Amazon: Innovation Sprints
PROJECT OVERVIEW
Each quarter, design teams within Last Mile XD pump the brakes on delivery work and explore what is coming in the next 1-3 years for the business — by looking around corners and diving-deep addressing unmet customer needs. We call these innovation sprints.
How might we ... demonstrate human-centered product development to a nascent org?
Each quarter, design teams within Last Mile XD pump the brakes on delivery work and explore what is coming in the next 1-3 years for the business — by looking around corners and diving-deep addressing unmet customer needs. We call these innovation sprints.
How might we ... demonstrate human-centered product development to a nascent org?
MY IMPACT
- Data-driven insights and validated prototypes as inputs for 2023 OP1/OP2 planning — Confidential
- I produced the first design-lead PRFAQ inside of Last Mile
- I presented outputs to Beryl Tomay (VP of Last Mile), changing the mental model that design is a strategic input not a pixel slayer.
Why Innovation Sprints?
Human-centered product development is all about customer obsession — understanding human needs from the very beginning and placing the outcomes customers want at the center of your process. There are two organizing principles across Last Mile Design: 1durable customer outcomes & 2innovation horizons. This allows design to establish itself as a strategic partner and not a reactive service.
WHY IS STRATEGIC PARTNER IMPORTANT?
The real value of design is not at the pixel but envisioning what is coming next for customers — ultimately de-risking areas where solutions are not yet defined, which propels business growth.
Most organization do not know this secret and perceive design as making things look nice and intuitive to use. To combat this perception, each quarter we conduct innovation sprints across Horizons:
- Horizon 1: we execute and deliver experiences for roadmapped projects. This feeds the development team for implementation and release.
-
Horizon 2: we explore where the next right solutions might be. This feeds future roadmap or Horizon1.
- Horizon 3: we discover where the next problems we should be solving could be. This feeds 3-Year Plans and Horizon2.
7 teams including (62 designers and 33 business partners) over the course of a month, identified the most promising problem areas to solve for their customer.
EXAMPLES OF PROBLEM AREAS:
EXAMPLES OF HCD METHODS:
The Opportunity
Like any good designer I had to questions the brief. Forging Voice of the Customer, delivery data, and 3YP goals I challenged the opportunity so that our ideas could lead to stronger focus with bar raising outputs.
We believe that ... qualitative data can help us better understand “the why” behind “the what” of the competitive benchmarking stating poor driver experience at Apartment Lockers.
Success Measures
- 100% participation from Product Management (min 1 hr)
- x% ideas in OP1 (quick wins)
- x% ideas for Rabbit 2.0 release
What I Did
DISCOVERY ︎ DEFINITION ︎ DESIGN ︎ PRESENT/INTEGRATE
DISCOVERY
Sprint Agenda:
Sprints core value is to quickly understand needs, explore a wide-range of ideas and test ideas with stakeholders and end-customers.
![]()
Our team’s focus for the sprint:
We narrowed in on the problem and customers with the highest impact and had the easiest access.
![]()
![]()
Sprint Agenda:
Sprints core value is to quickly understand needs, explore a wide-range of ideas and test ideas with stakeholders and end-customers.
Our team’s focus for the sprint:
We narrowed in on the problem and customers with the highest impact and had the easiest access.
Field Research:
This brought our hypothesis to life, squashed some of our assumptions, and gave us fuel to generate ideas.
![]()
![]()
![]()
![]()
![]()
![]()
![]()
![]()
![]()
![]()
This brought our hypothesis to life, squashed some of our assumptions, and gave us fuel to generate ideas.
Research Debrief:
After our ridealongs with customers we debriefed our learnings across a journey map.
After our ridealongs with customers we debriefed our learnings across a journey map.
DEFINE
We then identified a dozen opportunities and met with business partners to validate their viability.
![]()
![]()
We then identified a dozen opportunities and met with business partners to validate their viability.
DESIGN
Our team explored 100’s of ideas based on the opporunities we aligned with our business partners using a number of HCD methods.
Opportunity 1: Real-Time Locker Capacity
Opportunity 2: Scale Tribal Knowledge
Opportunity 3: Ring for Apartments
PRESENT/INTEGRATE
The real impact is in changing the way our business partners think about design. Below you’ll see dozens of presentation boards presented to our Vice President of Last Mile.
Final Output:
Ring for Apartments
Challenges Encountered
Hybrid of In-Person & Virtual Sprint Attendance
We had 5 teammates in-person and 1 virtual so making sure that collaboration scaled across mulitple formats was key.
We had 5 teammates in-person and 1 virtual so making sure that collaboration scaled across mulitple formats was key.
Knowing How to Carry the PRFAQ Forward
Since I wasn’t the “owner” of the Ring product, it was difficult to know how to best carry the PRFAQ forward. Ultimately, what was most successful was asking people “who should I speak with”? Which landed me with the Product Manager of an analagous product who was able to leverage our insights.
Since I wasn’t the “owner” of the Ring product, it was difficult to know how to best carry the PRFAQ forward. Ultimately, what was most successful was asking people “who should I speak with”? Which landed me with the Product Manager of an analagous product who was able to leverage our insights.
Being New to Amazon But Leading a Strategic Project
I didn’t have all the context for the problem space, nor did I know how Amazon “worked”, it was challenging to lead by example. But what I discovered was that knowing something at 50% produced better results than doing nothing at all.
I didn’t have all the context for the problem space, nor did I know how Amazon “worked”, it was challenging to lead by example. But what I discovered was that knowing something at 50% produced better results than doing nothing at all.