Share
Related Articles
How Language Can Drive More Bookings and Growth in 2021 for Travel Companies COVID-19 Communications During the Next Phase of Recovery Healthcare Challenges Language Can Help Solve
Newsletter

Lessons in Making AI Work: A Conversation with Ben Blake from Hotels.com

That’s according to Nielsen’s 2018 CMO Report. And they invest 20% of their marketing budgets on creative, as AdAge notes. Budgets grow, but creative remains highly subjective and returns remain poor. CMOs have much to say about their products and services, but they message mostly with guesswork, or “gut marketing.”

This must change, according to Persado co-founder Assaf Baciu and Hotels.com CMO Ben Blake. At an Oct. 2 session at the “Tech. Retail Week” event in London, the duo discussed the need to adopt AI in marketing and abandon the guesswork and subjective marketing methods of old. They also detailed AI adoption challenges and offered advice to successfully adopt AI.

The argument for AI in marketing

The 2010s mark the Age of Intelligence – specifically, machine learning and deep neural networks, deep science, deep analytics, and deep understanding of what drives consumers to interact with brands. People now read and write more than they talk or listen, so words and written language are more important than ever, for marketing and for the entire company since the choice of words impacts everything across the enterprise. Copy and creative can’t be left to the gut.

Brands today tend to communicate awkwardly, as if their backs are turned away from their customers, due to a lack of data-driven understanding of what actually resonates with individual customers, Baciu said. But CMOs can “turn and face their customer” using AI. Only AI can provide the necessary computation power, memory, and algorithms to help marketers choose the right words, adjust the words, and react to what customers respond to. Machines help CMOs face their customers again, get the feedback they need, and propel their brands into the future.

Learn more: How to successfully implement AI marketing technology

AI also complements humans and moves marketers away from the world of guesswork, which is approximate, subjective, and exposed, and into the world of mathematical certainty and accountability. AI applies science to always know the next best marketing message to better engage and win every moment across any customer journey and any channel at scale.

This is the next stage of marketing strategy – doing, and trusting in, what the machine says is best to do.

People now read and write more than they talk or listen, so words and written language are more important than ever.

AI exposes the limits of A/B testing

Organizations that grew during the Internet Age tend to continue to execute a test-and-learn strategy. This includes their marketing efforts, even as scale and speed are more and more critical. The classic way they test is A/B testing, or the champion/challenger model that compares challengers to each other until a champion emerges.

But as the business scales, A/B testing leads only to a narrow track of things that work.

The model eventually brings marketers to “evolutionary dead ends where the champion works only 50% of the time and the organization doesn’t retest for a longer period of time,” Blake said.

AI helps organizations scale much faster and more effectively than the champion/challenger model.

Without AI and machine learning, the champion/challenger A/B testing model equates to testing without memory. The memory comes from understanding what drives impact and what parts of the message resonate in the context of word choice. Without AI and machine learning, the only memory is a human who can maybe recall testing a concept several years earlier and whether it worked.

Machines don’t forget.

Without AI and machine learning, the champion/challenger A/B testing model equates to testing without memory.

Push for executive buy-in

Successful AI adoption requires executive understanding and buy-in. “It’s very easy to sit there and say, ‘We’re a machine learning-first company, now everybody go and make it happen,’” Blake said. “You will stumble on that at every stage, because it is a fundamental shift. That mindset, all the way to the CEO level, of starting to have a strong understanding of AI and having a mental model of what machine learning can do for your organization – rather than just a slogan in the strategy plan – is critical.”

Baciu agreed.

“For any AI to be pervasive and impactful at the company level it must have – and this is not just lip service – executive support,” he said. “The CEO, CMO, CIO…if they are not bought in, it just doesn’t work.”

Even the most experienced executives must learn to let go of their gut instincts and trust AI.

“Sometimes, the ad the machine spits out is not the one that I would have gone for, and I just have to trust the machine,” Blake said. “So, even in those old-school, very creative roles where marketing was all about guts, we’re eliminating that and saying we don’t hire anyone who is guts-only.”

Executive buy-in at JPMorgan Chase

Baciu cited JPMorgan Chase as an example. When North America’s largest bank hired Persado after a pilot for a deeper AI for marketing implementation, the decisions came from the C-level. After bringing Persado onboard, the entire organization lined up to build the processes that needed to change and clean the data that needed to be adjusted to take advantage of its capabilities and put them in the market. AI now generates or influences every marketing message Chase sends.

Organizations must also be prepared to open their systems for the vendors they hire. Baciu said he met recently with a CEO who had cultivated the right mindset. The CEO peppered Baciu with the right questions.

“His first question to us was: ‘Tell me, what do I need to unlock within the company for you to drive your capability? Where do I need to get involved?,’” Baciu said. “And it’s organizational skills, it’s technology stack – the ability to experiment for machine learning is critical because you have the word machine and the word learning.”

See beyond personalization

AI is more than just personalization. Not everything boils down into an audience of one, according to Blake.

“I don’t love the word ‘personalization’ to describe what we’re trying to do,” Blake said. “Personalization is actually a means to an end. What you are trying to do is build relevance for customers. And in many cases, true personalization isn’t necessary there. It’s the standard – the old adage of right person, right message, right time. We’re really trying to go after that.”

“Personalization is actually a means to an end. What you are trying to do is build relevance for customers.”

– Ben Blake, CMO, Hotels.com

Fundamentally, CMOs must shift to redevelop digital marketing creative in a new way to unlock the full power of the message. They must move away from creative developed using analog methods, such as humans and guesswork. These methods put the CMO and marketing team in a position of not really knowing how well the creative performs, making them unable to defend it.

AI for marketing creates measurable and predictable messages that deliver on-brand experiences at scale. And it ensures the CMO and marketing team can be accountable and defend the massive investment in digital marketing creative.