In-app content

As part of an ambitious app-wide experiment, I was asked to produce 10x our usual amount of Instagram-style 'story' content for our home feed in just one quarter. To achieve this without compromising quality, I developed an AI-assisted editorial workflow, combining automation with human insight.

I began by writing a set of prompt-engineering rules specific to pregnancy. Then, using a combination of user research, existing data, and medical insights, the team identified priority topics together – I was able to then use editorial expertise to come up with a snappy angle for each. I sourced trusted medical references for each piece, then used AI to generate first drafts. Every story was then carefully edited, reviewed by a medical expert, visually brought to life with a designer, and passed through legal and proofreading checks.

By streamlining processes week by week (and working at a breakneck pace), I was able to deliver 80 high-quality pregnancy stories in just eight weeks. The experiment resulted in a 5% uplift in free-to-paid conversions, and a 4% increase in Premium users returning to the stories feed on consecutive days. Some of the stories are now top-performing – the '3 surprising foods and drinks to avoid' story shown here, for example, has a 10% CTR.

This shows that AI, when used thoughtfully and with rigorous checks, can scale impact without losing trust.

Polycystic ovary syndrome (PCOS) impacts a significant number of women and people who menstruate globally, with estimates ranging from 4% to 20%. However, it takes years to get a diagnosis. Flo identified PCOS as a high-impact opportunity for early detection and intervention, planning a new diagnostic tool – Symptom Checker – which would include key information on the condition. The tool would also direct users to a dedicated PCOS content hub, known internally as an intent page. 

I proactively stepped forward to lead the content strategy for the Symptom Checker and its intent page. I gathered our existing data and insights into a Miro board to map user needs and pain points, then developed a clear, insight-led content plan to guide the work. I presented this to key stakeholders across Product, Content, and Medical, facilitating a collaborative brainstorm to refine the direction before writing and commissioning all content (including UX copy, infographics, articles and curated community posts) for the page. I also took full ownership of timelines and delivery, keeping everyone aligned and on track throughout.

The PCOS Symptom Checker now has the highest click rate in the Flo app, while the intent page attracts an average of 620,000 users within a 14-day period. Based on this success, Flo has launched Symptom Checkers and accompanying intent pages on a range of conditions, from endometriosis to uterine fibroids.