Website experienceExperimentationPersonalization

Improving the complete e-commerce journey

A structured optimization program using research, behavioral evidence, A/B testing and personalization to improve customer experience and commercial performance.

CompanyPUMA
RoleConversion Strategy & Program Owner
TeamUX/UI, Research, Analytics, Engineering
ToolsDynamic Yield, GA, MS Clarity, Figma, Jira

TL;DR

The program broke the funnel into distinct customer goals and measured the signals that predicted progression at each stage. My role connected strategy with execution: shaping hypotheses, aligning teams, prioritizing the roadmap and ensuring validated improvements moved into the product.

+115%
Conversion uplift
−30%
Return rate
+70%
Search engagement
+20%
Orders / user

1. Operating model

Marginal gains, applied systematically

The program broke the funnel into distinct customer goals and measured the signals that predicted progression at each stage. Page-level indicators guided decisions while transactions and revenue remained the commercial guardrails.

My role connected strategy with execution: shaping hypotheses, aligning teams, prioritizing the roadmap and ensuring validated improvements moved into the product.

Optimize each step for the decision the customer needs to make—then measure how those improvements compound across the funnel.

2. Optimization framework

Seven improvements across five stages

The same decision framework was adapted to different customer needs—from discovery to post-purchase engagement.

DiscoveryProductCheckoutAccountFit

3. Evidence

Validated improvements throughout the journey

01 · Discovery

A focused search fly-out

Hypothesis

Removing secondary elements from search reduces cognitive load and helps customers stay focused.

+70%Engagement
+38%Product views
Control
Search fly-out control
Variant
Focused search fly-out variant
02 · Discovery

Related search terms

Hypothesis

Relevant refinement terms reduce search abandonment and help customers explore useful alternatives.

+9.2%Product views
+3%Transactions
Refinement terms
Related search refinement terms
03 · Product

More product imagery above the fold

Hypothesis

A desktop image grid helps customers evaluate products faster and reduces PDP bounce.

−16%Bounce rate
+5%Transactions
Control
Product gallery control
Variant
Expanded product gallery variant
04 · Product

Behavioral social proof

Hypothesis

Showing when an item is selling quickly introduces relevant scarcity and increases add-to-cart behavior.

+19%Add-to-cart rate
+7.5%E-commerce CR
Selling-fast nudge
Selling-fast social proof message
05 · Checkout

Progressive disclosure in checkout

Hypothesis

Showing essential fields first makes checkout feel simpler and reduces abandonment.

+11.2%CTR to payment
+4.1%Checkout CR
Progressive form
Progressive checkout form
06 · Account

Personalized account recommendations

Hypothesis

Relevant recommendations create a new discovery surface and increase downstream conversion.

+40%PDP views
+20%Orders / user
Personalized recommendations
Personalized account recommendations
07 · Fit

Fit Predictor proof of concept

Hypothesis

Personalized footwear size recommendations reduce uncertainty, lift conversion and lower returns.

+115%Conversion uplift
−30%Return rate
Turkey proof of concept
Fit Predictor proof of concept
More selected workReturn to portfolio ↗