0 upvotes 0 discussions

Jay Speidell is a featured contributor to The Impact of AI on Modern Marketing: 50 Categories Ranked, 70 Experts Sound Off report.

Here's the complete contribution:

Artificial intelligence is going to continue to make marketing campaigns more targeted and efficient, and empower marketers with powerful predictive insights. Machine learning approaches like reinforcement learning can outperform A/B testing by choosing the best ads while still gathering data. Supervised machine learning can take data (like customer attributes and action history) and learn how it influences outputs (such as whether they purchased an item or how much they spent). This is a process called “training” and can produce powerful predictive models, such as Zillow’s extremely accurate sale price predictions. Unsupervised learning can create customer segmentations by analyzing distances between their data points in a multi-dimensional space that’s difficult for a human to visualize. Another unsupervised technique is called market basket optimization, which is what Amazon uses to recommend similar items that you might be interested in buying.

These techniques are already being implemented effectively and driving huge marketing and sales successes, but a new technique is emerging that will take marketing AI to the next level. This technique is called deep learning, and it is a sub-field of machine learning that effectively learns algorithms from scratch. A deep learning network consists of layers of neurons that can extract new features as data passes through them. This approach can solve much more complex problems, such as learning to drive a car based on video input, and is a driving force behind the future of marketing automation. From interactive chat bots that can converse naturally to deciding what strategy would work best to engage an individual lead, deep learning will enable companies to achieve the equivalent of personal, one-to-one interaction with a large number of people.

Today’s marketers need to be studying this technology. Even though they may not be writing code, a basic understanding of what goes on under the hood will empower future marketers to effectively integrate artificial intelligence into their strategy.

What's Next?