Case Study #1

Role

Director, Service & Product

 

Industry

Biotech, Public Health, Pharma

Organization

... leveraging scope creep to my advantage in complex, long-term projects.

I led a team of 4 to transform the supply chain of a small and scrappy startup into a more mature biotech organization. We implemented multiple solutions:

  • 3 new services that provide structure and protocols for routine tasks to reduce pressure on the supply chain.

  • 2 digital solutions—a new data entry software and dashboard software that helped facilitate better decision making, provided greater data reliability, and reduced friction around data entry.

  • 1 mindset shift towards a mature biotech mentality.

Key Metric

Streamlined a 20-hour task into 30 minutes

Impact

Enabled the organization’s pursuit of FDA approval by increasing reliability and decreasing friction.

Summary

OpenBiome, Finch Therapeutics

Duration

3.5+ years

Background

OpenBiome is a nonprofit biotechnology organization that manufactures a naturally-derived drug for microbiome-related illnesses.

Screen Shot 2022-01-25 at 12.15.17 PM.png
  1. My Team — I built the Product Operations team over the course of this project. Starting out as a “special projects” task force, I hired a data analyst and two operations associates that became project managers.

  2. Platforms Team for IT Product Implementation — For the digital solutions, the team included an internal IT manager, a quality systems director, a system administrator, and an external team of developers (based in Germany).

  3. Clinical Research Team — For the process improvements, team included a representative from 3 different functions.

 
Brief
 

I gave myself an assignment.

To be honest, this whole project started for selfish reasons: I was in charge of an inventory accounting report that took 15-20 hours to do each month. I loved working with our data, but hated this report because our data was so inaccurate and fixing everything manually was extremely tedious and inefficient. So, I gave myself an assignment: Find out why our data was so unreliable.

Methods
 

​I started with an analytical approach to understanding what was happening and then, sprinkled in a little qualitative understanding by interviewing team members that entered the data.

  1. Scripted different reports to slice the data different ways in hopes of identifying data error patterns; I used SQL and python to do this analysis.

  2. Brainstormed possible ways those trends could be happening; spoke to my teammates about my theories and asked for theirs.

  3. Compared brainstormed ideas with real life by observing processes in action and having conversations with the lab technicians and other data-adjacent teammates.

sql-mode-vsql.png

Observations & Interviews

Screen Shot 2022-01-25 at 12.15.52 PM.png

This project had so many people and systems involved that our team spent almost 4 months on the first big data gathering push. We took this time to watch 3-hour long processes from prep to clean up, ask why at almost every step, and observe unspoken team and workflow dynamics.

We talked to laboratory technicians, stakeholders, and experts in biotech-specific information technology systems, quality assurance and control, and regulated supply chain environments.

 

We engaged through observations and interviews, secondary research & expert information interviews as well as facilitated workshops, displacements and co-design via rapid prototyping. Together, we created iterative maps of our understanding of the organization.

Screen Shot 2022-01-25 at 12.15.45 PM.png
Screen Shot 2022-01-25 at 12.17.59 PM.png
769766CE-BE77-417C-BD0A-F893773DC314_1_105_c.jpeg

Secondary Research & Expert Interviews

In parallel, we completed extensive desk research that included expert interviews, sales calls, and internet research on best practices. To make sure we understood as much as we could to best facilitate appropriate solutions, during weekly syncs each member of the team would bring up something new they had observed and how they were going to research that new curiosity.

Facilitated workshops & Co-Design Activities

Co-design workshops were critical to success. In such a fast-paced organization, these workshops were dual purpose: maintain user and stakeholder interest and buy in as well as hear directly from teammates what they were thinking about their processes, frustrations, and goals. Throughout the project, we continued to use similar full-team sessions to review, troubleshoot, and plan for new features.

Screen Shot 2022-01-25 at 12.18.06 PM.png
Screen Shot 2022-01-25 at 12.18.13 PM.png
Screen Shot 2022-01-25 at 12.18.24 PM.png

We mapped everything as we listened, asked questions, and observed.

We made maps of process flows, information flows, data flows, physical movement through space within buildings and across the city of Boston, interactions between stakeholders, and interconnectivity of goals and workstreams.

We took especially careful attention when mapping anything cross-functional details to ensure extra flexibility and usability at those junctures.

Screen Shot 2022-01-25 at 12.20.06 PM.png

Early process flow maps

Screen Shot 2022-01-25 at 3.11.03 PM.png
IMG_3095.jpg
Screen Shot 2022-01-25 at 3.11.17 PM.png
1841F9F8-416D-489B-A323-DADF5B5E5E9B_1_105_c.jpeg

Process Flow Diagrams for Software Developers

Screen Shot 2022-01-25 at 3.12.57 PM.png
Screen Shot 2022-01-12 at 7.45.41 PM.png
Screen Shot 2022-01-25 at 3.12.10 PM.png

We learned a lot throughout this process and were able to build along the way! 

Insights & Theory of Change

1. Big Picture Insight - Growing Pains

OpenBiome was experiencing massive growing pains with service demands. Since beginning its public health mission, the organization went from helping ~50 patients per month to 600+ per month—that’s 1,200% more patients! However, all of the systems managing the organization's supply chain were being held together by employees’ grit—very little of the processes and systems had evolved with that growth.

Additionally, all of the stakeholders I spoke with mentioned the importance of scaling and being more efficient but were struggling to do so.

2. Functional Insight - Tediously and unnecessarily complicated

Each of the silos in the supply chain had their own problems that needed to be fixed to work within and across functions.

  • Conflicting protocols — e.g. there were 4 different drug products being produced with different requirements.

  • Clinical trials were chaotic — e.g. clinical trials were uncoordinated and tended to be bespoke, with custom labeling

3. Technology Insight - Under Resourced

Current digital solutions were hurting more than helping — e.g. there were time outs built into a home-brewed manufacturing system that caused technicians to mis-enter data; additionally, the developer controlling the system couldn’t provide full time support.

4. Cultural Insight - "Coming of Age" Tension

There was an itch for cultural growth as well as resistance to maturity. As the presence of OpenBiome grew, more attention towards their good work was encouraging more and more regulation which seemed to be at odds with the general attitude of the scrappy nonprofit.

5. Business Insight - Increasing Demand

OpenBiome’s for-profit sister company, Finch Therapeutics, needed more from OpenBiome for its own research projects: faster service to receive drug samples and more accurate sample data.

Theory of Change

 
Solution(s)
 

Research and implementation happened iteratively to make incremental progress and provide short-term solutions while the longer term solutions were developed.

In the end, my team and I had implemented multiple solutions:

  • 3 new services that provide structure and protocols for routine tasks to reduce pressure on the supply chain

  • 2 digital solutions—a new data entry software and dashboard software that helped facilitate better decision making, provided greater data reliability, and reduced friction around data entry.

  • 1 mindset shift towards a mature biotech mentality

3 services that support data reliability and FDA-readiness by creating clear protocols

1. Created Clinical Research Pipeline

To tackle the chaotic clinical trial problems we found (Insight #2 above), one of my associates and I led a co-designing effort with the clinical research and shipping teams to create a more consistent process for requesting bespoke orders. The hardest part of this process was managing a "living prototype," which acted as both a tool for building as well as the interim solution.

This service was maintained by the clinical research team.

2. Improved Donor Operations & Management Service

Most of the donor-related processes were done on paper. While we translated these paper processes into a digital format, we worked with the donor team to simplify the more complex nodes. The end results were repeatable processes that were easier to train new teammates on and more predictable logic to use when planning for poop supply needs.

This service was maintained by the donor operations team.

3. Data Dashboard & Reporting Service

We created processes around requesting data, reports, and dashboards as well as instituted a "State of the Poonion" meeting to increase awareness and understanding of our product data. This service accompanied the implementation of a dashboard platform (see below).

This service was maintained by my team, the product operations team.

2 products to facilitate better decision making, increase reliability, and reduce the friction of data entry.

1. Integrated Electronic Medical Record & Inventory Management System 

Capture.PNG

We translated almost the whole supply chain into an electronic medical record and inventory management system. This system was a configurable off-the-shelf solution chosen to support the needs we had come to understand from user and expert research. This FDA-compliant system created a central data hub and set a tone for how teams worked across the supply chain -- together and integrated.

2. Dashboard Platform

data dash.PNG

We created over 20 dashboards to support the daily needs of different teams. The lab technicians could view their daily stats and the executives could observe longer-term trends. The dashboard platform was a critical pressure release on the whole supply chain because it allowed for more nuanced understanding and expectations of the product data. We created "tiers" of data where the data request could be properly triaged and responded to--we leveraged the "live" nature of the dashboards to provide lower risk data more quickly and differentiated more serious reports outside of the platform. 

1 mindset shift towards a mature biotech mentality

While hard to see in the trenches of this 3.5+ year project, the organization had truly evolved in the end. ​ All of these elements worked together to transfer energy from reactively solving problems to proactively building for the future.

 

The organization had matured throughout this project by increasing structure, defining clear paths of action for different decision points, and making a habit of purposefully integrating new processes and initiatives with other teams.

Implementation
  1. Validation – Ran over 150 test scripts for the first phase in coordination with the Quality and IT teams

  2. Training – Designed week-long user-focused training sessions and additional follow up sessions

  3. Troubleshooting – Remained on-call for any system errors

  4. Standardizing a process for improvements – Built on the iterative design cycle to create a protocol for new feature requests and updates

 
 
Reflections

Everywhere along the way, the project had an opportunity to expand and encompass even more of the organization. While I've learned that sometimes you just have to say no, leading through this project with a centralized and consensus-informed approach allowed me to use the ever creeping scope as a way to reframe our problems.