What do pharmaceutical buyers want from bioinformatics vendors? According to panelists at a symposium last week, three little words: return on investment.
“Vendors should come with a clear value proposition,” said Shawn Ramer, vice president of R&D informatics at Bristol-Myers Squibb, during the session at the Bio-IT World Conference and Expo, where informatics executives from six biopharmaceutical firms discussed their life science IT strategies. Raymer’s sentiment was echoed by Rainer Fuchs, vice president of research informatics at Biogen, who chaired the panel: “It’s not the coolness factor, but the impact on discovery” that drives purchasing decisions in R&D informatics, he said. “I’ve seen some presentations [from vendors] and been very impressed by the technology, but ended up saying, ‘so what?’”
But quantifying the downstream payoff of information technology investments is easier said than done in big pharma. Even project managers proposing internal IT projects often have difficulty convincing higher-ups of the short- and long-term value of cutting-edge approaches. According to Boston Consulting Group, which presented some preliminary data from an upcoming bioinformatics market research report at a workshop during the conference, most IT investment in biopharmaceutical R&D is still in “process automation” technology — mainstream IT solutions to improve the efficiency, cost, and speed of the research process. Pharma is not yet investing in more innovative bioinformatics approaches “because efficiency is easier to sell from the point of view of the managerial decision,” said Charles-André Brouwers, vice president at BCG.
Despite the promise of bioinformatics to improve R&D productivity, “the value hasn’t been seen by many companies yet, so it makes the job of decision-makers very difficult,” he said. “Senior management in pharmaceutical companies have less and less of an interest in bioinformatics at this time. It’s very difficult to even bring it to the table.”
How to Pitch IT
The solution for both in-house and third-party informatics proposals, according to Brouwer, is to “articulate informatics in terms of the company’s research strategy and economics.” The tricky part is calculating the impact of a technology that may not provide a visible benefit until the tail end of a five-year project cycle in the quarter-by-quarter economic terms that upper management works with.
Luckily, for those struggling with this fuzzy math, BCG provided a template for calculating the projected value of a new technology investment. Using the example of a hypothetical predictive ADME-tox software product, BCG’s Detlev Biniszkiewicz provided a five-step plan for developing an economic model to evaluate the impact of a new technology (see box). The approach is applicable both for vendors seeking to put some meat on the bones of a sales pitch as well as in-house informatics teams assessing new technology options.
In the case of the hypothetical predictive tool, BCG plugged in some conservative industry parameters to estimate that the software could add $35 million extra to the $600 million net value of an average drug project. Biniszkiewicz demonstrated how the model could be tweaked to account for the effectiveness (or ineffectiveness) of the tool as well as the impact of the technology’s application position in the pipeline. Armed with an educated economic assessment based on this approach, the decision to proceed with the technology is then up to the individual needs and technology adoption philosophy of the research organization.
Brouwers broke down adoption philosophies into three categories: “believers,” the classic early adopters who are eager to try new technology in the hope it gives a competitive edge; “followers,” who tend to wait until a technology has hit the mainstream before adopting it; and “experimenters,” who fall in between. Experimenters are usually larger organizations who can bear the economic risk of trying new technologies, but tend to treat them as “toys,” not expecting to see an immediate productivity impact. The biggest challenge these organizations face, according to Brouwers, is effectively turning these toys into production-ready pieces of the technology infrastructure. As long as new technology is perceived as a toy rather than a tool, it won’t get buy-in from upper management.
Most big pharmas tend to follow the experimenter model, Brouwers noted. Yet in some new technology areas, pharma seems to be happy with the wait-and-see approach. For example, BMS’s Ramer noted that he’s keeping a close eye on new developments in semantic web technology, but “so far, I’ve been unable to confront a clear value proposition that can be communicated to how it can impact the top line of R&D.” Michael Braxenthaler, director of research informatics at Roche, said that large-scale grid computing technology won’t see widespread adoption at his company any time soon, because it doesn’t yet provide the “tangible, mid-term return on a current bottleneck” that executives want to see.
Despite the challenges of quantifying IT-driven process improvements, some informatics departments have developed effective strategies for assessing in-house projects. BMS is just winding up a five-year overhaul of its informatics group to ensure that its projects are in line with the company’s R&D roadmap. Part of that overhaul, Raymer said, was establishing criteria to assess the ROI of short-term, mid-term, and long-term projects in order to determine the level of investment in new technology. The next step, he said, is “asking the same questions for internal and external investments.”
This evaluation of third-party technology marked a key differentiator among the executives on the panel. Bernd Langhein, head of IT at Novartis, said his company already treats any vendor proposal in the same way it does an in-house project. On the other hand, David Pioli, director of lead discovery informatics at Aventis, noted that his firm could stand some improvement when it comes to vendor solutions. “I don’t think we’ve learned how to handle the vendor community properly,” he said. “We should treat vendor offerings the same as any other in-house project — evaluate benefit, cost, risk, upside, downside — but we don’t actually do that.”
The burden for now is on the vendor side to “show you’re bringing benefit to somebody, somewhere,” said Robin DeMent, worldwide director of genetics bioinformatics at GlaxoSmithKline. “You have to find something — a prototype, a pilot — that shows real value immediately.”
Additionally, for those vendors seeking tips on what the pharmaceutical industry is looking for, DeMent added, “there’s nothing worse than a vendor revealing that they know nothing about the drug discovery process.”
Further tips, only slightly tongue-in-cheek, came from Pioli: “Don’t talk to my boss’s boss first; and consider giving up the PowerPoint presentation and just telling me the truth.”