Skip to main content

Drug Development

Cutting (And Re-Cutting) the Research Pie

Here’s something that goes on inside almost every biopharma research organization, but always happens behind the scenes if you’re not right there at the meetings: resource allocation for projects. That “almost” is in there because the smallest companies don’t have this problem in the same way, since they tend to have one project that has to succeed. Arguing goes on (most certainly!) about how to make that happen, but it’s only when you have several projects or therapeutic areas going at the same time that you get a multi-front battle.

Simply put, money, people, equipment, and time are all limited. I’m not a fan of always seeing things as zero-sum games, but inside a given organization, the struggle for these resources is pretty much just that. It’s easier when a project ends, because the people involved are presumably free to be reassigned, but projects (for the most part) don’t end without going through a winding-down period, so it’s not like there’s a huge bolus of resources that are suddenly open for use. Month-to-month or quarter-to-quarter, though (depends on the company), the wrangling goes on among projects that are all still in progress.

“Number of chemists” is always a prized variable. The need for med-chem support varies over a project’s lifespan, but there are long stretches where it’s crucial. You see all sorts of behaviors, productive and otherwise, when it comes to allocating people. There are project (and project leaders) that jealously guard their head counts, because they know that people who are allowed to leave are hard to get back. So even if the peak staffing isn’t needed any more, they come up with rationales for why no one should leave at this key point – we’ve got that backup series going, we need to shore up the patents, we need to make intermediates for the tox scale-up, etc. There are also projects that are (if you listen to their leaders) seemingly always starved for hands to work on them. I well recall a therapeutic area head at a former company who, at every single allocation meeting, made the case that the projects in his area were absolutely crippled by the inability of everyone else to realize that every single chemist in the department needed to be working on them all the time. The first time I heard that speech, I was impressed by the boldness of the approach, but as time went on I realized (one) that it was pretty much the same pitch every time (“You’re killing us over here!”), and (two) that it was not having the desired “Cartago delenda est” effect.

Some organizations have a general perception that a project just isn’t “real” unless it has medicinal chemists assigned to it. As a chemist, I certainly wouldn’t want to minimize this respectful attitude, but at the same time, it’s true that if you’re still in the assay development stage that you might not need any just yet. And it’s also true that some projects are just intrinsically going to need more hands on deck than others. If the project settles down on a synthetically difficult lead series, for example, it’s going to be a resource hog, and you’re just going to have to live with it. I was on one that had deceptively tricky chemical matter, a small ring whose various steric hindrances and tendencies toward side reactions made it a constant annoyance, and the project team was constantly having to defend both its head count and its apparently low production of new analogs – which is not a good situation to be in, because no one wants to hear you moan about how hard the chemistry is, even when it’s the truth.

Other resources that can be in short supply and worth fighting for, besides chemists, can be high-thoughput screening slots, cell culture, protein purification (often the sequel to cell culture), cloning, and the attention of assay-development specialists. Those are often early-stage issues, while later on in a project the fights are over scale-up chemistry, toxicology screening slots, the attention and time of formulations specialists, and so on. Never once have I ever worked in an organization where these things were thought to all be in sufficient supply. And I don’t think that’s even possible, because if you always had enough to handle whatever project needs might arise, you’d be guaranteed to have a good number of people who were more or less idle another part of the year.

So the key is to find a way to allocate these things without bitterness or deception, and different organizations have had varying levels of success with that. My own take is that it comes down to personalities, for better or worse. If people have good working relationships, and feel as if they can trust each other, then there’s going to be less grandstanding and dog-and-pony-ing. If you’re working in a company where these issues are constantly a source of friction and bitterness, though, you’re probably seeing just part of the dysfunction that runs through the whole research organization – unfortunately.

25 comments on “Cutting (And Re-Cutting) the Research Pie”

  1. Vader says:

    It always comes down to personalities.

    Fortunately, our superb public education system puts great emphasis on building good character and constructive social skills.

  2. biotechtoreador says:

    I imagine in most places this comes down to ‘infighting is so vicious because the prizes are so small”.

    Same thing happens in the BD department where execs fight over what to buy/partner.

  3. RM says:

    I wonder if this is attributable to territorialism. That is, by clearly demarcating the projects and responsibilities (“you are working on the X project, your paycheck comes from the X project budget, your boss is the X project lead”), you set up the “us versus them” mentality.

    I’m wondering if it would be possible to set up the project organization in a more fluid fashion, so that people are working more for the organization as a whole, rather than a particular project.

    Not quite sure how that would work, but a possibility might be the “center” model, where all the, e.g., medicinal chemists report to, and are paid from, a “medicinal chemistry center” which then takes on projects as clients. You’d still get certain medicinal chemists dedicated to certain projects (as that’s most efficient), but the center directors would have more leeway to reassign people to other projects. Resource allocation would be based on overall business priorities, rather than individual project managers’ fiefdom-building. (This would also alleviate the bloody stupid tendency of “letting go” of some of your best chemists because they happen to be working on a project which got cut.)

    … though not having seen it done, I wouldn’t know if the cure is worse than the disease here. (E.g., without a strong project affiliation, would there be as much drive to see it to completion? Would interaction between medicinal chemists and the screening team become more segmented and less efficient?)

    1. Lora says:

      We had something similar to this at Wyeth – the biologists and biochemists were franchises whose work was supported by the medchem group and the expression/purification groups. It was …OK? I guess? Didn’t seem to be any more or less efficient than anyone else at finding drugs, but it was a fairly pleasant working environment until hecking Pfizer bought us. We did get some fussing about who was where in the queue and what projects would need more than the typical support, but overall it was relatively chill in comparison to just about everywhere else I’ve been.

  4. myma says:

    This is why a prioritization of projects from the top down is important. If not rank order, then at least in tiers. I was at a medium size biotech, and it worked reasonably well to get the noise down and constrained lively debate only on those projects on the bubble in the mid to lower second tier. Lowest tier, no debate at all, no way were they getting anything, so the seniormost person would allow the speaker to wax poetic about the favorite unresourced project for maybe a minute, and that was it.

  5. steve says:

    The problem is that the decisions are made by BD guys who have faulty market models they rely on. To truly be innovative the scientists with the vision should be an equal part of the decision making process. If you’ve been to a movie lately you’ll see what I mean – the marketing guys will always go for remake #5 instead of something that’s fresh and innovative.

    1. biotechtoreador says:

      “decisions are made by BD guys who have faulty market models they rely on”

      Similar to decisions made by R&D guys….

      1. Hap says:

        Except R+D generally has something to hang their hat on if they fail – some way to learn from their mistakes. Marketing estimates (*cough*Exubera*cough*Lipitor*cough*) seem to be based on voodoo exacerbated by groupthink, and don’t seem to retrieve any useful data to improve the models (other than maybe “Don’t require a bong to deliver your next inhalable drug” and “Customers aren’t willing to sacrifice potential lifespan and major mobility and money for minor reductions in pain.”).

        1. biotechtoreador says:

          I’m fairly confident BD people also are able to learn from their mistakes. It’s easy to mock the BD folks who pull absurd market projections based on very little out of the air, but I’m not sure it’s any sillier than R&D folks getting all fired up about making the 17th ATP-pocket binding kinase inhibitor.

          1. Hap says:

            Depends. R+D people get nuked if they fail (and sometimes if they succeed). I would have assumed that that’s not the case for the business people, but a lot of sales people have been getting whacked, so maybe I’m wrong. If everyone gets nuked when they fail then it’s either “learn or die”, and they’ll probably learn. If they don’t, probably not.

          2. steve says:

            McKesson did a study they published in Nature Rev Drug Disc. 12:737, 2013. They looked at market projections by pharma for the years 2002-2011. They found that most were wrong, often substantially; even 6 years post-launch the estimates were off by 45%. That means the marketing guys can’t even predict sales AFTER the drug has been on the market for 6 years. So why have any faith in their models whatsoever?

          3. Kelvin says:

            FYI, any sales forecast (or indeed ANY forecast, for that matter) that is based on precise point estimates, is guaranteed to be precisely wrong. That is why any decent forecast should always include a range to indicate the natural degree of uncertainty given the available information. Otherwise it’s like trying to describe the quantum world of probability distributions by classical Newtonian mechanics. Bottom line is: Nobody has a crystal ball, so let’s stop pretending that anyone does, and mocking them when it turns out they don’t (unless of course they pretend like they do).

          4. Mark Thorson says:

            Error bars on sales forecasts? I was thinking this is an absurd idea, but what if we had IBM’s Watson do it? The black box says these are your error bars. What do you have to say about that?

          5. Humulonimbus says:

            It’s sales forecasts without error bars that are absurd, despite the fact that they are ubiquitous.

            It’s not just error bars, forecastsof any kind ought to be seen as probability distributions. To quote Sam Savage*: “An uncertain number is a shape.”

            *I highly recommend Savage’s book “The Flaw of Averages” for detailed and compelling illustrations that point estimates are not only guaranteed to be wrong, but in many cases will lead to unwise decisions.

  6. MTK says:

    Protecting your turf and all that is fine until the reorg or merger comes along.

    Then management and the consultants come in and look to see who’s been most productive or efficient. That’s when constantly asking and receiving more resources can come back to bite you. You better have put those resources to good effect.

  7. SP says:

    I’ll add a related version- where management has decided that certain functions need to be at a certain level, whether it’s because they have in their head that you need x chemists per y screeners, or because they’ve done some calculation that for z projects you’ll need w crystallographers per project. So then when another group has a need that doesn’t fit these calculations, you can’t get funding to support even 1 person while at the same time there are 5 open job reqs for protein expression specialists.

  8. HTSguy says:

    One interesting thing a former employer did was to give significant bonuses to the entire research site based on Phase transitions. That had the downside of having the site leadership push stuff forward that maybe should have been killed (although it still need to get through upper management). The upside is that the fight for resources was not nearly so vicious. The success of your project could result in me getting a nice check.

  9. Crystallographer says:

    I’ve been on both sides of this as a project leader and more recently as a function leader. Assuming all have good intentions and sharing the rationale for decisions in real time really helps on both sides. The rationale may be difficulty of chemistry or project stage (ie, if the key need is PD/PK or a POC efficacy study, more chemists may not help….) or prioritization (we need a candidate this year). Most people in this business are smart and rational and will support difficult decisions (even if they disagree with them) if they have are included in the process.

  10. Peter Shenkin says:

    To address this problem (i.e, peak demand greatly exceeds average demand), computational guys have the Cloud. Pharma has CROs, which is harder…. We need Uber for med chemists. 😉

    1. Anon says:

      Which will pay like uber? I’ve heard they average $20,000/year.

  11. Kelvin says:

    This is why most pharma companies have a portfolio management function – to prioritize projects and resource allocation by expected return on investment. So if you want more resources on your project, you had better show that those resources deliver a competitive ROI compared with resources allocated to other projects. And expected ROI is quite easy to calculate based on the deliverables promised by adding more resources. The key is to make people accountable for delivering on those promises. So yes, you can have more resources for your project if you deliver XYZ, but you had better deliver XYZ or you lose your bonus, or credibility next time you ask for more resources.

  12. C_B says:

    Today’s xkcd highly relevant to this blog:

    (also, thanks for the AMA the other day, Derek!)

    1. Tim McDaniel says:

      Let me second the recommendation of XKCD in general, but this cartoon in particular.

  13. b says:

    Never mind all the BS above–It would be cool if pharma would hire chemists/biologists to do actual work rather than slave labor. Seams fair given that chemists and biologists outwork people like Ian Read (Pfizer ‘CEO’) 20:1, no??????

    1. Hap says:

      Except at this point it pays management better to hire cheap, temporary labor. They’re not going to be around in ten years to face their competition and their paychecks and options will already be cashed, so they don’t care.

      I assume management works hard – it’s just that they’re paid for failure or success, their incentives are ill-aligned with those of their companies, and the people who hire them don’t seem to care. In theory, if the people that own the company don’t care, then they can blow their money however they like, but I am not sure that the people making the decisions and most of the people that own the stock have much to do with one another.

Comments are closed.