Although two thirds of advertising spend is digital, companies still struggle to deliver on the promises of digital advertising and there’s a massive opportunity for marketers to use data and technology to drive digital transformation and innovation. The issue, however, does not lie with technology itself. While technology can be a driver of transformation and innovation, neither can happen through technology alone. Yes, there’s enough technology to solve almost any organizational challenge. But transformation is an organizational and cultural challenge.

How do marketing organizations make technology an integral part of their business processes, value creation, and innovation?


Firstly, marketing leaders need to promote agile mindsets, frameworks, and iterative ways of working. This empowers and enables their teams to translate business problems into smaller parts for which they can design and run experiments that can be enhanced through machine learning and algorithms. Treating experiments as components of a larger system enables fast and incremental value realization. It helps us understand how our learnings will scale and feed into the wider objectives. With the right processes and feedback loops in place, teams can learn from what didn’t work and scale what did work. This kind of approach doesn’t necessarily mean engaging in complex implementation. In fact, it most often is about doing simple things that create incremental value, and then iterating, improving and scaling. By addressing the smaller component issues and optimizing, an organization will solve its larger challenges.


An international opticians chain was delivering digital ads to consumers to generate eye test appointments, yet their shops often didn’t have near-time appointment slots available. Frustrated, potential customers would abandon the page without completing a booking. The client was wasting that ad spend. The hypothesis was clear. If we could, in real time, decrease demand where there were few available appointments and increase it where there were plenty, we would improve bookings and avoid much of the wasted ad spend. We started with a small experiment. We matched two data sets that were easy to capture: shops’ geographic locations and their available appointments. We used that data to forecast availability by location. Next we created an algorithm to define ad targeting parameters —  the most effective audience segments, creative elements, and pacing — and then used APIs to signal ad buys for those. Finally, we added machine learning  to steer ads in real time to locations with high appointment availability and away from shops with low capacity. And it worked. The retailer increased bookings and revenue by 15% without increasing media spend. From there, we scaled to more platforms across more markets, improving on our success.

The result? Our work increased retail bookings and revenue by 15% without increasing media spend.


The optician example demonstrates the simple methodology that I’d like to suggest: Design and launch a test. Assess what works. Then do more of that – and most importantly build an organizational muscle memory that makes this your operating model for problem solving. I strongly recommend that you act rather than spend a lot of time in strategic planning sessions. Stay away from the types of 100-slide PowerPoint presentations that are all too common. Instead, quickly launch a prototype. It also helps to be straightforward and use experiment methodologies that create trust and accountability beyond the marketing team.  

Design and launch a test. Assess what works. Then do more of that.


Marketers often create attribution models to assess each of the ad channels used. For example, we might take data for 100 sales and say that 30 of them came from search, 30 from social media, and 40 from banner advertising. That methodology works for channel optimization. But if we use that data to try to prove the effectiveness of digital media in driving purchases, a CFO will be rightfully doubtful that marketing led to all 100 sales. We need to instead design experiments that, in a statistically valid method, measure the incrementality and in return provide more real accountability and trust. We must focus on learnings gained from real experiments and show incremental business outcomes.


The idea is simple. Of course implementation will take focus, consistency and effort, but once this mindset becomes ingrained in your organization, you will have created a culture that enables you to continually iterate, learn, and rapidly pivot towards what releases the most value back into your business. Most importantly it will create an organizational readiness for the next transformation ahead of you.