Challenge to solve
Glovo asked Pivotal Labs to help them improve their internal back office tool that deals with the management of their end-customer facing delivery service.
To put it simple, for each day there is a certain number of delivery slots to fill with couriers, just enough to cover the forecasted demand of deliveries, and not too many in order to avoid having couriers without work to do.
To put it simple, for each day there is a certain number of delivery slots to fill with couriers, just enough to cover the forecasted demand of deliveries, and not too many in order to avoid having couriers without work to do.
There was a legacy admin panel in place, some functionalities needdd to be moved over to the new admin panel and also some new desired functionalities to be designed and developed.
Major goals were:
Major goals were:
• Automate courier management process
• Increase operations efficiency
• Increase utilisation rate of couriers
• Reduce courier churn
• Increase operations efficiency
• Increase utilisation rate of couriers
• Reduce courier churn
Time Frame
6 months
Team Setup
Remote
My Role
Design lead, user research, usability testing, workshop facilitation, service blueprint, UI/UX design, icon design
Project KickOff
Understanding the problem space and defining desired project outcomes
I facilitated a full day kickoff in Barcelona where we met the client team and defined together the business and product goals for the engagement. We created personas, mapped out a service blueprint of the current workflows and talked about potentials risks and mitigations.
Exploratory User Research
Let's learn all about the problem!
After the kickoff we stayed a week in Barcelona. The Pivotal engineers needed to get to know the tech stack and setup their environments. This was best done by pairing with Glovo engineers. The Product Manager and I paired on running some interviews with users, both in the Barcelona office and also remotely because Glovo has operations managers in various different countries and we wanted to get a variety of input. Our engineering colleagues helped to take notes.
In these interviews we asked the operations manager to show us their current workflow in order to manage courier slots and we asked them some open-ended questions about what is working well, what could be better and what functionality is missing.
We synthesised our learnings in personas and a service blueprint.
Personas
Meet Lola and Carrie
Together with the team, I created personas for the main users of the system in order to empathise with their needs and pain points.
We started with proto personas based on our team’s assumptions and after each round of user research we incorporated our learnings and updated them.
We focused mainly on the Lola persona as most of her needs would cover Carrie's as well.
We started with proto personas based on our team’s assumptions and after each round of user research we incorporated our learnings and updated them.
We focused mainly on the Lola persona as most of her needs would cover Carrie's as well.
Service Blueprint
The big picture of the product and scope
We mapped out the current workflow for the main personas and all systems and teams involved. It became clear that the operations managers had to deal with a very manual process using spreadsheets and uploading them to the legacy app, informing relevant teams through Slack etc. It became clear that we need to automatise all these error-prone steps.
Exisiting flow before the Pivotal Labs engagement
We then created another version to show the ideal MVP future state of the flow. This was a working document and something we kept updated throughout the project. At the end of the engagement this even served as an onboarding tool for operations managers.
Our planned work for the upcoming months was to transfer some existing functionality from the legacy system to the new admin panel as well as working on a few missing key features:
• Lola wants to copy values for bonuses from a previous week or day
• Lola wants to bulk edit courier slots
• Lola needs to add important events to courier slots
• Lola wants to schedule days for automatic enablement
• Lola needs to see the number of couriers which did not show up for a slots
• Lola wants to copy values for bonuses from a previous week or day
• Lola wants to bulk edit courier slots
• Lola needs to add important events to courier slots
• Lola wants to schedule days for automatic enablement
• Lola needs to see the number of couriers which did not show up for a slots
New flow put in place at the end of the Pivotal Labs engagement.
Design Studio
The power of collaborative sketching
I facilitated a couple of remote sketching exercises with the whole team to generate many possible solutions for specific scenarios.
By encouraging cross-disciplinary roles to think deeply about a problem simultaneously, we discover many more unknowns than if a designer sketched alone. Additionally, when developers are in the process they understand where the product is going and are able to make better decisions. It can save them time later as they're developing the product we create as a team.
By encouraging cross-disciplinary roles to think deeply about a problem simultaneously, we discover many more unknowns than if a designer sketched alone. Additionally, when developers are in the process they understand where the product is going and are able to make better decisions. It can save them time later as they're developing the product we create as a team.
Prototyping
Testing our assumptions in the fastest way possible
After our sketching sessions I was building prototypes with Sketch. I always start with the most important feature first and then build onto it using a THINK-MAKE-CHECK process.
Usually, the first few prototypes are pretty simple, built in order to validate one specific use case. I show it to the users and give them a scenario and a task to perform using the prototype. Then I iterate based on learnings from these usability tests.
Usually, the first few prototypes are pretty simple, built in order to validate one specific use case. I show it to the users and give them a scenario and a task to perform using the prototype. Then I iterate based on learnings from these usability tests.
Usability Testing
Deciding what to keep, kick, change
During the project we did several rounds of testing with prototypes and also with the working software. After each round with 3 to 5 participants in 1:1 interviews we have collected a lot of notes and validated or invalidated assumptions.
I usually did facilitate sessions of affinity mapping and a KEEP-KICK-CHANGE analysis in order to come up with action items for the next iteration.
Design Iterations After User Feedback
Lola confused the coverage indicator with enabled days
One example of an iteration I did was the indicator for the day status. In the first version I used coloured stripes. When we tested this, the 'Lolas' were confusing this system with a similar slot coverage indicator on the courier slots page, a feature that indicates if enough couriers have signed up for a slot.
In the new version, I used dots instead, a legend for the status of the days AND I added the coverage feature as well since this was a helpful new information for Lola on the overview calendar.
Iconography
Lola needs to see what is happening in her city at a quick glance
I needed to work on a set of icons for special events that might be happening during a slot and that help Lola to make decisions on bonuses, compensation and couriers needed.
For some of the categories it was a bit tricky to find good and intuitive symbols, so I decided that I need help and organised a sketching brainstorming with some designers and product managers from the Paris office who were not part of the project. In fact, I thought it was helpful to ask some unbiased colleagues to come up with ideas.
After the sketching session I worked on the vector icon set in Illustrator. In the courier slots page these would be displayed in a rather small size and needed to be simplified. I added a specific colour to each event to help scan them more easily throughout the table.
Interaction Design Examples
Lola wants to schedule days for automatic enablement
Lola wants to copy values for bonuses from a previous week or day
Project Outcomes
A lot of happy Lolas!
The Pivotal team never got access to exact metrics of courier utilisation rate, courier churn etc. but we heard several times from client stakeholders that the results are doing very well and that we have significantly increased operations efficiency.
Throughout the project we always had easy access to users who were more than happy to give us their honest feedback about our work and who were therefore our main performance indicator. Towards the end of the engagement, we did a final round of interviews with the developed product and had lots of positive feedback.
Throughout the project we always had easy access to users who were more than happy to give us their honest feedback about our work and who were therefore our main performance indicator. Towards the end of the engagement, we did a final round of interviews with the developed product and had lots of positive feedback.
My colleagues and me in the Glovo Barcelona headquarter
