Need confidence they will receive the groceries they intended to buy.
Quality is a promise the whole marketplace has to keep.
Instacart fulfills online grocery orders for same-day delivery through a network of more than 600,000 shoppers.
A customer expects to receive what they ordered, or a next-best option they can trust. Delivering on that expectation touches nearly every part of the shopping and fulfillment experience.
Quality was being improved across distinct product areas with different owners, objectives, and timelines. Yet a decision in one area could change the experience somewhere else. Improving the outcome meant treating those connections as part of the product work itself.
Need clear tools, achievable expectations, and fair measures of their work.
Need confidence that digital demand translates into reliable fulfillment.
Seven connected areas. Five teams. One quality outcome.
These weren't seven steps in a journey or seven independent teams. They were overlapping product investments, each shaping how a customer order ultimately came together.
Understand what's available
Store intelligence
Collect shelf conditions through shopper scans to create fresher inventory signals.
Inventory expectations
Help customers understand availability before committing to an order.
Help orders come together
Replacements
Translate customer preferences into practical decisions when items aren't available.
Finding items
Give shoppers meaningful aisle, shelf, and visual guidance in physical stores.
Progress and notifications
Surface order changes and decisions at the moments customers need them.
Contextual communication
Connect customers and shoppers to resolve item-level questions and exceptions.
Make quality sustainable
Shopper incentives and protections
Recognize high-quality work, explain expectations, and account for circumstances outside shoppers' control.
I made the connections part of how teams worked.
My role was to keep decisions from becoming isolated within individual product areas, while making sure each piece of the experience met a high bar on its own.
Shared visibility
I encouraged more work-in-progress sharing in team crit, giving designers exposure to related decisions and opportunities beyond their own projects.
Collaboration by design
I put the consumer and shopper Replacements designers into the same working group and meetings, so dependencies could be caught as the work evolved, even without me in the room.
Active experience oversight
Through project reviews, I surfaced conflicting approaches, questioned gaps in the thinking, and pushed teams to consider how their decisions affected the broader experience.
The goal wasn't to route every decision through me. It was to help teams make better connected decisions themselves.
Where my direction changed the work.
Three examples of different kinds of leadership: hands-on creative direction, evidence-led product judgment, and cohesion across connected experiences.
Shopper Quality Score
Making performance understandable and actionable
Quality metrics could affect shoppers' access to work. Early concepts simplified the score so much that shoppers couldn't tell what raised or lowered it, or what to do differently.
I challenged the simplified approach, proposed specific alternatives, critiqued the information hierarchy, and pushed for multiple rounds of research. Shopper feedback supported a more explicit explanation of score drivers, which materially changed the experience that shipped.
I also pushed beyond reporting issue counts. We connected the experience to customer-reported support issues, practical coaching, and deeper lessons, so shoppers could understand what had happened and how to improve.
Beyond the shipped experience: I brought the Quality Score and Picking teams together to explore how guidance could appear while shoppers selected items, before an issue occurred. That direction influenced roadmap discussions and remains planned work.
Finding Items
Making location guidance recognizable in the real world
We could describe an item's location through retailer aisle data, digital maps, shelf layouts, nearby products, or actual shelf photos. Each represented information differently, and shoppers had little time to interpret it.
I pushed for a guiding principle: make the location as immediately recognizable in the physical store as possible. I supported broad experimentation rather than prescribing one solution, then used qualitative feedback and experiment findings to evaluate what helped.
Shelf photography proved particularly useful because it matched what shoppers were looking at in front of them. It reinforced our direction while leaving room for other approaches across different retailers and item types.
Replacements
Keeping customer intent and shopper decisions connected
Replacements span two distinct experiences: how customers communicate what they want, and how shoppers act on that information when something is unavailable.
With a Staff Designer leading the work, I stayed closely involved in reviews and brought the consumer and shopper designers into the same working group. I focused on cohesion between the experiences and challenged proposed specificity against what catalog data and ML systems could reliably support.
That questioning contributed to a more grounded level of taxonomy detail in context-aware replacement guidance, without losing the broader ambition of making customer intent easier to honor.
Craft in smaller moments
Not every meaningful intervention is a major product redirection.
When shoppers were asked to scan shelves to improve inventory intelligence, I pushed for content explaining how those scans would help other shoppers find items in the future. It connected the task to a benefit shoppers could recognize, rather than presenting it as work done for the company.
Better decisions across the system.
This body of work supported improvements in fulfillment reliability, clearer shopper expectations, and more useful decisions during the order. The initiatives had different stages of maturity, from shipped improvements and observed experiments to future roadmap investments.
Clearer quality expectations
Shipped Quality Score experiences gave shoppers more detail about performance and actionable ways to improve.
More effective item finding
Experiments and shopper feedback supported shelf photography as a valuable tool for locating items.
More cohesive decisions
Shared working groups, broader crit visibility, and active design reviews helped connect changes across product boundaries.
These outcomes reflect the work of multiple design, product, engineering, research, and operations partners. My contribution centered on experience direction, connecting decisions across teams, and raising the quality of the work.
Leading quality at this scale meant staying close enough to the design to improve individual decisions, while keeping enough perspective to help the entire experience make sense together.