Designing optimal interaction flows for container usage
Case Study: Using design thinking to suggest efficiency improvements in Ginkgo’s container inventory consumption
Skills: Literature Review, User Flow Mapping, Workshopping, Interaction Design, UX Strategy
Summary
Background
Scientists at Ginkgo use containers as vessels to conduct their experiments in. However, Ginkgo had over 204 different types of plates in its inventory, causing difficulties for both scientists and the procurement team. Ginkgo needed a better way of ordering and using its plate consumables to reduce their costs and make it easier for users
Challenge
Aggregate findings from past research efforts and leverage the expertise from subject matter experts to ideate and design a new streamlined process that would alleviate pain points and reduce time and cost spent in relation to the container consumables.
Solution
I designed a process flow that increased relies on more automated forms of communication, a single source of truth, and providing resources to lab scientists so that they can self-troubleshoot and more easily know what is in stock and where
A screenshot of the final service blueprint flow
I handed it off to the procurement team and it was well received h SEQ scores
“Great job on your report, lot of effort and time went into this initiative. Very well done, we love your passion. Thanks for the very well structured report. There are so many ideas in this report for this team to consider. Thanks for also taking the time to walk thru the findings and recommendations.” — VP of Procurement
Background
Scientists perform their experiments in vessels called containers (or plates)
Containers all have the same length and width dimensions, but can vary in material, well number, well shape, coatings, and manufacturer.
The type of experiment a user is performing determines the type of container they would want. For example, growing cells works best in a tall container with a round well bottom whereas containers to be put on a plate data reader instrument need to have a flat and clear bottom.
An image of a container with 96 wells and a round bottom
Scientists use different lab automation to speed up or enable their work
Lab automation can read concentration values, can image cells, can heat up and cool down their plates, and can increase the throughput of transferring and mixing samples, and much more
Each container needs to be carefully calibrated for each piece of lab automation to ensure accuracy
Because the dimensions within a container (like well width and height, distance between wells) vary between different container types, the Automation team must carefully calibrate each one
For liquid handlers this ensure that the pipette tips pipette into the center of the well (or exactly where the user wants it to). This makes transfers more accurate and avoids instrument errors
It takes about 3 hours to onboard a new container type
Ginkgo consumes a lot of containers in its day to day operations
Both users and the procurement team place orders for containers
Based on a SQL query of the databases:
There are 204 distinct container types onboarded into our software
Over 32,000 containers were used in May 2024
In order to decrease overall spend and increase our runway, Ginkgo needed to look at options to optimize their consumable spend
Maintaining a diverse inventory was taking up a lot of warehouse and shelving space and effort from the procurement team
Scientists were wasting time deciding on and looking for containers on the shelves, and wasting money ordering one-off and potentially duplicate container types with minimal supervision
Each time a new container type gets added to Ginkgo, the automation team spend hours onboarding each one into their system
There were previous disparate research efforts around container decision making, usage and consumption
2 different people and a hired UX consultancy conducted research around container usage at Ginkgo within the past 3 years from different perspectives
None of the material was aggregated
Much of the material about the status quo and pain points was in the reports, but no action had been taken on any of the recommendations
Design Challenge
Given previous research and subject matter expert knowledge, how can we better design a process for ordering, deciding, finding and using plates at Ginkgo that reduces time and money spent?
Actions
Literature Review
Aggregate findings and pain points from 3 past research projects
Diagramming the Current Flow
Take findings and pain points and create diagrams to illustrate how interdependent the steps and pain points are
Ideation Workshop
Work with SMEs to ideate ideas to address pain points
I joined a working group aimed at optimizing consumable spend
The working group consisted of members from the Automation team, the Procurement team, the Environmental Health and Safety, Continuous Improvement, and lab scientists.
We had weekly checkin meetings to update everyone what we were working on and to ask any questions
I did a literature review of the artifacts from the past research efforts to find explanations of the process and pain points
I combed through all of the murals the UX consultancy made, the confluence pages researcher 1 made, and the murals, dovetail and final reports researcher 2 made
I added all of the materials into a mural, and grouped findings into various buckets to reveal the container usage life cycle and the interdependent stages
A screenshot of the mural I used to aggregate findings from the 3 research projects
I aggregated all of the findings to create a flow diagram of the full process of interacting with plates
I used the buckets I made from synthesis to understand the distinct stages of the lifecycle
At the time many of the stages were interdependent, so it was tricky trying to create a clean process flow diagram with minimal overlaps. It took a few tries to get it right!
I followed up with people from automation, procurement, and stocking to fill in my knowledge gaps
Prioritization Workshop
Work with SMEs to prioritize the ideas from workshop 1
Diagramming the New Flow
Take workshop outputs and create a new recommended flow
An early version of the process flow diagram.
Final process flow diagram to share out with the working group
Final process flow diagram to with related pain points associated to their steps
I also aggregated all of the pain points and created a map to show how many are interdependent
While going through the literature review, I noticed that many of the pain points caused other pain points in other locations in the system
I wanted to draw attention to this in the working group, so I made a separate diagram showing all of the interdependencies
Many pain points stemmed from communication issues, high mental load on users for deciding on and finding the containers they use, and a lot of manual inventory maintenance processes
A screenshot of my pain point diagram
I broke down my findings and diagrams a ‘follow along’ google presentation to make it easier for the working group to follow
I shared the presentation with the working group in slack and in our weekly check in meeting to get everyone on the same page before we started the workshops
A screenshot of my slides in my walk through presentation
After sharing out this background work with the working group, I planned a two part workshop series
The goals of the workshop were to leverage the expertise of subject matter experts in plate consumables from all angles to generate ideas, create connections between, and prioritize them based on the identified pain points
Because the scope of this work was large and many of the members of the working group were leaders in their area, it was difficult to schedule large swaths of time. Rather than a half day session, I decided to schedule two 1.5 hour sessions. One to ideate, and one to prioritize
Separating the workshop into two allowed me to synthesize the ideas between sessions to make it easier for the working group to internalize and prioritize
The first workshop had the working group ideate as many ideas as possible in response to the identified pain points in each part of the flow diagram
I organized the workshop to be a quick, rotational activity with each main section of the process flow be it’s own section for ideation
Many of the participants had not done a workshop like this before, so I had to add additional prompts and instructions to get them more comfortable
I reviewed the pain points with the participants, and also had them available in each of the sections they would be ideating in
A screenshot of the workshop setup with some ideas from workshop 1
After the workshop, I themed the ideas into buckets in preparation for the second workshop
I had the participants try to theme the ideas in breakout rooms, but since many were new at it, they didn’t finish in time in the time I had allotted. The ones they did get a chance to theme were great!
I themed the rest of the ideas and shared them out in the weekly meeting
A screenshot of the themed ideas within their section
The second workshop had the working group prioritize their ideas. I asked them to create their own idealized flows
I used the synthesized ideas from workshop 1 as a ‘word bank’ for them to pull from. I asked them to pick no more than 5 ideas from each colored box so that they would need to prioritize which ideas they chose
I told participants to grab ideas from workshop 1 and create their idealized workflows. “Imagine that you can wave a magic wand and have your ideal process enacted tomorrow. What process do you think Ginkgo should follow to keep our plate consumables easily used and maintainable?”
Additionally, I asked participants to jot down what goal they would like to achieve through their flow.
Users shared out their flows to the whole group
Many of the participants came up with similar flows, sharing similar priorities like simplifying our massive inventory and providing scientists with tools to better find containers that meet their needs
A screenshot of two of the participant flows
Given the information from the literature review and the two workshops, I ideated an ideal process for container usage at Ginkgo, as well as recommended transition activities to achieve that process
I used the Nielsen Norman Service Blueprint as a blueprint for my final diagrams
I worked closely with the Automation strategist to iterate and prepare the diagrams for shareout
I sent the final diagram out for comments after the first iteration
Initial draft of the idealized process flow with comments
Second draft of idealized process flow in Service Blueprint formatting
Solutions and Results
I created a final ideal process and laid out the transition activities as workstreams
In addition to creating an ideal process, I also listed steps I recommended the Procurement team take to create the support processes that would be used as the foundation for the new usage flow
This new flow relies on more automated forms of communication, a single source of truth, and providing resources to lab scientists so that they can self-troubleshoot and more easily know what is in stock and where
A screenshot of my finalized workstreams to create the support processes for my idealized flow
A screenshot of my finalized process flow in Service Blueprint format
I handed it off to the procurement team for the to execute
They liked my ideas and incorporated them into their roadmap and plan
“Great job on your report, lot of effort and time went into this initiative. Very well done, we love your passion. Thanks for the very well structured report. There are so many ideas in this report for this team to consider. Thanks for also taking the time to walk thru the findings and recommendations.”
“Yes excellent work @Hanna
to research these complex, highly interdependent and crossfunctional processes and create these logical and easy to understand stand workstreams and recommendations!”