R&D Productivity: A Dialog

DR. SOCKS:   Shall we discuss R&D productivity?

DR. LYSIS:  It’s on everyone’s mind in the pharmaceutical industry.

DR. SOCKS:  Indeed in many industries. Perhaps we should start with a definition.

DR. LYSIS:  It’s all about producing. Let’s get production up in R&D.

DR. SOCKS:  Don’t they have approaches for doing that already – isn’t that what Six-Sigma’s all about?

DR. LYSIS:  Yes, but in the pharmaceutical industry we refer more to effectiveness in getting valuable products out the end of the research pipeline. Pharmaceutical companies have been employing Sigma techniques to optimize R&D activities for years, yet products still aren’t making it into the marketplace. Many make it far down the R&D production line, but in the end they often don’t show enough improvement over existing products to justify the expense of launching them commercially. In manufacturing we know what we’re building and we can more accurately predict increased sales as a result of our optimization work. In R&D we simultaneously design and build the product. We can be fully optimized in our build, and yet still be working on the wrong design.

DR. LYSIS:  A 2006 pharmaceutical industry survey+ on R&D productivity uncovered the inconvenient fact that it’s all about individual behaviors. Fix the behaviors and you’ll fix R&D productivity. It’s inconvenient because most managers would rather not deal with other people’s behaviors: it’s easier to ignore them and focus improvement efforts on the inanimate (e.g., technology, reorganizations, and acquisitions).

DR. SOCKS:  Fix the behaviors. That’s a pretty generic statement. Knowing you I’m pretty sure you’re not caught up in the circularity of the argument: fix the behaviors (i.e., make them more productive) and you’ll fix R&D productivity.

DR. LYSIS:  Thanks. We look at specific behaviors that affect R&D productivity, but we must also step outside any narrow definition of these behaviors, i.e., thinking we can fix a behavior and hence gain more productivity. It’s like the carnival game of whac-a-mole, where the mechanical mole pops its head out of a hole and you smack it down with a mallet only to have another one pop out of a nearby hole. The better you get at smacking them down the faster they come out of the holes. As soon as you smack down one unproductive behavior or habit, another one pops up to take its place.

DR. SOCKS:  Agreed behaviors are ugly. That’s probably why most people would rather not deal with them. Is it worth it?

DR. LYSIS:  Exceptional leaders intuitively know how to deal with other people’s behaviors, and their own. One approach is for corporate management to focus their time looking for that next generation of exceptional R&D leaders. Our approach is to take merely good leaders and make them exceptional. This means we must get them focused on behaviors. For example when the deadline is tight and the customer is screaming, exceptional leaders do not heedlessly jump in to do the work themselves. The screaming customer is a chance to demonstrate leadership by helping others to continue to work coolly and effectively toward a solution.

DR. SOCKS:  So elaborate on what you mean by dealing with behaviors? At a certain point it’s a matter of experience and maturity: the school of hard knocks.

DR. LYSIS:  We can’t set it up so leaders overcome a hardscrabble life, learning to deal with every conceivable type of behavior, and then through sheer strength of will make it to the top. But, we can design mechanisms and controls that take care of 90% of behaviors typically found in R&D, giving leaders more time to focus on the remaining 10% of troublesome behaviors. We design autonomous mechanisms that leverage and grow today’s limited leadership capacity in R&D. We help merely good leaders become great by setting up the work in a way that more or less autonomously deals with unproductive behaviors.

DR. SOCKS:  Autonomous mechanisms? Are we computerizing evidence gathering and decision making?

DR. LYSIS:  Couldn’t be further from the intent. It’s the people that matter the most in creative enterprises. We design our approach to the work of R&D in a way that elicits productive behaviors from most of the team most of the time, allowing leaders to give individualized attention to those most in need. Behaviors become in effect the critical design criteria for all our work activities. In creative endeavors individuals often have their personal identity wrapped up in their work. Design the work to elicit productive behaviors and you talk to the core identity of many individuals.

DR. SOCKS:  And how does this relate to traditional approaches to managing behaviors, for example incentives, bonuses, promotions, etc.

DR. LYSIS:  In R&D at least, it exposes their role as a crutch. Their intent is to try to get something done despite poor management. Why should you pay a bonus to get someone to do a great job? If the work is exciting and structured in a way that allows them to do a great job, then what need do we have for further incentives? Incentives, bonuses, etc. are signals+ of poor management. Another problem with most traditional approaches and tools is they are often open to gaming. They tend to be one-size-fits-all+, and when the measure doesn’t fit my need, I’ll look for a way to game it.

DR. SOCKS:  So that’s it. We look for ways to increase productivity by focusing on the work that needs to get done and how it affects behaviors? For example I examine the work activities of governance and see how they affect team productivity. We analyze the behaviors of both the governors and the governed. For example, when the governance body personalizes work-in-process (e.g., prototype+ Acme001, Acme002), we show how this makes individuals stay within the walls of their specialties, to the detriment of productivity. Why do the governors behave this way? How do we design governance to get rid of these behaviors? When we complete this analysis across all the work areas we end up with a comprehensive program for increasing R&D productivity (Exhibit 1).

Exhibit 1:  Template for increasing R&D productivity in the pharmaceutical industry. See here for a discussion of this image.

DR. LYSIS:  For the pharmaceutical industry at least that’s directionally correct: customized by team and for a specific phase of the work. We avoid one-size-fits-all in creative endeavors. The behavioral needs of teams change as their work matures. If something no longer works we quickly discard it and try something else. But in general you’ve got it. And this thinking may apply as well to other industries that rely on creativity in product development. But let’s think more broadly about this.

DR. SOCKS:   That’s where I was going. Here’s the generic template. For any industry we list the driving strategic goals, and the means toward achieving those goals. So if an industry needs increased sales, the means may be as simple as getting more qualified feet on the ground. We then list all the key work areas and the behaviors needed to achieve the means. We redesign the way we do the work areas so that the work itself more or less autonomously encourages behaviors that support the strategic goals.

DR. LYSIS:  Precisely, and the answer you’ll get will be just as facile as the analysis you described.

DR. SOCKS:  Explain.

DR. LYSIS:  Our discussion so far is based on a retrospective view of how I came up with the solution for the pharmaceutical industry. It’s a model that cleverly hides much of the underlying complexity. The path to the solution was much more convoluted. But there are still lessons. The key was in the voyage.

DR. SOCKS:  Sounds very Confucian. Continue.

DR. LYSIS:  I was smitten by an idea, the idea of Explore-Confirm+ developed by Dr. Orloff of Novartis. Based on my 20 years experience in the industry it fit too well not to be true and I was determined to see it work. I started off quite naively with the notion that the path to Explore-Confirm lie simply in improving evidence, so I spent close to a year investigating Standards of Evidence, finding the best way for entities to gather and analyze evidence, make a decision, and take action based on that decision. By sheer coincidence I stumbled across Nassim Taleb’s book Black Swan which gave voice to my discomfort on how evidence is gathered and analyzed in the pharmaceutical industry. I felt I was on my way.

DR. SOCKS:  So far that’s consistent with what we’ve been saying.

DR. LYSIS:  I uncovered the obstacle of the decision maker and the need for greater flexibility in the project mandate for teams. I can’t improve effectiveness in evidence gathering without giving teams the time and resources they need to consider valid alternate explanations that fit their evidence. Decision making needed to change if we were to change evidence gathering. Nothing new yet. But what I found was that in tackling each new obstacle, the solutions surfaced new questions that I had never before considered. For example when I studied solutions from the field of economics I was drawn into the concerns common to that profession: the revelation+ principle, adverse selection+, etc. These could not be safely ignored. Each new solution opened up new obstacles. My journey carried me across two dozen academic disciplines, for example, education theory, the evolution of cooperation, and decision theory. Hundreds of new questions surfaced. If I had started with only a simple template, my solution would have been greatly impoverished. For example, I never would have considered the importance of regrets+, churn+, or sunk costs for investment behaviors+. In each industry you must take a similar voyage and come up with solutions unique for that industry. Use the content of your partial answers to help you build the template, but remember the template is merely a mnemonic device.

DR. SOCKS:  For the moment I’ll pretend I know what you mean by regrets, churn, etc. Behaviors will always be important, as will be your focus on the work areas. It merely remains to determine toward which ends.

DR. LYSIS:  Behaviors are important certainly in creative endeavors. It’s also true that 80-90% of staff merely want a paycheck. The mortgage is due each month. Our focus on behaviors is more or less intense depending on how much we rely on the self motivation and creativity of our employees for our productivity.

When you say it merely remains to determine toward which ends, you overlook the hundreds of obstacles, rejections and false leads I faced during the course of developing a solution. If I hadn’t been smitten with the elegance and explanatory power of Explore-Confirm I never would have finished. You need the dedication of a researcher willing to continue to push ahead, often in the face of contradictory evidence, for the sake of a compelling idea. You need that original Explore-Confirm ah-ha moment for each industry in order to pull off such a revolution in thinking.

DR. SOCKS:  Granted, but as you say most folks in industry are just looking for that next paycheck and prefer facile answers. You’re going against human tendencies to ask for so much extra curricular devotion.

DR. LYSIS:  There are some key lessons we can learn from my pharmaceutical industry experience. First, as we’ve seen, the thinking behind the original insight, Explore-Confirm, is far removed from the final behavioral solution. It barely rates a mention. This is consistent with what we see with insights in the biological sciences. The final drug is often far a field from the intent of the original biological insight. Don’t settle for the first, or even the first few answers that grab your imagination.

Second, there are many insights both in the pharmaceutical and management sciences that just go nowhere. History is written by the victors and is often viewed through the lens of retrospective distortion+. We rarely see the underside of ah-ha moments gone bad. Who wants to write about failures? We admit humbleness in our ability to analyze and organize large endeavors involving that complex human animal. We build flexibility into our program for change.

It may turn out that improved evidence will have nothing to do with R&D productivity. Perhaps all we can do is wait for that next exceptional R&D leader or further maturation of the basic biological sciences. In the meantime my arguments do serve as an organizing principle for change. In the pharmaceutical industry it’s clear what we’re doing now does not work. We need people motivated and organized around a compelling vision for change, yet remain fully aware that implementation may take us far a field from our original intent. It’s the journey that counts.

DR. SOCKS:  So have we concluded our dialog on R&D productivity?

DR. LYSIS:  Between us yes, but it’s just starting in industry. Maybe next time we’ll have a dialog on the productivity of government.

DR. SOCKS:   I look forward to that discussion.