YouTube Facebook LinkedIn Google+ Twitter Xinginstagram rss  


By Richard Dweck

Oct. 9, 2002 | Summing up the difficulty of relationships at the end of Annie Hall, Woody Allen jokes about the man who complains to his psychiatrist that his brother thinks he's a chicken. When asked why he doesn't commit his brother, the man replies, "I would, but I need the eggs."

Life science companies face a similar dilemma looking for informatics solutions. The need for innovative informatics in R&D is significant and growing, and there is no shortage of companies offering a rich variety of software and services. But the stability of these "egg" providers is anybody's guess. There is a list of failed companies (DoubleTwist, MAG), companies gutted through acquisition (Genomica, NetGenics), and companies with radically altered business strategies (Viaken, Accelrys).

Even so, the offerings coming out of this sector will have a significant positive impact on life science companies. Hats off to anyone who can predict which bioinformatics business models operating today will be thriving in five years. Because many smart people have lost big money on this question already, it's unreasonable to expect a predictable landscape anytime soon.

Characteristic of an emerging sector, life science managers face the prospect of their informatics investments being stunted by failures, restarts, and acquisitions while the technology becomes smarter and faster. Decision-makers are caught in the crosshairs between uncertainty and lackluster progress if they fail to act.

In-house Vs. Outsourcing 
Determining the benefits of using external resources versus the perceived security of doing everything in-house is key. In-house development does carry a fair amount of risk and fosters an insular mentality that might cause you to miss the boat on the benefits of new innovation. At the end of the Cold War, the Soviet bloc had every industry necessary for self-sufficiency, but few survived.

Informatics investment decisions are complex. As difficult as it is to devise and gain support for a winning informatics strategy, the subsequent tactical decisions may prove just as challenging. In the end, you will likely handle a great deal of the deployment internally, while still relying on an external party for key portions of the solution.

Time is scarce, but fortunately, only some informatics decisions warrant your attention. Delegate the rest. If the investment is high in its strategic and/or operational importance to your organization, then it requires commensurate managerial attention. Some questions to ask include: Could this investment result in a competitive advantage, or could its absence result in a competitive disadvantage? Is it required to store or manage mission-critical data? Are other critical systems going to be tightly tied to it? And finally, is critical downstream work dependent on it?

How to Proceed 
If you do conclude there will be significant operational or strategic dependence, consider the following:

* Be clear about what you are buying and how you will replace it. For example, a package may include software tied to a subscription database that is updated and cross-checked by the vendor. If the vendor disappears, how will you replace both services? Even if you could replace or support the software internally, do you have the means to update the database on your own — or could you outsource that responsibility cost-effectively?

* Take a good look at in-house development muscle. The technical and project management skill levels required to develop a software system are high, particularly if the informatics technology is new to your organization. Internal managers must be realistic about the skills of key individuals and whether the staff assigned might be already stretched too thin. Avoid becoming dangerously dependent on the one key in-house expert whose knowledge of the system is irreplaceable.

* Cut smart deals. Front-end arrangements can reduce potential risks. Negotiate for source code (and appropriate documentation) to be escrowed with a third party in the event that the vendor is no longer able to provide ongoing support. Ensure that your confidential information is fully protected under all possibilities. Require the vendor to support you and address your confidentiality needs even if they become part of — or tightly aligned with — a competitor. You need to involve experienced attorneys to protect your interests before project initiation.

* There is no safe haven. Should you go with a startup company or an industry stalwart like Sun or IBM? Unfortunately, there is no guarantee that the stalwarts will want to serve you in five years. One of Jack Welch's early moves at General Electric was to divest from all businesses where GE was not in the top two. The big guys are investing heavily now, but will they only stick around if the rewards add significantly to their bottom line?

* Know where your data go next and plan for the worst. If the system is storing or managing mission-critical raw data from instruments or other sources, it is crucial to have an exit strategy that includes getting the data out and replacing the functionality of the discontinued system. Realistically, you may only be able to support it if it is a mature system — providing that you have secured use of the source code as mentioned above.

* Imagine owning the data, even if there isn't any. Product maturity makes a big difference if you have to take over a system with tens — if not hundreds — of thousands of lines of code. Mature products have been largely debugged and are potentially supportable in-house. Better still, the likelihood of it being taken over by another competent third party or of finding someone well-trained to maintain it increases.

* Consider the users and how much they cost. Replacing a system may have a minimal impact on the IT staff but a big impact on training costs associated with bringing a new system into general practice. The number of users and complexity of the system is directly proportional to the costs borne by the user community in a transition.

* Plan for the best. In an industry where advances in knowledge, techniques, and technologies require a steady stream of new or enhanced information systems to compete, the useful life of many IT investments may be shorter than in most other industries. If you have solid experience with an informatics vendor, take steps to build on it. Be quick to benefit from their experience with other customers. Be up front with them in your business goals so that they have the resources to provide even greater value to your organization.

As much as the industry endures shifts and volatility, new informatics technologies are still essential for more efficient drug discovery and better drugs. The advantages that can be gained justify enduring the upheaval, and yes, even some boiling and scrambling. Because at the end of the day, we really do need the eggs.* 

Richard Dweck is the founder and CEO of 3rd Millennium, a bioinformatics consultancy in Cambridge, Mass. He can be reached at

For reprints and/or copyright permission, please contact Angela Parsons, 781.972.5467.