smartest management decisions I thought Immunex ever made. Like some form of judo, our emotional responses were harnessed to find innovative solutions to a facilities problem.
And, as I recall, the design of the cubicles won an award. Not too bad for an insurrection.
5) Vet everything in the open. Then vet it again. The increasing size of Immunex began to make information flow too sluggish to arrive at good decisions. So we developed novel ways to exchange the needed data in ways that enhanced our abilities to innovate and make decisions.
Unlike many companies, we openly examined every research project in depth every four months. We leveraged our management infrastructure to create a matrixed system, one where department heads responsible for budgets of multiple projects butted heads with project chairs whose projects worked across departments.
After going through all the projects, the management would prioritize each project based on resources, money, etc. Then four months later we would do it again.
This sort of competitive conflict helped focus our scarce resources.
But initial problem because of those limited resources was that too often the process became a zero sum game – for one project to get ahead, another one needed to be cut. As we got larger and information flow degraded, things like politicking and personality began to encroach on decision points rather than purely objective, empirical information.
The process increased the chances of making poor decisions.
So, in typical Immunex fashion, we applied our own creative approaches to an internal company process, enhancing the information gathering aspects of the process and permitting decision points to be more effectively reached.
We did this by opening up the process, making the vetting transparent and finding a way to prevent some the adversarial aspects of this approach from overpowering our decisions.
Before the meeting, every project was assigned a chaperone – a neutral department head who would sit down with the project chair before the review and go over the progress, ask questions about problems and get a good idea of what was going on. Then at the meeting, the chaperone would introduce the project and discuss the relevant information, not the project chair.
Attention was thus directed to the chaperone – a department head – rather than the project chair. In fact, the project chair was often there to answer questions, not to defend the project. We could focus on the science.
By now having both a department head and a project chair driving the discussion, the entire dynamic of the meetings changed from an adversarial form – department heads vs. project chairs – into a more collegial one – committed people all striving for the same goal.
As time went on – because of their excellent ability to transfer information – we made these meetings open to anyone. Many employees at all levels could sit in on them just to get quickly up to date on all the projects.
This added vetting provided all sorts of benefits. Entirely new information that had not been known to either the project chairs or department heads would come out. Some of the discussions sparked entirely new directions of research.
It became easier for us to all move in the same direction, together.
More than anything else we did at Immunex, I believe these meetings overcame many of the growth problems we had experienced.
But they did so much more, because they leveraged the large increase in size Immunex had undergone to allow a much larger and more diverse group of creative minds to examine major research decision points.
In some ways, it allowed us to attack a much wider and more complex set of problems than when we were only 50 people.
Open vetting did more than bring the best of the insurrectional minds at Immunex to bear on a problem. It also put all the information out in the open, permitting all of us to understand why a decision was made.
Now when the decisions – move forward, kill it, change direction – were made for each project, people were pretty much in agreement because they could see the rationale for the decision. They had seen the information presented; their concerns had been heard by all. They understood.
It was not one group deciding what another group would do. The vetting was not a zero-sum game of project death matches but a consensus-driven community united in its goals.
This process actually reduced the “fog of war” problems found when dealing with complex biological systems, even as we got larger.
Openness can enhance all the other lessons.
These lessons I learned are interconnected. The handicaps of a large company can be overcome by increasing openness and leveraging creativity, in order to enhance information flow. The faster information gets to the right people, the faster the company can move out of the gray and towards decision points that are successful.
Any company that can utilize these 5 lessons will be well on the way to a successful and important one. And its influence will not wane, even if it disappears from the Web.
Immunex proves that.