From reading complaints to owning the product

Mandate

Make one product serve older learners, children's guardians and teachers safely as paying membership passed one million.

Outcomes

• Replaced an automatically converting trial with an end-of-trial choice; trial refund requests fell from 3,900 to 760 a month
• Moved children's practice videos to guardian-controlled visibility; weekly review items fell from 41,000 to 9,500
• Restricted messaging between unconnected adults; solicitation reports fell from 8,600 to 1,900 a month
• Paying members grew from 0.92 million to 1.38 million between early 2021 and June 2026

Decisions in context

My appointment as chief product officer brought children's courses, community features and safety tooling into a remit that had previously covered adult learning alone. The company had passed 900,000 paying members and was being asked questions by parents, regulators and its own reviewers that the product had not been designed to answer. I reorganised the function around the people using it: older learners, guardians and children, teachers, and the reviewers whose tools we build. The three cases on this site come from this period, and each involved a disagreement with a capable colleague. I present a product-risk report to the board every quarter but hold no directorship. Our co-founders and two venture investors weigh growth differently, and part of my job is to state plainly what a safer default will cost before they are asked to approve it.

Reason for the move

Present role; I would add one board seat alongside it rather than leave.

Mandate

Turn a set of recorded calligraphy lessons into a membership product that older beginners could use without help.

Outcomes

• Launched a reading mode with larger type and fewer choices per screen in 2017
• Introduced family payment so that a relative could pay while the learner kept the account
• Established a standing panel of 140 members aged sixty and over for release testing

Decisions in context

When I joined, the company was a year old and had about sixty staff. Its lessons were filmed with care, but its app assumed users who already knew how to search, scroll and pay. Our members had often been given a phone by their children and had learnt only video calls. I spent my first month in community centres watching people try to find a lesson. The reading mode, voice search and family payment that followed came from those sessions rather than from analytics. I also learnt to refuse features that pleased younger staff in testing and confused the members they were meant for. Paying membership reached roughly 900,000 by the end of the period, and the product team grew to 120. The arrangement in which a relative paid while a learner used the account later shaped how we designed for children and guardians.

Reason for the move

Appointed chief product officer when children's courses and community features were combined under one product leader.

Mandate

Move accounting and teaching-certificate preparation from desktop courses to mobile, and reduce disputes over course packages.

Outcomes

• Mobile share of study sessions rose from 22% to 71%
• Replaced seven conditional refund promises with three plain package terms
• Payment disputes fell by about 45% while sales of the most expensive package declined

Decisions in context

As product director I inherited packages sold with promises such as a refund for a candidate who failed after completing every lesson. Few people could claim them, and the disputes consumed customer service and damaged trust among the candidates who most needed help. I proposed three simpler packages with terms a candidate could check for herself. Our sales director warned that removing the conditional refund would reduce sales of the premium package, and it did. Disputes fell by more than sales, but the trade-off was real and I reported both figures. The mobile work was the larger programme: practice questions, short lessons and progress tracking designed for commuters. I worked closely with the chief operating officer and the teaching centres, and I learnt that anxious customers read product terms very differently from the staff who wrote them.

Reason for the move

Joined an early-stage company building learning products for older adults.

Mandate

Build the company's first mobile apps for practice questions and short lessons.

Outcomes

• Launched a practice-question app used by 1.1 million candidates in its second year
• Introduced offline lesson downloads for commuters with limited mobile data
• Designed the first in-app purchase flow, with the full price shown before confirmation

Decisions in context

Smartphones were spreading quickly among the young professionals preparing for qualification examinations, yet the company's courses had been built for desktop computers. I led a small team building its first mobile apps. We started with practice questions because candidates studied in spare minutes on trains and in queues, and costly mobile data made long video impractical. Offline downloads were the most valued feature of our first year. I designed the purchase flow so that the full price and what it included appeared before confirmation, a choice shaped by my years handling subscription complaints. The team was small enough that I wrote many specifications myself and listened to support calls every week. The apps gave the company a mobile channel it had lacked, and their growth formed part of the case for the private-equity investment of 2012.

Reason for the move

Promoted to product director.

Mandate

Grow forums and comment sections on the handset portal while keeping content within the portal's rules.

Outcomes

• Community channels reached 4.3 million monthly users by 2009
• Introduced a priority queue so reviewers saw serious reports first
• Added a review-workload estimate to every community feature proposal

Decisions in context

The portal served people who browsed on basic handsets, often over slow connections. I managed forums, comment sections and a question-and-answer service in which readers helped one another. Growth was rapid, and the review team of twenty-five struggled to keep pace with posts. I began sitting with them every second week and saw that some of our features generated work no one had estimated. A reply-notification feature, for example, doubled the volume of reported posts within a month. I proposed a priority queue so that reviewers handled serious reports first, and I added a workload estimate to every new community proposal. My manager initially regarded the estimate as a delay. After the reply-notification episode it became a standard part of our planning papers, and the habit of costing review work before launch has stayed with me.

Reason for the move

Moved to mobile learning products as smartphones changed how people studied.

Mandate

Prepare daily subscription content and reconcile customer complaints with subscription records.

Outcomes

• Matched complaint cases with subscription logs for operator review
• Rewrote sign-up messages to show price and frequency before confirmation
• Prepared a monthly complaint analysis used in product decisions

Decisions in context

My first employer sold weather notes, proverbs and quizzes by text message, charged monthly through operators' bills. I wrote and scheduled content for several services, but my most important work turned out to be complaint analysis. I matched each complaint with our subscription logs and prepared cases for the operator. Many customers had technically agreed to a subscription without understanding that they had done so. When our largest operator partner asked for a clearer confirmation step, I helped rewrite our sign-up messages so that price and frequency appeared before the customer agreed. Subscriptions fell, and my manager accepted the loss because the alternative was losing access to that operator's billing. I left with a lasting interest in the moment a person agrees to pay, and in how an accurate record can still fail to explain a charge to the person who receives it.

Reason for the move

Joined a mobile portal to work on products that customers chose to visit.