SAN FRANCISCO - The complex interactions among various pieces of marketing technology software might lead you to believe that there are very few roles left for humans.

But some of the top people in the field outlined why they think that presumption is premature on opening day of MarTech: The Marketing Tech Conference here.

No matter what marketing technologies you use — and yes, you could be using plenty – you still need strategy and the nuance that comes from thoughtful decision making about ethics, goals and other concepts that go beyond mere data and analytics. Companies need to tap into what we’ve learned about science and human behavior to craft large-scale strategies.

‘Natural Laws of Strategy’

Gord Hotchkiss is the author of The BuyerSphere Project, which explains how how business buys from business in a digital marketplace. His work with companies is primarily about the need for an overarching strategy that is flexible for the things that can’t be control. He used an example of a straight line, which rarely exists in nature.

Real-life often comes in waves and unpredictable cycles, and any solid business strategy must be flexible enough to adapt.

“We get too focused on tactics and we don't focus enough time on strategy,” he said. “We need to start looking at the sciences, which shows we have an illusion of control and need to rethink what strategy looks like. You can't ignore nature forever.”

Since the goal is still a good customer experience, any analysis of metrics and user data must be brought back into a core strategy.

Moving Toward Cognitive Marketing

Putting the mind to work in another matter can work too. Gerry Murray, a research manager for IDC, has coined a term “cognitive marketing.”

Learning Opportunities

While it may sound like another buzz term that’s yet to catch on, the idea has a lot more depth behind it.

“Machines can handle large-scale interactions with humans,” he said. “But there are still particular nuances that computers haven’t mastered yet.”

As an example, he said the planning for how machine learning and higher-level computing is deployed must still be undertaken by humans, who still surpass computers when it comes to anticipating behavior. He cited an example of chatting with an online help specialist, who he was unable to determine at first if it was a bot.

However, with a couple of simple questions it was clear a real person was on the other side. By asking those same questions to a chatbot, the results were wildly off topic. The lesson isn’t that automation and large-scale learning isn’t helpful, but that’s it must be considered as part of a strategy that takes human behavior into account.

Given the large number of companies who are reaching into marketing technology, using cognition and strategy are becoming ever more important as one’s software stack becomes increasingly multi-layered.