Defined the growth problem
Connected inspiration, product discovery and basket-building into one commercial opportunity.
Product Management
Growth
E-commerce
Shop the look
Turning styled product imagery into a measurable path from inspiration to basket.
TL;DR
Styled product photography creates desire, but the shopping journey breaks when customers have to identify and find every item separately.
I reframed “Shop the Look” as a growth hypothesis rather than a feature request: reduce friction in the product discovery, increase the number of shoppers reaching product pages, and encourage multi-item purchases.
The work turned the idea into a decision-ready MVP proposal with competitive evidence, a deliberately narrow scope, a measurable funnel, an A/B testing plan and a staged roadmap.
My contribution
Moving the conversation from interface to investment
Scoped the smallest useful MVP
Separated the core customer loop from substitutions, curation and personalization.
Designed the measurement model
Defined a causal test, leading indicators, commercial outcomes and experience guardrails.
1. The opportunity
The customer problem was discovery, not inspiration
The onsite product photograhy showed customers how individual products could work together and inspired them to purchase the whole look. Yet the experience treated that inspiration as a dead end: finding the jacket, shirt, trousers and shoes meant starting separate searches and reconstructing the outfit manually.
This extra effort created significant friction at the moment of highest intent. The opportunity was to preserve this momentum and turn existing product imagery into a conversion area of its own.
1. See the look
A styled image creates desire and provides a complete outfit idea.
2. Identify items
The customer must work out which specific products the model wears.
3. Search again
Every extra query or click increases friction and abandonment risk.
4. Rebuild the outfit
Items added to the cart only if the customer successfully reconnects the products.
Growth hypothesis
If customers can access the exact products worn by a model at the moment of inspiration, more of them will progress to product and basket—and complementary items can increase units per order and average order value.2. Market evidence
Shoppable looks are an expected pattern
Competitor examples showed several versions of the same underlying behavior: product lists beside editorial or social imagery, hotspots on models, complete-look drawers and direct size selection. This reduced interaction risk, as the pattern was already well understood by fashion shoppers.
The benchmark was used as directional evidence, not proof of PUMA-specific commercial impact. The latter still required controlled validation.
3. MVP strategy
Scope the first release around the riskiest assumptions
The first release did not need to solve everything like, e.g. substitutions or personalization. It needed to prove that the core loop was technically feasible, clear and commercially valuable.
Prove the core loop
- Map model imagery to sellable SKUs
- Show exact items, price and availability
- Link directly to the relevant product
- Handle sold-out products clearly
- Minimize manual merchandising
- Track the image-to-order funnel
Delay sophistication
- Alternative products for sold-out items
- Seasonal or editorial curation
- Behavior-based recommendations
- Personalized substitute ranking
- Broader cross-sell orchestration
4. Product experience
Use a familiar pattern and shorten the path to basket
The interaction was based established shopping cues: a visible entry point on the model image, a complete-look view, item details and a clear path to selection. Familiarity meant that customers did not have learn a new behavior or pattern.
The product decision behind the interface
The goal wasn't maximizing clicks on a new widget, but to preserve the intent across the journey: model image → look → product → basket. Each interaction had to bring the customer closer to the purchasing decision.
5. Growth measurement
Define the metric tree before committing to scale
Engagement could diagnose behavior, but it could not justify investment on its own. The measurement plan connected interaction signals to incremental commercial value and protected the surrounding journey from unintended harm.
Commercial lift
Incremental revenue per exposed user and look-to-order conversion.
Journey progression
Entry-point engagement, look-to-PDP rate, add-to-cart rate, units per order and AOV.
Experience quality
Page speed, PDP conversion, returns and exits caused by unavailable products.
Validation method
Instrument the full funnel, randomize eligible traffic into exposed and control groups, and compare incremental outcomes through a controlled A/B test. Segment diagnostics can explain where the journey changes; the randomized result determines whether the feature creates value.
6. Investment logic
Evidence—not enthusiasm—determines the next release
Before build, the proposal required an engineering estimate and a comparison between an owned solution and third-party tools. The decision was therefore structured around explicit gates rather than an unqualified revenue forecast.
01
Ship the MVP only if expected incremental margin can reasonably clear build and operating costs.
02
Compare an owned build with third-party cost, flexibility, maintenance and time-to-value.
03
Fund substitutions and personalization only after the basic journey demonstrates causal lift.
7. Product roadmap
Sequence investment from certainty to sophistication
Exact products
Show the products worn by the model and measure incremental journey and commercial value.
Relevant alternatives
Prevent dead ends by offering appropriate substitutes when an exact item is unavailable.
Ranked substitutes
Use behavior, affinity and context to determine which relevant alternative each customer sees.
Outcome / Next step
A feature idea became a testable product decision
The work produced a coherent growth case: a defined customer problem, a causal hypothesis, competitor evidence, an MVP boundary, an instrumented funnel, explicit investment criteria and a roadmap that delayed complexity until the core loop proved itself.
The honest next step was not a full rollout. It was engineering sizing followed by a controlled pilot on eligible product pages. Because the available draft does not contain verified launch results, this case intentionally stops at the decision-ready proposal rather than presenting forecast impact as achieved performance.
Product lesson: a credible growth case connects customer value, commercial value and evidence. The interface is only one part of that system.