It’s an arbitrary measurement, but the ‘first 100 days’ has become a benchmark for judging the success of new initiatives, such as a data platform. Originally used to gauge the success of new U.S. Presidents, the first 100 days are now widely used as a yardstick in businesses to measure upfront impact. While achieving true data competency takes time, companies should be able to see an immediate impact and reap benefits within the first 100 days of investment.
Delivering tangible success early is essential. It allows businesses to begin realizing a return that justifies their investment. A well-thought-through and executed data strategy is transformative across the entire organization. The sooner this becomes evident, the better.
It also cements senior leadership and business-wide buy-in. For a data project to deliver to its full potential, everyone needs to be on the same page when it comes to goals and what is needed to achieve them. Early success that reveals what the data project is delivering helps to get everyone aligned throughout the project.
Finally, demonstrating impact within 100 days is the hallmark of a successful data approach that understands the business, accounts for potential sources of friction, and takes a strategic approach to achieving business goals. If the project isn’t starting to deliver within a relatively short time frame, it could point to a problem with the approach or with the level of buy-in achieved.
In this blog, Ross Katz, Principal and Data Science Lead at CorrDyn, talks specifically about how we achieve this in the biotech sector, an industry where each organization has a bespoke set of requirements.
To address each of its client’s unique needs, CorrDyn uses tried-and-tested core building blocks that are tailored to suit each business. We look at how biotech firms can approach data competency in a strategic way that sets an organization up to achieve results.
The starting point for any data project is assessment and discovery, which are the foundation of any data strategy. Biotech CMOs (Contract Manufacturing Organizations) often have bespoke equipment that generates data in a particular way and involves unique downstream processes to generate the desired results. This makes a tailored approach essential. We look at documentation for the existing systems and explore the data they produce in order to understand their structure, how they are captured, and the manufacturing, scientific, or engineering phenomena they represent.
We also interview various stakeholders to understand their priorities better and establish what success means for the individual organization. As with any customized approach, the discovery phase will vary based on the individual needs of an organization, but broadly speaking, we aim to answer the following:
This is fundamental to defining the right data approach for organizations. By understanding what a company wants to achieve, we can ensure the data strategy delivers. Drug manufacturers may need to improve data collection for FDA compliance; a biotech CMO may wish to reduce machine downtime and improve yields; or there may be a need for wet-lab automation to boost efficiency. By understanding the long-term business goals and the short-term functionality aims, we can tailor a data capabilities roadmap for each organization.
A key part of the assessment process is understanding the current structure, format, and availability of the data and what needs to be done for your organization to get more value from it. This involves understanding how much data already sits in the cloud and how much on-premise equipment is needed to capture data and push it to the cloud for processing. In many biotech firms, simply migrating data from equipment to the cloud has a significant impact as it opens the door to better machine visibility and more responsive maintenance. Understanding the status quo is an integral part of the initial project that helps establish the long-term data roadmap.
It may be budget, security, compliance, latency, or a range of other areas, but every business looking to implement a data strategy has constraints. We look at what they are and how they interact with the priorities to determine what makes the most sense within the unique context of the company we are working with. For example, if the business manufactures drugs where cold chain management is critical, latency is a vital consideration and needs to be a priority.
Once we understand where an organization is today, what its goals and priorities are, and the constraints that it faces, we are well-prepared to move to stage two.
Following the initial assessment, the resulting information and insights are crafted into a Data Infrastructure and Capabilities Roadmap, a customized proposal that addresses all the dimensions uncovered in discovery.
It presents the findings from the initial assessment and recommends a phased solution that delivers value, starting with the lowest-hanging fruit and building toward long-term strategic priorities. It also weighs alternative approaches and explains why the suggested approach works best to address a client’s target business capabilities, goals, and constraints. It is a deep dive into an organization's data that establishes key insights and gets everybody on the same page about what we're aiming to accomplish.
CorrDyn recently developed a machine data analytics solution hosted in the cloud for a biotech CMO. The road mapping phase involved interviews with stakeholders across the manufacturing organization in order to understand:
The assessment was completed in 6 weeks, and the solution was deployed to production within 100 business days.
The first 100 days are about setting up for long-term success, and by taking this approach, which looks comprehensively into the business and its data, we can deliver this.
Once stakeholders and CorrDyn are aligned on the Data Infrastructure and Capabilities Roadmap, we start implementing foundational elements, which are usually part of the first phase of the roadmap. First, we address on-premise requirements to send the data to the cloud. On-premise data services aim to stream or batch-process critical datasets that can lead to immediate, valuable insights and reporting requirements. At the same time, for this to be a practical approach, we look at minimizing costs and maintenance burden so that the new data infrastructure generates value, not workload.
At this stage, we take care of the practical considerations and constraints identified as part of the discovery process and addressed in the roadmap. This may be replicating existing processes to minimize disruption. For equipment subject to Good Manufacturing Practice (GMP), we define a pre-planned approach for handling data to ensure consistent, compliant production processes. Security is always a priority, so we ensure best practices in all our data systems in line with our customer’s requirements (i.e., encrypted data in transit and at rest, private networks and other network security precautions, least privilege access, data de-identification, etc.). For biotech companies, specifically CMOs with audit requirements, we can ensure that the data is saved, handled, and stored in a way that meets local and global regulatory requirements.
Whatever the tailored requirements are, the comprehensive nature of the roadmap means that when it comes to establishing the architecture needed to achieve the client's overall data goals, we can ensure a smooth deployment and maintain business as usual.
As the architecture takes shape, we start to get ready for Stage 4 by putting in place cloud data pipelines that enable warehousing, ad hoc analysis, near-real-time reporting, and machine learning ready to support the immediate analytics opportunities that have been identified in the roadmap. This is essential to making existing data ready to start working for the business and is an ongoing process as we continue to build out data pipelines to support longer-term goals.
During the process, data is partitioned and optimized. It is then organized in the correct state for analysis in the ways that are most valuable for your organization. The data is prepared to be used for the immediate purposes defined in phase one of the roadmap.
As mentioned, in practice, we begin working on Stage 4 while still establishing the baseline architecture. This allows businesses to see and feel the impact of the project and generate ROI within a short timeframe.
When data is successfully warehoused, modeled, and organized, the data strategy becomes fully customizable, depending on business requirements, and starts to drive results. The instant impact opportunities ultimately depend on the application and the insights the organization wants to realize, but let’s look at some examples.
We can process data in batches every 1-24 hours or near-real-time if latency is a concern. (Realtime is achievable but not without substantial costs, so we usually find that near-real-time meets the needs of most use cases.) As an illustration, we worked with a Director of Engineering for a biotech manufacturer to implement a flexible IoT data pipeline that could process thousands of events per second at a 40% lower cost than their existing solution, with even better economies of scale as the solution evolves. This met their specific requirements for near-real-time insights and proved extremely valuable in allowing them to improve product quality and reduce machine downtime.
Another example, again looking at manufacturing equipment, is developing a methodology for Overall Equipment Effectiveness (OEE). By giving our clients 360-degree visibility into sources of waste using machine data, they receive insights that can guide more efficient and effective manufacturing processes. This information helps organizations minimize avoidable costs and machine errors, driving tangible value for the business.
Depending on its particular goals, the immediate impact will look different for each biotech organization. But by starting a project with in-depth discovery, CorrDyn can make sure that the quick wins align with strategic goals and deliver for the organization as soon as possible. The medium and long-term benefits, where an organization reaps the rewards of increased automation, machine learning, and integration between data systems and manufacturing systems, all come in time. However, seeing upfront ROI is essential to generate business-wide buy-in, which is crucial to support the data strategy over the long term.
We use multiple metrics to determine the success of a data campaign; cost reduction, efficiency gains, business insights, etc. Looking at the impact in the first 100 days acts as a ‘temperature check’ that allows us to make necessary corrections, and ensure the project is on track to meet the desired business goals.
As we outlined in the introduction, an effective strategy should demonstrate ROI within 100 days to gain stakeholder support and confirm it is working. If a data campaign is not achieving this within the 100 days window, adjustments need to be made. This not only ensures short-term goals are met but also establishes a solid foundation for long-term data ambitions.
The CorrDyn approach aims to fully understand the objectives and challenges unique to your organization in a sector where we have real-world experience. Starting with a deep understanding of the business is at the heart of what allows us to begin to provide tangible results, with campaigns that are designed to hit the ground running from the outset.
If you're interested in learning more about how CorrDyn can assist in defining a data strategy that generates immediate impact, click here.