Joanna Leung

𝗖𝗢𝗡𝗧𝗘𝗫𝗧
After onboarding improvements increased paid product creation, sales didn’t follow automatically. Leadership needed to know where the remaining trials were being lost and which lever to pull next.

𝗧𝗛𝗘 𝗣𝗥𝗢𝗕𝗟𝗘𝗠
Of all trials, about 68% created a product, 38% reached 10+ store visits, 13% reached 100+ unique visits, and 6% reached $100 in GMV. Creating a product was no longer the binding constraint. The largest losses sat after launch: 66% of trials with 10+ visits never reached 100+.

𝗛𝗬𝗣𝗢𝗧𝗛𝗘𝗦𝗜𝗦
If creators publish stronger listings with less effort and are prompted to promote each product with a link back to the store, then more trials will reach meaningful store traffic and a first sale by day 14.

𝗗𝗜𝗦𝗖𝗢𝗩𝗘𝗥𝗬
Funnel analysis, interviews, and observed onboarding sessions showed:

Among trials with a product ready to sell (about 77%), writing the listing was a top friction point, often spanning multiple sessions. Creators also could not tell what a high-converting listing looks like.
Among trials with 10+ store visits, 45% had made no marketing post. Of those who posted, 57% drove no store visits because they lacked a call-to-action.

𝗣𝗔𝗜𝗡 𝗣𝗢𝗜𝗡𝗧𝗦
High effort and skill required to write a listing.
No visible quality bar for a listing that converts.
Little post-launch promotion, and promotion without a link or call to action.

𝗦𝗢𝗟𝗨𝗧𝗜𝗢𝗡
Three connected interventions mapped to the funnel:

Launch-time share prompt (Case Study 1), addressing missing promotion at the moment of highest momentum.
AI Product Listing Generator, building a complete listing from a short description, modelled on top-performing creator listings.

Any-time promotion nudge, prompting creators to share each new product with a call to action, extending the launch-time prompt beyond onboarding.

𝗘𝗫𝗘𝗖𝗨𝗧𝗜𝗢𝗡
Built the funnel baseline and located the largest stage drop-offs.

Segmented trials by intent to target listing work with creators who had a product ready.
Launched the listing generator as a staged experiment, gated on listing acceptance rate (55% threshold) rather than usage.
Shipped the promotion nudge in two phases: first at store launch, then at any point after a product launch.

𝗜𝗠𝗣𝗔𝗖𝗧
Redirected roadmap priority from product creation to store traffic, the largest underserved leak.
The launch-time share prompt contributed to a +7.9% lift in trials making a marketing post. This moved the behavior behind the largest drop-off, though the effect on store visits was not measured.
The listing generator cleared its 55% acceptance gate [confirm achieved rate] and opened to all trials.
The any-time nudge and the listing generator did not reach a conclusive conversion readout before handoff, so their effect on first-sale rate is unmeasured.

𝗣𝗢𝗦𝗧-𝗟𝗔𝗨𝗡𝗖𝗛
Open questions carried to the team: the first-sale effect of the listing generator, read intent-to-treat on the experiment split; whether the any-time nudge lifts store visits and not just posts; and whether promotion prompts help creators with small audiences.

Role

Founding Product Manager

For

Creators

Date

Oct '24 – Apr '25

Type

GenAI, web app, experimentation, onboarding, growth

URL

www.stan.store/