The art of innovation: How Doug Sutcliffe is redefining agile tech development for AI at L’Oréal

In this interview, we explore Doug’s journey from entrepreneur to trailblazer, discussing the challenges, inspirations, and successes he’s encountered along the way. With a focus on harnessing artificial intelligence (AI) and machine learning (ML), Doug shares insights into driving innovation, fostering leadership, and preparing for a future where technology and beauty converge seamlessly.

How have your entrepreneurial experiences shaped your approach to data-driven marketing and product management?

My journey into product management starts with my entrepreneurial background. Since I was young, I’ve always had an entrepreneurial mindset. From selling cookies and renovating houses, all the way through to college when I started my first businesses—including a digital marketing company, a 40k member ride-share organization, and a 25 employee house painting business.

These experiences showed me that you can achieve results by approaching challenges from multiple angles and taking ownership of outcomes.

What were the first obstacles you faced while integrating technology in a corporate environment?

At L’Oréal, I’ve had the opportunity to apply my passion for and knowledge of technology. Early on, I identified a gap in how we were using data in our sales processes.

Sales reps didn’t have the right data to make informed decisions, and the data they did have was not accessible or actionable. So in my second month, I took it upon myself to bridge that gap by building predictive tools that helped reps recommend the ideal assortment and inventory level for each individual store.

This initiative wasn’t just about building a tool; it was about understanding the human challenge behind it and using technology as a tool to solve that challenge. I needed to create the tool for the sales reps, as well as tools and processes to help business leaders document promotional strategies in ways the rep-facing predictive model could understand.

After successfully applying this approach within my region and generating significant revenue uplift, I gained the support to scale it across Canada. This experience solidified my path towards product management, showing me the power of combining technological solutions with an entrepreneurial approach to solve business challenges at scale.

That’s when I realized that product management and data-driven marketing were where I could make the most impact—leveraging technology to solve complex problems. By focusing on both availability and actionability of data, I drive decisions that lead to tangible outcomes across a large scope of business.

How significant is mentorship in your leadership approach? How is this method effective in driving innovation within your team?

I feel that mentorship is an essential element of effective leadership. While investing in technology can improve the efficiency of teams, turning that efficiency into innovation and business results depends on highly developed team members.

This is because anything you automate or build with technology accelerates your own viewpoint. But when you teach the skill set of strategic innovation and problem-solving to other people with completely different perspectives, that variety of viewpoints opens new doors to even better results.

In my teams, I’ve been fortunate to always have people from different countries, educational backgrounds, genders, and so on. By enabling them with the skills to create and innovate, we can leverage our complementary skill sets to achieve more together than any one of us could alone.

What frameworks do you employ to develop your team’s dual lens of technical skills and broader strategic vision?

Technologists are amazing when it comes to detail and recognizing potential risks and gaps, but they are not trained to see the bigger picture. One of the challenges they face is the inability to see why we’re doing something and decide what problem we should be solving in the first place.

I’ve developed an approach for training these really smart team members to also think at a macro scale. By combining the macro and the micro, you get hyper-effective innovators.

This involves a lot of workshopping, providing guidelines on how to approach problem breakdown, and showing by example that it’s possible to achieve success even when there is no clear path.

Solving problems is not easy. Being persistent, determined, and willing to inform yourself about all the different domains involved in the challenge is key to finding the solution.

Really, it’s about providing your team with frameworks and structures to help them achieve solutions, and then stepping back to ensure that their own perspective can create an array of solutions different from what I would have come up with on my own.

Can you provide a specific example of a transformative program or initiative that you spearheaded at L’Oreal, detailing how it was developed, the direct impact it had on improving operational efficiency, and the broader benefits it delivered to the company?

The 30 brands I collaborated with across L’Oreal often sought my expertise for their data systems needs, recognizing my ability to accelerate their projects and improve outcomes. While helping these teams launch their projects, I frequently encountered significant inefficiencies in their approaches to tasks that consumed a large portion of their time and wanted to help improve their efficiency, so that they could focus on the creative and inspiring marketing work they were hired for.

Seeing that the largest category of inefficiencies related to data management, I designed an enhanced Excel training to help give them the tools they needed to speed up their tasks, allowing more time for strategic marketing activities. I went on to build several reusable training tools and train over 500 employees across Canada. The outcome was a notable increase in productivity, with staff reporting up to 10 hours saved each week, equating to over $6.2M / year across the organization and proving the effectiveness of the course content.

How does questioning a task unlock new paths to value creation?

Questioning the underlying purpose of a task encourages a deeper dive into the core objectives. For example, when asked to determine the size of our East Asian audience, instead of providing just a number, I sought to understand the underlying goals driving this request. At face value, it seems straightforward, but it’s the deeper ‘why’ behind these questions that can unlock real value. I decided to dig deeper; to thoroughly understand what the brand was aiming to achieve with this data.

In doing so, I uncovered that the one audience was being used for a range of purposes measuring the success of diversity programs to identify potential customers with a high likelihood of interest in skincare products. This revelation was a game-changer. It shifted my focus from a wide-net demographic approach to an AI strategy that was both more accurate and more inclusive of all ethnicities.

Next, I pivoted from traditional demographic segmentation to analyzing purchasing behaviors, engagement metrics, and responses to previous campaigns. This wasn’t just to identify statistically common interests across groups, but to understand the actions and preferences of each individual. This approach allowed us to uncover patterns that defined the top skincare buyers, transcending demographic lines.

This strategic shift not only aligned more closely with our brand’s goals but also led to a more effective marketing strategy. I wasn’t just throwing out broad nets based on demographics; I was fishing with precision, targeting those who showed a genuine interest and engagement with skincare products. It’s a classic case of quality over quantity, and the results spoke volumes.

Preparing your team for the integration of AI and ML requires a foundational education before practical application. How do you structure this training process?

Preparing a team for AI and ML starts long before the dive into the technologies. It’s more about creating a foundation where innovation is a natural extension of the team’s day-to-day activities. For me, it’s about fostering a culture of continuous learning and curiosity.

I really focus on demystifying AI and ML. There’s a lot of hype and sometimes anxiety around these terms, so I start with education—not just technical training, but discussions about what AI and ML are, how they’re being used in our industry, and the potential they hold. This helps in setting the stage and aligning expectations.

What methodologies and practices do you employ to transition from theoretical knowledge to practical application?

To use AI and ML, it’s crucial to build a solid data foundation. AI and ML are only as good as the data they’re built on, so ensuring our data is clean, organized, and accessible is a priority. I work closely with data teams to ensure that our infrastructure supports current needs and is scalable for future AI and ML projects.

A key aspect is creating an environment for safe experimentation. This means using modular design, canonical data structure, and strong data governance. These three elements simplify development, as teams can jump straight to building a great model rather than spending significant amounts of time preparing the data.

Likewise, when good solutions are discovered, consistency across systems and business units makes scaling an almost trivial effort—or at least reduces it to a minor testing task. Together, these efficiencies enable you to minimize the cost of failed tests while maximizing the impact of successful ones, mitigating the fear of failure and encouraging the team to embrace AI and ML as powerful tools for innovation.

By pairing this technological foundation with a fail-fast, learn-fast culture, built by rewarding teams based on shots taken and learnings gained, you can create a constant flow of product innovation. Not every AI or ML initiative is going to be a home run, and that’s okay. What’s important is that we learn from each experiment and iteration, applying those lessons to continuously improve our approach.

What process and strategies have you implemented to enhance collaboration between teams?

AI and ML projects often require cross-functional teams to work together. At L’Oréal, I have achieved this by converting teams organized by skillset into teams organized by deliverable.

I identified that each new AI tool was being launched through the coordinated effort of 5-7 different teams, each of which had to plan the work around their other monthly priorities. For fear of blocking capacity too early and wasting it on delays, each team would wait til the following month’s sprint to start their work. This could result in a 5-day project stretching out over 5 months.

I brought together the team leaders and led them through a transition of deliverable-based team structures, including budget allocation, vendor renegotiations, and process changes. In the end, this project accelerated AI launches by 10x, while reducing costs by 40%—enabling teams to launch over 200 innovative AI solutions like augmented reality (AR) makeup try-ons in a single year!

What are the critical structural and strategic considerations for building a sustainable AI organization that remains at the forefront of innovation?

Everyone seems to look for the flashy, ‘sexy’ AI projects. But in my experience, the less glamorous work of building system structures is where the real value lies.

It’s not the individual AI technologies that drive financial success; it’s the productivity, efficiency, and systems that support technology execution. We developed this really cool AI tech, sure, but the much larger value comes from the system structure that allows us to launch hundreds of AI tools each year.

It’s all about crafting a repeatable structure, honing in on targeted goals, and creating data flows that are both reusable and repeatable. Across industries, this approach has proven to not only enhance the efficiency of AI development and deployment, but also significantly improves the retention of top-quality data scientists. This is crucial because a large part of a data scientist’s job can end up being data prep and engineering, rather than the data science they were trained for and find exciting.

By streamlining the prep work—the data structuring and engineering—I enable our data scientists to focus on what they love: innovating and pushing the boundaries of what AI can achieve. It’s like being a chef in a kitchen; if you’ve got a sous chef taking care of all the prep work, washing, and cutting ingredients, you’re free to focus on creating beautiful, flavorful dishes. Highly trained chefs, or in our case, data scientists, don’t want to be bogged down by prep work.

This foundational focus is why I believe building a sustainable AI organization is more impactful in the long term. Sure, I can build you cutting-edge tech today, but without the right structure, it’ll be outdated in six months. The question then becomes, ‘How do you build a type of structure within which you will always be on the cutting edge?’ That’s the key to sustainability in the fast-evolving world of AI.