Compared to most other industries, life science and biopharmaceutical companies have been slow to embrace digital technologies. Part of the problem is that lab managers are afraid of failure, and that fear is preventing them from realizing the many potential benefits digitization offers. Digital tools can improve process development, streamline tech transfer, reduce errors, and much more.
Early adopters of digital technologies have learned many lessons during the implementation of these tools—and those lessons offer precious advice for new adopters. Here are the biggest pitfalls lab managers should avoid:
1. Relying too heavily on customization
In a recent survey of pharma/biotech companies conducted by Jigsaw Research, one in four respondents reported building software in-house to precisely fit the functionality they needed. This approach is especially common among biopharma development teams, whose needs have historically not been met by commercial, off-the-shelf software. They feel they have no choice but to try to modify systems designed for other purposes.
In reality, software for biopharma has significantly improved in recent years. Cloud solutions provide the agility needed to support collaborative, dynamic projects and this is vital for emerging and translational therapies where the lines between discovery, development, and manufacturing are often blurred.
While developing a custom solution may seem like the best way to meet the needs of your business, it usually isn’t. Custom builds are lengthy, expensive, and resource-intensive, and they pose a real risk that what’s delivered will meet the stated requirements—but not actually solve your business’ challenges.
You will have to rely on potential end users to define requirements for custom software, but it’s often hard for them to visualize how the system will work, and users can easily get caught up in minute details without considering the bigger picture. Then there are other classic build versus buy considerations, such as figuring out the logistics behind training, upgrades, maintenance, and validation.
2. Failing to consider long-term extendibility and sustainability
The extendibility and sustainability of digital solutions is important but often overlooked. For example, do multiple templates need to be created when the same step is required for different processes, or can a common template be used instead? These systems should be extendable to a variety of scenarios, such as the need to exchange data with partner organizations or incorporate new business units into a process.
Another key consideration is the need to define and set parameters. Do multiple protocols differ only by a particular condition, material, or attribute? The answer will govern the design of digital solutions. You can reduce the number of templates that need to be developed and maintained—and strengthen your ability to interrogate your data—by leveraging existing parameters and having a shared vocabulary that defines them.
Finally, custom-built or hybrid systems are frequently brittle and difficult to maintain. They require complex data synchronization and loading processes, which are prone to break as data models evolve or systems go on- and off-line. The niche skill sets needed to perform maintenance can be hard to find.
3. Ignoring opportunities to standardize and implement best practices
The lack of data standards is a known issue in the life sciences that industry consortia such as BioPhorum, Pistoia Alliance, and Allotrope Foundation are working to address. Standardization is so rare that often, for example, groups within the same company that work at different sites will have completely different naming systems and documentation practices—or even distinct ideas of what comprises a “project.”
Let the adoption of a new system be your opportunity to operationalize data best practices. Data governance standards are relatively simple to formulate, but difficult to implement. Build your data governance into the system you implement, using business rules, picklists, or conditions to ensure that data is model-ready from the moment it is collected. This will be far more effective than trying to conduct retrospective data clean-ups, which are reactive rather than proactive.
4. Letting existing systems drive requirements for new technology
When it comes time to replace an existing system, familiarity with legacy systems can forestall innovation and the potential to get the most out of a new investment.
Let use as an example an existing biopharma lab retiring its early 2000s ELN in favor of a modern lifecycle-based data management platform. The new platform meets the biopharma lab’s strategic objectives around data science and future use cases for data. Approached for requirements, users describe the functionality of the legacy platform rather than the business need. After all, it “does what they do.”
Will the new system provide maximal benefit when its use is modeled on a legacy system? Of course not.
5. Losing sight of the big picture: solving business problems
Like the Aesop’s fable of the man, the boy, and the donkey, try to please everyone and you will please no one. A common objection to some digital solutions is “this will take me more time than it does now with pen and paper.” If you look at just that single point in time then yes, the digital approach may even be slower. But if you consider how long it would take to search for every instance when a given parameter went above a threshold, across sites, and/or departments, in a mainly paper-based versus digital system, the picture changes dramatically.
Additionally, the “pen and paper” days do not address a modern approach to data. Data is a biopharma lab’s number one asset, only after its people. Data is acquired now not just to solve the question of the moment, but to formulate questions that haven’t been posed yet. Efficiencies in data capture are increasing, and the volume, velocity, veracity, variety, and value of data are now the focus.
The challenge lies in determining where compromises can be made, and which requirements are truly non-negotiable. There are no hard and fast rules for this but keeping in mind the overarching business objectives both now and in the future is a good place to start. Where possible, work with a trusted vendor to see if any key requirements that aren’t addressed by current functionality can be incorporated into future releases. Digitization and modernization vendors with a focus on biopharma lifecycle management work closely with organizations building future capabilities into their product developments. Chances are, if it’s critical to your business it will be critical to someone else’s business as well, so it’s a win-win for all involved.