UX Research · Service Design · 2023 · Shipped
UX Designer & Researcher
Claro.
digital cancellation.
Turning Claro's manual, 13-click cancellation into a clear digital journey — one that retains customers by understanding why they actually want to leave.
¶ Snapshot
Client
Claro
Year
2023
Role
UX Designer & Researcher
Duration
~6 months (survey Jul 2022 → proposal Q1 2023)
Team
Cross-functional: UX, operations, backoffice & stakeholders
Platform
Web · Minha Claro residencial
Status
Shipped
Scope
UX Research · Service Design · Product
My contribution
Led the end-to-end research and journey design for residential digital cancellation: desk research, interviews with the operations and backoffice teams, stakeholder interviews and user interviews. Mapped the full cancellation journey and its pain points, then proposed the redesigned journey and the UX metrics to measure it.
¶ Problem
What was broken
Context
Cancellation at Claro was a fully manual process on the Minha Claro residencial site. The digital journey took 13 clicks — and customers needed up to 15 clicks just to find where to cancel — so most gave up and called the contact centre instead. Fewer than a quarter of cancellation requests could be reversed by the company.
Baseline
<25% of requests reversible · 13 clicks to cancel · up to 48h to first contact · 15+ clicks just to find cancellation
If we did nothing
90% of cancellation requests came from customers who only wanted a discount — not to leave. Handling them by phone cost up to 56 BRL per à-la-carte cancellation (up to 7 calls) against a 40 BRL average ticket, making it one of the most expensive flows to run manually.
¶ Constraints
- 01Legacy backoffice flow: request → SMS within 3h → email within 48h → manual reversal team, with no way to qualify customer intent up front.
- 02Outdated registration data made customers hard to reach — only 25% could be contacted in time to negotiate.
- 03A cancellation could be requested by a family member rather than the contract holder, and separate NET/SMS cancellations sometimes failed, generating undue-charging complaints.
¶ Process
How we worked
Research
A mixed-method study run between July 2022 and Q1 2023: desk research, interviews with the operations and backoffice teams, stakeholder interviews, user interviews, a full mapping of the cancellation journey (front- and back-office) and its pains, and the definition of UX metrics to track the new experience.
Hypotheses
- H1The main reason for cancellation requests is not that customers want to leave Claro — validated: 90% use the journey to get a discount.
- H2Cancelling à-la-carte by phone costs more than doing it digitally — validated: up to 7 calls, ~56 BRL cost against a 40 BRL ticket.
- H3Informing customers of fines at cancellation only postpones churn — validated: customers retained by fines don't understand the real value of Claro's products.
¶ Decisions
Each choice listed beside the alternative we deliberately rejected — and why.
D01
We chose
Cut the digital journey from 13 to 4 clicks and let customers self-serve à-la-carte cancellations.
Over
Keep funnelling every cancellation through the call centre.
Because
Phone handling cost up to 56 BRL per cancellation against a 40 BRL ticket, plus a 48h delay that killed any chance to negotiate. A short digital flow removed both the cost and the delay.
D02
We chose
Qualify intent up front — separate customers who genuinely want to leave from those hunting a discount.
Over
Treat every request as churn and counter with a fine or a price.
Because
90% of requests were discount-seeking; leading with fines only postponed churn and eroded trust. Qualifying intent let retention focus where it actually mattered.
D03
We chose
Show only the products the customer actually has, in plain language, with a clear next-bill amount.
Over
Keep the dropdown listing every Claro product and the generic 'cancel' form.
Because
Customers didn't understand 'à-la-carte', couldn't say why they were cancelling, and didn't trust what they'd be charged next. Clarity — not friction — was the real retention lever.
¶ Craft
¶ The backoffice flow
Before the redesign, every cancellation was handled by hand — a chain of automated messages racing a 48-hour clock before a manual reversal team could act.
Request generates a protocol
SMS to the customer within 3h
Email contact within 48h
Reversal team acts if they reply
↳ Understand the reason
Wants to negotiate a discount
Technical problems — signal, wi-fi
Billing problems — undue charges
Genuinely wants to cancel
“When customers call us, up to 85% of the time we are able to reverse their intention to cancel.”
— Claro reversal team
¶ Mapped pains
Across operations, backoffice and user interviews, the same friction kept surfacing.
The form only said “cancel” — no reason, no context to negotiate.
Outdated data: only 25% of customers could be reached in time.
A 48h window too short to make a retention offer.
13 clicks to cancel — and 15+ just to find where.
A dropdown listing every Claro product, not the customer's.
Customers unsure what their next bill would be.
¶ Why people ask to cancel
Six reasons surfaced in research — most solvable without losing the customer.
- 01I don't use this product anymore
- 02I didn't like the product
- 03The price is too high for me
- 04I had technical problems
- 05I had billing problems
- 06I'm moving to a competitor
¶ Finding cancellation
Research moved cancellation from a buried, 15-click dead end into a clear route in the main menu.
Before · Jul 2022
After · Q1 2023
¶ What shipped in the MVP
- From 13 to 4 clicks to request a digital cancellation
- From 48 hours to 30 minutes to first contact
- À-la-carte cancellations customers can self-serve
- Intent qualified up front — leavers vs. discount-seekers
- No more duplicate à-la-carte cancellations
- Retention negotiated on value, not just fines and price
¶ Measuring it — the HEART model
Success was framed across happiness, engagement and adoption, each with its own signals and metrics.
Customers can cancel products they don't value
À-la-carte sales · NPS · positive feedback
Can we still retain customers?
Cancellation requests · plan changes · new contracts per CPF
More people using the digital flow
Cancellation access · à-la-carte purchases · returning users
¶ Outcome
4
Clicks to cancel
digital cancellation request
30min
Time to first contact
85%
Reversible by phone
when the customer was reached
Qualitative signal
Research reframed cancellation from a churn event into a retention moment: most people requesting cancellation only wanted a discount or clarity, not to leave. That insight shifted the team away from fines-and-price arguments toward qualifying intent and showing real product value.
Business impact
The proposed journey shipped after research (Q1 2023): a shorter self-serve flow that lets customers cancel individual à-la-carte products without losing the contract, qualifies who really wants to leave, avoids duplicate à-la-carte cancellations, and moves expensive manual handling off the call centre.
¶ Reflection
What I'd do differently
The biggest unlock wasn't the interface — it was reframing cancellation as a retention moment rather than a loss. Once we could tell discount-seekers from genuine leavers, the journey almost designed itself. Next time I'd instrument the à-la-carte behaviour from day one, since the open question we left was simple: once it's this easy to cancel, how do people actually behave?
Next case · 02 / 04
Open ↗Grupo Boticário
store experience.
