When one party provides services to another party, he is called a “service provider.” It is the oldest, most widespread relationship in human history. It must work! And it does, in every industry and in every culture.

It is almost inescapable today that CROs and biopharma study sponsors refer to each other as “partners.” New contracts for long-term or multi-program work assigned to CROs apparently are not exciting enough for what they are — long-term relationships based on significant awards of work — and need to be termed “partnerships.” The semantics are obviously important for all sorts of reasons, having almost nothing to do with the achievement of efficient clinical research. The partnership concept is also seriously misleading and sometimes dysfunctional. I suggest the heresy that there is nothing to be gained by characterizing service providers as partners. Indeed, using language accurately and managing accordingly can lead to better performance.

Everyone who provides a service these days wants to position themselves as your partner. Your dentist is your “oral health partner.” Your car mechanic is your “automotive mobility partner.” Really? The Oxford Dictionaries define a partner as “a person who takes part in an undertaking with another or others, especially in a business or company with shared risks and profits.” Does your dentist share the risk of toothache if she does not completely clean out your cavity? Does your auto mechanic share the risk of being stranded at night if he does not securely reattach a wire? Or stated positively, does your dentist feel the relief in her mouth when your abscessed tooth is pulled? Does your mechanic get to drive your car after he’s fixed it?

These are not silly analogies. Except in the rarest of circumstances, no matter how often the term “partner” is misused, a CRO is, like a dentist or car mechanic, simply providing a service to the sponsor. A sponsor and a CRO both may benefit from a successful study or both be harmed by an unsuccessful study, but they do not share risks and profits in kind nor in degree. Relatively speaking, the CRO’s risk is low and its modest profit is predictable and near-term. In contrast, the sponsor’s risk is high and so are its potential long-term profits.

When one party provides services to another party, he is called a “service provider.” It is the oldest, most widespread relationship in human history. It must work! And it does, in every industry and in every culture. Somewhere along the way, the relationship in which a CRO is a service provider and a study sponsor is a customer has become passé, frowned upon, and almost shameful. There is nothing wrong with this kind of relationship! In a properly defined customer/provider relationship, the roles and responsibilities are clear, the risks are articulated and objectively managed, the tasks, costs and profits are short-term and straightforward, and both sides care just enough to ensure a successful result. If not, the relationship is broken and a new one is formed, quickly and with much less angst than in a divorce of partners.

Let’s look at three substantial reasons CROs and sponsors are not partners: profit, risk and care.

Profit

If you and I were partners in a lemonade stand and the price of sugar went up 20%, we would sit down and discuss what to do: Should we raise the price? Switch to a corn syrup sweetener? Make lemonade that is less sweet? We would both care about our product’s taste, market success, and cost to produce, because it is “our” business. But if it were my lemonade business and all you do is mix the ingredients, or sit by the road after school and sell the lemonade, what do you care about the price of sugar? How does it affect your ability to stir, or shout to passing cars? It doesn’t, and it shouldn’t.

Indeed, there is hardly any overlap between the business of a biopharma and the business of a CRO. The CRO performs functions parallel to and derived from what biopharmas do, but the businesses are completely different. The core financial drivers of the two entities are profoundly different; the owners therefore have profoundly different interests and concerns.

A drug in development represents a decade’s investment that is unlikely to, but might, lead to a brief but enormous profit. A staged succession of successful development projects can produce some continuity of business over time and has created some of the largest corporations in the world. Investors judge biopharmas based on current revenue and profits, of course, but also on the likelihood that their science and investments could produce a series of low-probability successes.

In contrast, CROs live on numerous, relatively short-term projects from multiple customers, with relatively low but predictable margins. With enough customers and staged projects, CROs can grow steadily with relatively stable profitability. Investors pay more attention to the current financial results of CROs because they are more predictive of future results (squishy backlogs and multi-year “partnerships” with sponsors notwithstanding). There are no blockbuster drugs in the future of any CROs. On the other hand, the failure of a study or the loss of a customer is bad news, though probably not catastrophic.

These fundamental differences in the business models of biopharmas and CROs drive fundamental differences in their attitudes toward a given study. It is impossible to align the goals of two such different organizations in an effective partnership. On the other hand, the customer/service partner relationship is designed to handle such differences. You pay your dentist for competent work. You pay your car mechanic for competent work. It’s a bit of an oversimplification, but why not just pay your CRO for competent work without all the “partnership” trappings? Look at any recent press release announcing a new sponsor-CRO partnership. Every single service or advantage listed as part of the partnership can be purchased individually, or in any combination, from that CRO by any sponsor at any time without a partnership agreement.

Risk

It should be obvious to even the most ardent proponents of “partnership” labeling that sponsors and service providers do not share the same risks. Do both parties have “risk”? Sure, but the nature of the risks is quite different, and they are certainly greater for the sponsor. To suggest otherwise is misleading.

The service provider’s risk in non-performance is mostly one of tarnished reputation. While news of non-performance spreads quickly in our intimate industry, such news is far from rare and the responsibility for failure is usually obscure. Theoretically, a service provider could be investigated by regulatory authorities, but the number of such incidents is very small. A greater risk for service providers is not being paid, although sponsors are notoriously loathe to pursue any sort of penalties for their service provider’s non- performance. If there is a sanction, it would most likely be loss of additional work. But anyone familiar with CRO contracting knows that sponsors routinely continue to give work to service providers who have failed them in the past.

Compare this level of risk to that of the sponsors. A poorly conducted trial risks regulatory rejection, late market entry (which could significantly reduce lifetime revenues), patient safety, and the return on millions of dollars of investment. It threatens international reputation, rapport with the medical community, and employee morale. At best, it means delay, re-work, higher costs, or even repeating the trial. Clearly, the sponsor has much more to lose.

Care

Considering the above differences, it simply cannot be said that the service provider cares as much about the performance of a clinical trial as the sponsor. This is not to say that service provider personnel are not professional and do not take pride in their work — surely most of them do. Neither are all sponsor personnel undyingly committed. But, no matter how professional your dentist, it’s not her tooth; it’s yours. Your mechanic is not repairing his car; it’s yours.

The practical result of this gulf in caring is potential (not universal) problems, such as CROs assigning junior people to a project that is mission-critical for the sponsor, or a slow response to a data quality issue or protocol revision, or tolerance of process inefficiencies that the sponsor would deem unacceptable (if it knew about it).

At the most fundamental level, it is impossible for organizations with such different interests to be “partners.” But can they play a critical role in each others’ success for high mutual benefit? Of course.

Improving the Service Provider/Customer Dynamic

The practical way to ensure that the relationship of service provider and customer will work effectively is for the customer to actively manage three basic dimensions: cost, risk and care.

Cost Management

Effective cost management begins with clarity on both sides of the relationship. Many sponsors, even those with years of outsourcing experience, put out RFPs to CROs that are too vague, too open-ended, and with insufficient detail, especially for contingencies. CROs are equally guilty of responding in kind, including vague assurances for specific contingencies. CROs so love the partnership concept because they can use it as a license to issue change orders without restraint, and change orders are where the profit is. If a sponsor complains about excessive change orders, the CRO that has a major multi-year “partnership” agreement with that sponsor is very likely to assert that this is what partnership means — in return for dedicated staffing, guaranteed capacity, and such, the service provider needs to be able to make up for the unexpected (a given in clinical research) with a steady flow of change orders. Since it is a “partnership,” both parties are “in it together.” The CRO can submit change orders with confidence because both parties “share” in the successes and failures of the partnership.

Eventually, however, the sponsor may realize that the partnership is not working as expected and be brave enough to upset the partnership status quo, a rare event in today’s outsourcing environment. Clarity of expected costs and responsibilities, honest expectations of a trial’s changeability, analysis and prevention of change orders, and a steady eye on integrated cost reporting will help both parties plan their timelines and resources better in a healthy service provider/customer relationship.

Risk Management

The first step in proper service provider/sponsor risk management is to recognize the imbalance of risks and start to manage the project accordingly. Accept that the two parties are not partners sharing equivalent risk. Risk management is best addressed through clear project accountability, which means placing project control in the hands of the party with the most risk (the sponsor), and detailing what the service provider will be accountable for, within the limits of their responsibilities. This means dropping the fashion of both parties having equal voices in a Joint Operating Committee (JOC), which suffer from the failings of all committees and also the ambiguity of who’s in charge.

The parties should agree upfront on a clear procedure for replacing CRO personnel who are not performing adequately. This procedure must spell out how such replacements will be made quickly, rather than allowing the CRO to stall for time, which the trial cannot afford. The parties should also agree on what trial performance metrics are important and how often they should be reported, with what explanations. For instance, EDC tools can tell the sponsor if, when and how often CRO personnel are “in” the sponsor’s data, which is one indicator of whether the CRO is actively performing the necessary work. Good CROs embrace such reporting as an opportunity to prove the quality of their service and other competitive advantages.

Caring

Eliminating the partnership myth will help the sponsor understand the inherent limits on how much a service provider will care about what is really important to the customer. It will also help the service provider understand why the sponsor needs to be in charge, no offense intended. The parties can then act and communicate honestly according to their natural, differing priorities. The things each party cares about will not be the same, nor do they need to be. The motivation for performance, not a shared reason for it, is important part.

Conclusion

Both sponsors and service providers have different, but powerful reasons to see the sponsor’s development succeed. No sponsor today is likely to return to a time when it did all clinical development work itself. While shared objectives are essential to successful clinical study execution, pushing the “partnership” myth has been useful for CRO marketing but has done no favors to clinical trial execution and customer management.

The criticism of a pay-for-service relationship in favor of something somehow more lofty is misplaced and misleading. The oldest form of relationships may be the better way to develop new medical therapies.

“If only we were as adaptable as viruses. Our clinical research processes unfortunately evolve at a much slower rate.”

Flu virus evolves continuously, so much so that a new vaccine is needed every year. It’s the fast replication of virus generations that produces such rapid mutations. Typical viruses change so rapidly, through natural selection, they are able to respond quickly to new environmental conditions. If only we were as adaptable as viruses. Our clinical research processes unfortunately evolve at a much slower rate; clinical research can often take a decade or more to make much less dramatic “genomic” changes. Can we learn as fast as natural selection?

Excuses

I can hear the objections to this comparison already: biology is “objective,” process is “soft.” The excuses we have for the slow pace of change are many:

Regulatory requirements. Biopharma quickly takes refuge in the argument that the reason why things don’t change is because we are a regulated industry. This is rarely relevant to process discussions. There are actually surprisingly few regulations on clinical research process (for instance, nowhere do the regulations say we must perform 100% SDV), and on more than one occasion (I can think of at least three in the past decade), FDA has actually led the industry by issuing guidances describing processes more advanced than industry was initially willing to accept.

It’s not broken so why fix it? This is a typical reaction among those working in a long-lived, experienced and successful (i.e., drugs approved) development organization. But it is broken, and if you look, you will see the facts to prove it. You will find inefficient policies, overlapping responsibilities, unexploited technologies, and political speed bumps slowing your processes.

The “not broken” argument is used a lot when considering ePRO, for instance: “well, I know the paper data is probably bogus but it’s the standard of approval.” Well, no, not anymore, but how long does it take for that fact to filter through research departments who are still rejecting ePRO because it costs too much? That in itself is misleading: what costs “too much”? Don’t they really mean that they didn’t adequately budget for the cost of reliable data?

We don’t have enough evidence yet to justify the upheaval of change. This is false rigor and laziness. When management resists change for a lack of “data,” ask them where is the data to justify current processes? Try taking a zero-sum approach to how you work, and see what proportion of your current processes would survive the analysis.

It’s an unfair comparison – viruses are much simpler organisms than human. – Ah, now we are on to something. If only we could examine human behavior the way we can a genome.

A Method

Actually, we do have a method for accelerating human evolution – it’s called learning. And in the jargon of process improvement, we call it “Lessons Learned”. Many biopharmas would point out that they do Lessons Learned exercises routinely. So the questions are how do we do them, what do we do with them, and where can we see the impact?

One concern is that Lessons Learned exercises have become perfunctory and disrespected. Is the lesson from the “Lunch ‘n’ Learn” session in the cafeteria excreted later in the day? Is the Lessons Learned binder carefully shelved, accumulating dust like autumn leaves? Are those who run these sessions prepared to make them productive? We are neither born teachers nor born learners.

Is the whole Lessons Learned concept thought of as a training function (and therefore underfunded and mostly ignored)? At best it is probably a Clinical Operations function, so the lessons do not permeate medical, biometrics and planning departments. More importantly, where are the lessons coming from? That is, do we have the skills or training to analyze properly the experience before us, and derive the lessons needing to be learned?

Information technology is an important tool ready to help us learn – data on clinical research performance has never been so plentiful or accessible. But no company does a comprehensive job of mining that data for the most relevant process indicators, and no technology vendor does as well as it should in providing easy tools for analysis.

If we’ve outsourced research operations, do we think we can outsource learning too? Are those we have outsourced to learning from their experiences? Must we pay them to learn, otherwise they will have no incentive to improve?

We can and must do much more to learn from our past work. First, learning needs to be incorporated as an expectation and requirement of all functions, and not shuffled off to some ancillary group. I am amazed, for instance, when study teams are not held accountable to document the good and the bad of their experiences. Who were the high performing investigators? What approach worked best in getting drug supply produced on time? Which data manager should be promoted because she was so effective? And I am equally amazed when functional specialties do not take the time to assess their effectiveness, to learn from each other, to routinely identify means of self-improvement.

Be Afraid, Be Very Afraid

The best source of lessons learned is in the head of our experienced staff, and here is where we should be very concerned. At the moment when we most need to learn from past experience, that experience is walking out (or being shown) the door. As we send staff to CROs or retirement or forced career changes, we are losing the chance to capture and define and articulate the lessons they could be teaching us.

We have to care much more about learning. As always that starts with upper management. It continues by allocating money and time. And it is capped by a willingness to learn (change) and a method for applying lessons quickly.

George Santayana famously said, “those who cannot remember the past are condemned to repeat it.” That was more than a hundred years ago. And here we are decades later with that lesson still unlearned. Our industry cannot afford to wait that long to apply the many lessons available at our fingertips. We may not be able to match the adaptation rate of viruses, but we will fall victim to the disease of ignorance unless we organize the learning from our experience.

©Waife & Associates, Inc., 2011

What did you do in the last meeting you attended? Assuming you weren’t running it yourself, did you listen to what everyone was saying? How many times did you check your email? How many websites did you visit? What proportion of the meeting would you say was important to you personally in getting your work done? I know the answers to these questions. I don’t know the answer to this one: why do we do this to ourselves?

After securing food, shelter, clothing and sex, early humankind’s next step (or detour) on the road to civilization was probably having a meeting. The next step after that was to not pay attention when someone else was grunting about something irrelevant to the next wooly mammoth hunt. Today we can look back with pride and marvel at how far we have advanced from our primitive beginnings. Now, unlike our distant ancestors, we have agendas!

I have written before about the strange rite of the weekly meeting, whose frequency is determined by an arbitrary astronomical rhythm, and not at all by a connection to the work at hand. But if meetings are perhaps inevitable, at the very least let’s pay attention! What is rampant throughout clinical development, and probably the entire pharma enterprise, is what should be called “unattended meetings.” We should suggest to Microsoft, makers of the ubiquitous Outlook software for scheduling meetings, that they add another choice to meeting requests other than “Accept”, “Accept Tentatively” and “Decline”: “Attending But Not Listening.”

This Time, Technology is Not the Answer

One thing we know, technology has only worsened the unattended meetings phenomenon. Twenty years ago the worst thing we probably did when we weren’t paying attention was to doodle (with a pen, on a piece of paper). It is technology that enables us to supplant doodling with checking, and writing, emails; surfing the web on phone, tablet or notebook; or just doing other work. Simple telephone conferencing has created the modern marvel of enabling us to not pay attention to meetings from across the globe, in real time! Alexander Graham Bell could never have anticipated such value added.

What’s the Cause?

Why are we not attending to the meetings we attend? The list of reasons is long: certainly a lack of relevance and a surfeit of meandering discussion is the first cause. But there are worse reasons, such as attending the meeting only in order to assert authority, or defend a staffer from another department, or because your company values “consensus” at all costs, or because we are afraid of being left out. Then there’s misplaced politeness, basic mistrust, and the ultimate time-waster: if she’s going to come, you’re going to come too!

One of the key motivations for attending, yet not attending, meetings is that there may in fact be a moment in that meeting – brief but critical ­– when there will be a kernel of information that actually is relevant to your work. But what a cost to get it! And what a risk that you’ll miss it, even if you are not surfing!

The root cause of this time-wasting may be that we mistake a meeting as a medium of getting information. There can hardly be a more inefficient means of sharing, getting or requesting information. Instead, this is what today’s technology is indeed useful for. Clinical development is an arena filled with software applications generating data at our fingertips. While the software might not be good at turning that data into information, that’s what we humans can do, as long as we don’t spend our time in meetings instead.

Justifiable Risk

There are really only two good reasons to have a meeting: discovery, and decision-making. Discovery is a powerful purpose; decision-making is a necessary one. A meeting for discovery is how a staff group formally explores a research or process problem; if well run, these meetings can be exhilarating, and put other meetings to shame. Group decision-making may indeed be critical, and if so, no one should feel the need to drift to the electronic ether while the decisions hang in the balance.

Short of eliminating all of the meetings that we actually do not attend, we should be helping each other scrutinize the reasons why a meeting is being called. Is it because we don’t think about the topic until we walk into the room? Is it because the way we use our technology tools is failing to inform us? Is it because we outsourced the work but can’t trust who we outsourced to? Is it because we are afraid to act on the obvious, without using a multidisciplinary group to cover our tracks? The time we are wasting is our own. No one can afford it, and we all pay for it. Let’s have a meeting to discuss this.

With the rise of mega-CROs, and the widespread willingness of large pharma to outsource clinical development to them, is there anything left for a sponsor’s clinical development group to do? Indeed yes. More than ever, there are hyper-critical functions that sponsors must execute effectively to produce success in clinical development. This can be summarized as implementing a performing process for managing outsourced development.

It is a truism that pharma “hates” their CROs, while at the same time pharma is using those same service providers more and more. The current fashion for “strategic” relationships has not ameliorated this at the operational level, although perhaps at the executive level. Much like what happened with software providers, where the failure of clinical development applications could often be traced to poor implementation by the sponsor customer, so too can frustrations with CRO performance be traced to suboptimal vendor management by sponsors. Both sponsors and their providers can benefit greatly from improved provider management, which can create a positive cycle of trust, high performance, and higher value.

Those with long outsourcing experience understand that it takes robust internal project management to ensure high provider performance. This is also true when offshoring functions. Too frequently, however, this increased oversight is not budgeted for. Most importantly, as similar as the functions of clinical operations are, whether done in-house or by a provider, the process for managing, contracting, measuring and ensuring high performance must be substantially different in the two scenarios. Again, too frequently, new processes, and the subsequent re-definition of roles and responsibilities, along with concomitant training, are not designed and implemented.

Moreover, if outsourcing was chosen in order to eliminate interdepartmental conflict, it is likely that the conflict will continue when the services are outsourced; essentially the conflict is simply outsourced along with the function. If clinical and data management mistrust each other when both are in-house, this mistrust is likely to continue if outsourcing one or the other is all that is done to try to fix it. Without open recognition of existing issues with governance, interpersonal conflict, differing philosophies (about data cleaning, about the purpose of monitoring, about the role of science, etc.), these factors – all creators of suboptimal performance – will persist and likely be exacerbated in an outsourced operational model.

How would we recognize a performing process? There are a number of indicators:

· Fewer agenda items about conflict or performance at the Joint Operations Committee meetings

· Lower staff turnover on the service provider team, and seamless execution when it does happen

· Full transparency for both sponsor and provider on the status of all trial elements (from protocol changes to deliverable delays) and on the reasons for these changes

· Fewer amendments, fewer change orders

· Both sponsor and vendor feeling safe to tell each other the truth

· Ready knowledge of which vital few metrics are important , and where “we” (sponsor and provider) stand on those metrics

· Easier contracting and faster study startup because of the trust and transparency built – not from a “strategic” handshake, but from demonstrable high performance.

We can’t successfully outsource what we don’t respect or don’t understand. The long-term danger in prolonged outsourcing is that the fundamental and critical knowledge of operationalizing a protocol will have disappeared from sponsor staff. We must be vigilant to prevent this. Active and fully funded vendor management is one step in preserving operational understanding in the midst of scientific inquiry. This ensures an environment that creates the opportunity for innovation in high performing processes and development excellence.

Whenever something is wrong, something is too big. – Leopold Kohr, The Breakdown of Nations

Size does matter when it comes to optimizing clinical development, but perhaps in ways we don’t expect. Certainly few in the clinical research industry seem to believe that “small is beautiful” anymore. Mergers of large pharmaceutical companies seem to occur monthly. Acquisition targets now even include large biotechnology companies, creating ever larger entities with contrasting cultures. The mergers and acquisitions continue despite producing equivocal benefits. The usual explanation is that somehow a larger mass of researchers will be (this time) the “critical mass” that will produce results. Meanwhile, CROs are, as usual, mimicking their customers, and so they too seek to grow as large as they can. They argue that being “global” (i.e., big and broad) is equivalent to providing the best service. Neither of these assumptions are absolutes, and there are good reasons why they should be re-examined.

Large Can Be Pretty

Merged companies usually merge their individual internal functions. So two monitoring groups become one, two data management groups become one, etc. Why? Well, either it is assumed that one department is more efficient to maintain than two, or because one company’s department is perceived as a better performer than the other. Ideally, a smaller proportionate overhead (so-called “non-performing assets” like trainers and standards/quality assurance personnel) is needed for the larger group of “deliverable-performing assets.” Each company’s methods can be examined for what seems to be more efficient, and these “best” (or better) practices can be applied to all.

Merged functions are also assumed to be more fungible (i.e., interchangeable, providing more flexibility in assignment). In this perspective, more personnel are always better than fewer. More people mean it is easier to respond to boluses of work or new initiative. More people can mean more time for process improvement because there are more “idle” hours to spread around to work on such initiatives.

Large is also assumed to be required for trial execution globally. And all trials are trending larger as we know, for various regulatory and scientific requirements, so we must be larger to respond and support. The large CROs base their business model on scale, and in fact do win a lot of business on scale alone.

And Ugly

While large means fungibility, it also means mediocrity, by definition. Why? Because of the inevitable human bell curve. The abuse of school grading (where a “B” is the new “C”), and Garrison Keillor’s Lake Wobegon (“where all the children are better than average”) notwithstanding, in real life if one gathers a large group of people together — be they the proverbial “man on the street” or all university-trained statisticians – the bell curve will apply, especially as the sample size increases. The more people, the larger the middle of the bell, the larger the pool of mediocre, or “average” talent. It’s not possible to hire a hundred outstanding data managers, or three hundred outstanding field monitors. For all sorts of reasons: vetting, timing, training, integration, consistency, quality control, motivation, life circumstances, instincts, etc. Nor, for the same reasons, is it possible to train them all to the same point of excellence, despite many companies and vendors assertions to the contrary.

Of course, large also exacerbates the extended litany of human foibles: bureaucracy, hierarchy, misunderstanding, mis-communication, mistrust. And large runs counter to the general corporate trend to minimize or reduce overhead, especially facilities. This in turn drives large organizations to geographically disperse, encourage working at home, and ultimately to outsourcing. Each of these in turn creates new dilemmas.

Let’s take a typical example of the impact of CRO scale on the quality of their services. First, all research sponsors have their tales to tell about mediocre project managers, who turn over inordinately. The big CRO has lots of project managers to throw into the fray, but a sponsor does not benefit from volume, only quality. And if a sponsor chooses to rely on a large CRO for a large trial it is running, precisely because of scale, and then learns that the CRO project manager is actually working from home a dozen states away from the sponsor and the rest of the CRO team, what advantage did we get from scale, from large size?

Small Can Indeed Be Beautiful

It may feel like scale is a necessity in today’s global research enterprise, but it should not always be considered an advantage. Indeed, some of the best industry operations examples are small, sometimes referred to as “skunkworks,” which live inside large organizations.

Why is small better? A number of reasons:

• A better environment to get the most out of individual talents

• Creates a group which is more unified, dedicated, and motivated

• Provides otherwise “silo’ed” specialist professionals with a common purpose

• Enables the goal to be “within sight,” leading to interim, achievable victories

• Easier to control

• Easier to prevent mediocrity –the bigger the scale, the more certain the bell curve rules.

One of the truisms of a career in pharma clinical development is that so many people can go through their entire working history and only touch on a series of compounds or devices passing by. And as functional groups get bigger, roles within the function can get ever more specialized, further reducing the scope and sense of ownership for each individual. Smaller functional groups help preserve a breadth of professional contribution, and can be structured so that it is more likely the staff will see a candidate through the end of clinical development.

The Best of Both

It is very important to emphasize that pharma can and should structure their development organization to gain the best of both large and small models. A key component of this is the misunderstanding people have in reacting to words like “standards,” “innovation” and “scale.” These do not have to be incompatible concepts. While we associate “standards” with large bureaucracies, and “innovation” with small teams, we can definitely have both.

Innovation can operate within standards, especially those established primarily as professional foundations, which is what most clinical development folks mean by “standards”. And of course ultimately, the regulations are the ultimate common standards that we all must live by. Even within our highly regulated process, innovation can and does flourish, if the opportunity is provided. Most importantly, small scale does not require loose standards in order to be effective. In fact, standards and other controls free the small group to focus on their output and not their process or compliance.

Nor does small have to mean inefficient. Using the principle of core functions, one can maintain professional and process standards as much as desired across the company, thereby ensuring consistent use of methods known to be optimally efficient in your company. A functional core is not a silo under another name; rather it is a professional development center which provides the time and place for standards creation & maintenance, and a place for common training and professional development. This provides two additional key benefits: it addresses the concept of fungibility without mediocrity, and it provides employees a familiar professional identity that will mean “security beyond the team” during their career.

E. F. Schumacher wrote what was then a landmark book, in 1973, called Small is Beautiful. Among other nuggets, he memorably and presciently wrote:

The most striking thing about modern industry is that it requires so much and accomplishes so little. Modern industry seems to be inefficient to a degree that surpasses one’s ordinary powers of imagination.

This certainly strikes powerful echoes for clinical development, more than a generation later. For a large organization to work, according to Schumacher, it must behave like a related group of small organizations. Interestingly, the title of his book was in turn a quote from his teacher, Leopold Kohr (quoted at the beginning of this column), who went on to say:

It is because human beings, so charming as individuals or in small aggregations, have been welded into overconcentrated social units.

Large or small, size does matter. And what matters most is whether we are managing correctly to the size we have chosen. Functional managers personally face the challenge of making the choice succeed. In the end, small may indeed be beautiful, and should be tried more often in clinical development.