Enabling conversion optimization at a global scale

1. The situation

Regional teams running by their own playbooks

At PUMA, regional market teams often have their own agenda for many things. While successful in its own right, this strategy can create friction when it comes to shared digital resources and global governance. It also leads to the same work being duplicated across multiple regions and to inefficient use of the budget.

The situation was exacerbated by local teams hiring their own conversion agencies and running separate conversion optimization programs without coordinating them with each other or the global team. This resulted in:

• Overall spend exceeded $750K annually with subpar growth results
• Changes to the shared codebase had a negative impact on other regions
• Regions had historically owned the website before governance shifted to the global team and remained protective of their autonomy

Despite its flexibility, this approach created cross-regional friction, wasted spend, and risk to global UX and brand consistency. With more markets coming online, the problem would only compound.

2. The task

Alignment without a mandate

As Global Experience Digital Lead, I had to align regional and global teams to improve the CRO governance and experiment quality without having the authority to impose one on the other. If alignment failed:

• Continued financial waste and opportunity cost.
• Growing UX inconsistencies across regions.
• Loss of credibility for the global product ownership model itself.

3. The action

Building shared ownership

1. Diagnosed the root cause

I started with a detailed review of previously run tests and associated costs to confirm the subpar ROI. While the overall traffic volume allowed for decent sample sizes, many experiments were built around cosmetic changes with no clear hypothesis tied to business-relevant outcomes.

I mapped how experiments run by local teams were creating code and interest conflicts between regions. Also, I sought to understand the teams' perspective who feared losing autonomy and control over their work.

2. Built trust and reframed the problem

I continued with listening sessions to validate the expertise and surface real concerns before proposing anything. I then reframed the issue as a shared pain, not a criticism of any region specifically: higher-than-necessary costs and code conflicts that affected every region's roadmap.

I connected the fix to shared goals and established common principles, e.g. better insights from fewer, higher-quality tests, not just less autonomy. This helped lay the foundation for a global governance framework.

3. Created a structured, shared process

I created and set up a global testing framework with a lightweight experiment review and prioritization steps. Each test had to fulfill certain requirements which served as guardrails for quality and covered:

• Hypothesis structure and clarity
• Success KPIs
• Expected uplift range
• Estimated duration
• Required sample size
• Impact on other regions
• Other mechanics of well structured A/B tests

Regional teams retained ownership of their own tests, as long as they operated within these guardrails. Shared experimentation documentation in Confluence was set up to give every region visibility into what was being tested and why.

I also introduced feature-flag-based rollout for winning experiments, so a winning local test could go live for a given region without a deploy that risked breaking other regions' experience. Additionally, to make sure the insights collected during testing are instituionalized and to make planning easier, I introduced quarterly business review sessions.

4. Managed resistance strategically

In direct partnership with local teams I continued with a few quick, so the first results were wins that regional teams helped to produce. I celebrated those results publicly emphsizing local involvement, to build cooperative credibility rather than compliance.

I deliberately avoided top-down mandates throughout and emphasized mutual accountability with the help of the RACI framework.

6. The lessons

What Made It Stick

• Influence is earned through trust, shared goals, and visible wins not authority.
• The framework and the new approach became possible because teams helped shape it, not because it was handed down.
• Empathy and framing matter most when teams feel that their autonomy is threatened; the real objection is rarely the one stated initially.
• Transparency and shared accountability drive change that survives after you leave the room; mandates only last as long as enforcement does.