How a Single A/B Test generated $1M+ in Revenue

“Experimentation is one way that a company that’s product driven can guarantee that it’s data driven.”

I recently had a chat with Tristan Burns, ex-Global Head of Analytics at Pizza Hut Digital Ventures (PHDV), and this sentence stuck with me.

Tristan told me the story of how a single experiment generated a projected £1 million in additional annual revenue and how he built the data team at PHDV from scratch.

Starting From Scratch: The Early Days of Data Analytics at Pizza Hut Digital Ventures (PHDV)

Tristan is a veteran in the data space with 15 years of experience across data, financial analysis, and strategy roles. He has built and scaled data teams in both startups and corporate environments and also founded two marketplace businesses.

PHDV operated as an internal startup within Pizza Hut International, set up to build and scale commerce solutions for all non-US markets. Operating in over 120 countries, their mission was to create consistency in online experiences across all international markets while managing restaurant backend technology and data infrastructure.

When Tristan joined PHDV as a Senior Analyst in 2018, the data landscape was minimal. The entire data operation consisted of a single web analytics contractor looking at Google Analytics. There was no data warehouse infrastructure in place, and the team was starting from the ground up.

Building the Data Organization

Tristan’s first key hire was what he called at the time a “technical analyst” – a role that's comparable to an analytics engineer today. This person:

→ Set up their BigQuery instance 

→ Built the first data warehouse 

→ Set up Domo as data visualization tool

His next crucial hire was a data solutions architect who: 

→ Implemented tag management tools 

→ Set up product analytics tracking 

→ Made their analytics accuracy “much closer to the truth and much more reliable”

The team continued to grow, eventually incorporating conversion rate optimization specialists under the data team’s umbrella. By 2022, the team had expanded to 7-8 people, operating within PHDV’s structure of about 200 employees.

Driving Business Impact Through Experimentation

The experimentation program became one of the team’s most significant achievements. They went from doing no experiments at all to building “probably something like one or two new experiments per week,” creating a complete process for handling idea intake, execution, and iteration.

A standout example was their add-to-basket experiment. They noticed that many customers didn’t want to customize their pizzas but were forced through a customization flow. The original design was aesthetically pleasing but created friction.

Their solution?

Add a simple “add to basket” button that bypassed customization. While the variant was “kind of ugly” with its chunky call-to-action button, it significantly improved the customer experience and projected to generate an additional million pounds in revenue when extrapolated over a year.

10x your data team’s impact on the business

Tristan’s experience is very similar to my own: If you own experimentation and conversion rate optimization as a data team, your impact on the business will rise dramatically - and your “perceived impact” (as perceived by C-Level leadership) will probably 10x.

When I was Head of Business Analytics at Rocket Internet, the world’s largest venture builder, my team took over conversion rate optimization and website personalization.

We fundamentally changed the way experiments were designed across Rocket’s portfolio company.

One key change was to stop optimizing on conversion rate and instead push for profit contribution per visitor which allowed us to see the impact of experiments across sessions and on the bottom line.

The visibility of our team increased massively after taking over experimentation and I would not start any data leadership role if I don’t own this area.

Advice for Future Data Leaders

Tristan’s advice for data leaders is clear and practical: “Care about what the business cares about.” He recommends: 

→ Spend time with senior stakeholders 

→ Ask what keeps them up at night 

→ Focus on preventing threats and enabling opportunities 

→ Build a culture where data is trusted and valued

As Tristan puts it, “Technology is downstream from everything.” The priority order should be:

  1. Create a culture where data is useful and trusted
  2. Own experimentation
  3. Develop a clear data strategy
  4. Only then focus on technology implementation

This approach ensures that data teams remain focused on delivering real business value while building strong stakeholder relationships.

Want to join us?

Join 500+ future data leaders for tips, strategies, and resources to build impactful data teams and live a better life 🏝

Error. Your form has not been submittedEmoji
This is what the server says:
There must be an @ at the beginning.
I will retry
Reply
Emoji icon 1f64c.svg

Whenever you need me, here's how I can help you:

Data Action Mentor Masterclass : 🏭 From dashboard factory to strategic partner♟️

A digital, self-paced masterclass for experienced data professionals who want to work on high-leverage projects (not just dashboards). 📈

Knowledge Base

Free content to help you on your journey to create massive business impact with your data team and become a trusted and strategic partner of your stakeholders and your CEO.

​10X Data Team Collective 🦸​ 

We build 10X, AI-first data teams. Together.

A curated community for ambitious data leaders who generate outsized business impact (and outsized career growth) by building the AI-powered 10X data team of the future. For the price of less than $1 per day.

You'll get expert content, hype-free conversations, and curated 1:1 matchmaking with forward-thinking professionals in the data & AI space.