Power Subscriber Diary Study
Senior UX Researcher at Amazon
Summary
The gap: How Subscribe & Save's most engaged customers managed the program at a scale that wasn't visible in standard usability sessions or transactional surveys, leaving the team without a clear read on how power subscribers actually interacted with tools and any work-arounds they may use.
What I built: A 6-week longitudinal diary study capturing power subscribers' real Subscribe & Save tasks in context, adding subscriptions, swapping out-of-stock items, skipping deliveries, changing variations, managing payment, etc. through screen recordings written reflections on the Dscout platform.
Why the design mattered: I built the study to be refreshable from the start, so the same instrument could be re-fielded later and produce findings comparable against the original baseline.
Scale: 2,018 applications, 25 participants selected, 23 completing both study missions across two rounds (2021 and 2025).
Where it shows up today: Surfaced key findings that have shaped how the team understands its highest-value segment: out-of-stock communication, the need for bulk actions, duplicate subscriptions, and workarounds power subscribers invented to control delivery management. The 2025 refresh gave leadership a longitudinal view of which problems had been solved, which persisted, and where new friction had emerged.
The Full Story
The Problem
Subscribe & Save's most engaged customers, those with 11+ or more active subscriptions, were also the hardest to learn from. They were managing the program at a scale most users never experienced, which meant the friction they were running into wasn't visible in standard usability sessions or transactional surveys. A 60-minute lab study can't capture what it's like to skip the right delivery in a list of 100+ items, or how customers actually behave when an out-of-stock alert appears before a delivery they were counting on. A 6-week diary study launched in May 2021.
The team needed to understand how power subscribers were actually using the program in their lives, week to week, what they reached for, where they got stuck, and what they did when the tools didn't quite work. And later, four years on in October 2025, the team needed to know whether the picture had changed: had the friction shifted, had new pain points emerged, and were the original findings still the right basis for prioritization?
The Approach
I designed a 6-week diary study to capture power subscribers' experience in their own context, in their own words, on their own schedule, and built it to be repeatable so the team could come back to it as the program evolved. The 2021 and 2025 studies were recruited and designed consistently.
Recruitment built for the audience. I targeted US Subscribe & Save customers with 11+ active subscriptions. Of 2,018 applications, I selected 25 participants and 23 completed both study missions. The high application rate confirmed the segment's engagement; the completion rate confirmed the study design respected their time.
Two-mission structure. I built the study in two parts, each answering a different kind of question. A short mission asked participants to reflect on their overall experience, what they liked most and least, plus the highest and lowest moments that stood out, to give me a baseline read on satisfaction and the emotional shape of the program at scale. A long mission ran across six weeks and asked participants to record themselves completing real Subscribe & Save tasks as those tasks naturally came up: adding new subscriptions, changing variations, swapping out-of-stock items, reviewing emails, skipping deliveries, managing payment methods. Participants submitted screen recordings, written reflections, and structured ratings through the Dscout platform.
Designed to capture intensity, not just volume. Each diary entry asked participants not only what happened but how strongly they felt about it, using a 1-to-5 intensity scale alongside open-text reflection. This let me distinguish a mildly inconvenient moment from a deeply frustrating one, and identify which friction points were merely common versus which ones were doing real damage to the customer relationship.
Built to be refreshable. I designed the study with comparability in mind from the start, so the same instrument could be re-fielded later and produce findings that could be tracked against the original baseline. In 2025, four years after the original study shipped, I refreshed it , same recruitment criteria, same mission structure, same intensity scale, to measure how power subscriber behavior had evolved as the program and the underlying tools changed.
Synthesis and stakeholder distribution. I performed the analysis across all submissions for both rounds, organizing findings by task type and weaving together quantitative tagging, behavioral observation, and direct customer voice. I drafted both reports, presented findings to product and engineering leadership, and distributed the work across the broader Subscribe & Save organization.
The Impact
The original 2021 study became a foundational reference for how the team understood its highest-value customer segment. The 2025 refresh confirmed which problems had been solved, which had persisted, and where new friction had emerged, giving leadership a longitudinal view few research programs at the company can offer.
Four findings carried strategic weight across both rounds:
Out-of-stock and delivery delay communication was the highest-intensity pain point. Vague messaging left customers unsure whether to wait, skip, buy in store, or cancel altogether, and the lack of clarity was often more damaging than the missing delivery itself. This was the top dissatisfaction driver in 2021 and remained a leading friction point in 2025.
Power subscribers needed bulk actions, not item-by-item management. Customers with 50+ subscriptions described navigating long scrolling lists and being unable to find specific items, particularly when they didn't know an item's next delivery date. Editing each subscription individually didn't scale to how power subscribers actually used the program.
The variation and substitution flow pushed customers to create duplicate subscriptions. When customers wanted to change a flavor, scent, or size, the path of least resistance was often to start a new subscription rather than edit the existing one, leaving them with two active subscriptions when they only wanted one.
Customers were inventing workarounds to control delivery timing. Rather than using skip, change frequency, and CoPa as designed, power subscribers would set every subscription to monthly, or change frequencies to indirectly move a delivery up or push it out — strong evidence that the existing tools didn't match how this segment actually thought about timing.
What I Took Away
A few things from this project I bring to every role:
Match the method to the audience. Power subscribers couldn't be studied in a one-hour session, their experience only made sense across weeks of real use. Designing the study around their actual rhythm is what made the findings credible.
The richest signal often shows up in workarounds. When customers set every one of their 100+ subscriptions to monthly just to gain control over delivery timing, that's not a user error, it's evidence that the existing tools don't match how power subscribers actually think. Workarounds are some of the most valuable data a study can surface.
Design longitudinal research from the start. Most studies are built once and forgotten. This one was built to be re-fielded, which meant four years later the team could measure change rather than start over, a much more valuable position for leadership to be in.