Farshad Amiri
← All case studies

Case study 01 · NatureLand · Subscription

From repeat calls to a flexible subscription

Loyal parents kept calling to change their monthly diaper bundle. I used design thinking to turn that friction into a subscription they control, on the app and the website.

Role
Senior Product Designer
Team
Tech, Product, Marketing
Platforms
Mobile app and website
Markets
Kuwait (A/B test), GCC (launch)
Reading
8 min
Visit NatureLand ↗
NatureLand Subscription: “Subscribe once. Edit before every delivery.” The website’s Subscription List with active, paused and pending bundles, a Diaper bundle · Size 4 card, and the app’s subscription detail with delivery discount progress and the list of deliveries.

At a glance

Flow A won the A/B test and launched across the GCC

Calls about bundle changes almost disappeared.

−96%

support calls about bundle changes

About 40 a week before, about 7 a month after launch

3.7×

payment completion, Flow A vs Flow B

44% vs 12% · 500 users per flow · Kuwait, one month

118

active subscribers across the GCC

Net of cancellations and pauses, three months after launch

Flow A

launched to all users

Chosen interval, growing discount, edit, pause and cancel

Process

Design thinking, end to end

Five stages, each one feeding the next. The sections below follow the same order.

  1. 01

    Empathize

    Reviewed 80 random callers and profiled customers with RFM.

  2. 02

    Define

    Framed one problem statement and two How Might We questions.

  3. 03

    Ideate

    Mapped ideas by impact and effort, and chose a subscription.

  4. 04

    Prototype

    Designed two subscription flows for the app and the website.

  5. 05

    Test

    A/B tested both flows with 1,000 users for one month.

01 · Context

A bundle people loved, and a reorder that could not flex

Natureland sells monthly diaper bundles in six sizes: 4 packs of water wipes plus 8 to 10 boxes of diapers, about 10% below list price. Before this project, “Reorder” repeated the last bundle exactly. When a baby moved up a size or a family needed fewer wipes, the only option was to call.

~40

calls and tickets a week

about bundle changes, around 520 every three months

Manual

edits

support and admin retyped every changed order by hand

6

bundle sizes

size 1 to size 6, ordered about once a month

Before: the reorder journey

  1. Bundle arrives“This works for us.”
  2. Diapers run low“Time to reorder.”
  3. Opens My orders“Let me tap Reorder.”
  4. Reorder repeats the same bundle“But she needs size 5 now.”Pain point
  5. Calls customer service“I hope they get it right.”Pain point
  6. Admin edits the order by hand“Is the price still right?”Pain point
  7. Delivery“Did the right size come?”

Steps 4 to 6 are where the experience broke: the site sent people to the phone, and the team did the work by hand.

02 · Empathize

Listening to the people who called

With no time for interviews, the team reviewed the profiles of 80 random callers from the three months before the project, each counted once. All 80 were active, loyal customers by purchase, payment, repeat rate and the marketing team’s score.

What the callers wanted

  • 80callers reviewed100%
  • 69wanted to change their bundle86% of all callers
  • 33also wanted regular delivery48% of the 69 · 41% of all callers
Call review · 80 random unique callers · real data, scaled

Two needs, one group of people: flexibility (change the bundle) and routine (have it arrive on its own).

Who the callers are — RFM profile from order history

  • Value

    The bundle saves money, so callers keep it instead of buying items one by one.

  • Routine

    Callers reorder steadily, about once a month. They already behave like subscribers.

  • Predictable change

    Babies move up a size every few months, so the same request comes back again and again.

Patterns shown; exact figures under NDA.

Persona · built from the call review

Sara, 31

Mother of a 10-month-old · Kuwait City · orders on her phone

Goal
Never run out of diapers, always the right size, at a fair price.
Pain
Reorder sends the old size, and calling with a baby in her arms is hard.
In her words
“I just want the same bundle, one size up. Why do I have to call?”

03 · Define

One problem, two questions

I turned what the callers wanted into one problem statement and two How Might We questions, then listed the assumptions to test before designing.

  • HMW 1

    How might we turn repeat buying into a subscription that delivers on its own and lets people change size and contents before every delivery?

  • HMW 2

    How might we keep the bundle’s savings and give people a reason to stay subscribed?

Working assumptions — each one was tested before the design was final

Working assumptions — each one was tested before the design was final
AssumptionHow we tested it
Most changes are a new size, a box count or wipesSurvey Q2
“Regular delivery” means automatic delivery at a set interval, editable before each oneSurvey Q6 and Q7
The 80-caller review represents all callersSurvey Q1, compared with the review’s 86% and 48%
The business keeps the bundle discount after the contents changeMeeting with the business team

04 · Ideate

Many ideas, one direction

Ideas were grouped by the need they answer, then placed by their impact on calls and their cost to build.

Quick wins (High impact, Low effort)

  • Editable reorder
  • Next size suggestion
  • Same flow for the support team

Big bets (High impact, High effort)

  • Flexible subscription
  • Edit window before each delivery

Fill-ins (Low impact, Low effort)

  • Reminder before running out
  • Saved preferences

Not now (Low impact, High effort)

  • Personal bundle builder

05 · Validate

A survey that turned needs into requirements

Interviews were not possible, so I designed a 10-question Google Form for callers, sent before the subscription was designed. Q1 repeats the call review’s funnel, so the survey checks the review before it adds anything new.

  • Q1–Q5 · Why they called

    Q1 sorts people into three groups: a bundle change, a change plus regular delivery, or another reason. Q2 to Q5 ask what they wanted to change and why.

  • Q6–Q7 · Automatic delivery

    How people decide when to reorder, and what would stop them from automatic delivery, read separately for each Q1 group.

  • Q8–Q10 · Use and contact

    How steady their use is, what calling gives them that the site does not, and one open question.

From answers to requirements

From answers to requirements
RequirementWhySource
Delivery at a chosen interval, monthly by defaultMost callers reorder when diapers run low or on a fixed day, and they already reorder about monthlySurvey Q6 · RFM
Automatic delivery, with a reminderMany reorder only when they remember, sometimes lateSurvey Q6
Change the size inside the subscriptionA new size is the most common change and the main barrier to automatic deliverySurvey Q2, Q7
Edit box count and wipes before each deliveryUse changes from month to monthSurvey Q2, Q8
A discount that grows for stayingPrice is the main reason people keep the bundleSurvey Q4 · RFM
A summary to confirm before each deliveryMany called just to be sure the order was rightSurvey Q9
Pause or cancel at any timeCommitment worried people who only wanted a changeSurvey Q7

Survey results are summarized as findings; their figures are not shown.

06 · Prototype

Subscription inside the shopping path

Flow A puts subscription inside the shopping path. Every bundle card carries a Subscribe badge, and from the bundle’s page people choose Subscribe instead of Add to cart, adjust the items, pick an interval, set when and where it arrives, and pay — all in one continuous flow.

  1. Bundles list · Subscribe badge
  2. Bundle page · Add to cart or Subscribe
  3. Adjust items · pick interval
  4. Start date, days, time · address
  5. Review checkout
  6. Pay
  7. My subscriptions in profile

Flow A · App

Flow A · Website

07 · Test

Flow A’s payment completion was 3.7× higher

1,000 active users in Kuwait were split at random, 500 per flow, and both flows ran for one month in the same period. We measured payment completion: of the users who tapped Subscribe and reached the cart, the share who completed payment.

Payment completion after Subscribe → cart

  • Flow A44%
  • Flow B12%
Bar length = share of users who reached the cart after Subscribe and completed payment. Scale 0 to 100%.

+32 points

difference in payment completion between the flows

44% vs 12%

3.7×

higher payment completion with Flow A

Than with Flow B

500 + 500

users per flow

Kuwait, same month, random split

08 · Outcome

What launched, and what changed

Flow A launched across the GCC. The subscription now delivers at a chosen interval, gives a discount that grows with each delivery, and lets people edit, pause or cancel at any time.

  • Before

    ~40 calls and tickets a week

    About 173 a month, each one edited by hand

  • After

    ~7 calls a month

    About 96% fewer calls about bundle changes

  • Three months in

    118 subscribers

    Active across the GCC, net of cancellations and pauses

  • Active subscribers grew every month

    • 34Month 1+34 new
    • 78Month 2+44 new
    • 118Month 3+40 new
    Active subscriptions at the end of each month, net of cancellations and pauses · GCC
  • Discount by delivery

    A reason to stay subscribed, answering the “price” finding.

    • 10%1st delivery
    • 15%2nd delivery
    • 20%3rd delivery
    • 25%4th delivery

Shipped in Flow A

  • Chosen delivery interval
  • Growing discount, 10% to 25%
  • Edit details
  • Pause anytime
  • Cancel anytime

09 · Reflection

What I learned

Three lessons I carry into every project, and what I would do next.

  • Calls are research

    Support calls were the clearest signal we had. Reviewing 80 of them shaped the whole project.

  • Flexibility makes routine possible

    People commit to automatic delivery only when they can change it before every delivery.

  • Test the route, not only the idea

    Both flows met the same requirements. The route through them made a 3.7× difference.

Next steps

  1. Confirm that “Edit details” covers size, box count and wipes.
  2. Add a reminder and an order summary before each delivery.
  3. Run the survey with live subscribers and track retention by discount step.