This is the first of three posts in a blog series associated to Predictum’s on-demand webinar, Reactivating Your Dormant Data to Extend its Business Value.
Experimentation leaves vast amounts of dormant, single-use data in its wake, so why aren’t companies doing more to explore the potential of that data and put it back to work?
Several of our clients have accumulated literally hundreds of thousands, if not millions, of single-use data files and have taken action towards their reuse. Given the significant initial investment of money and resources in experimentation, other companies should place higher value on their dormant data. Why miss out on the insights and opportunities that could be shared readily from bringing dormant data back into the experimental equation?
Let’s consider the following scenarios related to experiments and the data that they generate:
- Engineers and scientists may start their experiments with less evidence because comparable, historical data is inaccessible. It’s a duplication of effort when they have to generate data that already exists, only to get the same insights.
- Unstructured data that is collected manually, for example, in single-use spreadsheets, and stored by individuals, which prevents it from being sharing in a meaningful way with their peers across the enterprise.
Many companies have adopted design of experiments (DOE) to improve productivity and efficiency in their experimentation. Designed experiments are patterns of factors arranged in a group and are run in a random sequence. The pattern has properties that enable the maximum learning to be gained from conducting a minimum number of experiments. Methodical experimentation is made easier with the use of JMP® Software, which enables engineers and scientists to design, execute, and analyze the most advanced and accessible experiments.
DOE improves experimental productivity substantially, but why stop there? And for those companies that have not yet adopted DOE, why not at least take other steps to improve productivity, at least what you can do about it now? The next initiative for companies, whether they have adopted DOE or not, should be to reactivate the data from past experiments and maintain the active status of each observation as long as it remains relevant.
Consider doing an initial assessment of your data practices to get a sense of the dormant data in your organization:
- What dormant data have you accumulated?
- What are the sources of your dormant data, such as shared drives, Electronic lab notebooks (ELNs), Laboratory Information Management Systems (LIMS), Microsoft Access databases, Microsoft Excel files, PDF files?
- What is the ratio of recency vs. relevance of your dormant data?
Wayne J. Levin is President and CEO of Predictum Inc. Connect with Wayne on Twitter and LinkedIn. Also, connect with Predictum Inc. on Twitter and LinkedIn.
For more insights in this blog series, catch our second post, Behind the Scenes of High-Volume Data Extraction, and our third post, The Recipe for Success in Research and Experimentation.