Skip to content Skip to sidebar Skip to footer

Source: Forbes as of 02-03-2020

The marketing industry currently finds itself at a crossroads. On the one hand, you have an industry that prides itself on its creativity and the ability to come up with surprising and innovative ways to market products. On the other hand, whether you realize it or not, it is an industry that is increasingly technology-driven, relying on the latest in artificial intelligence and deep learning to reach consumers as many times and in as many different ways as possible. Deep learning has already changed how marketing operates in ways both obvious and subtle and has transformed how brands interact with consumers as well as how consumers relate to brands. It has allowed brands (or their algorithms) to gain a more complete understanding of how their customers think, react and purchase, while also allowing for the complete overhaul of internal organizational structures.

In and of itself, deep learning is a mechanism that allows for prediction and analysis on a vast scale. Deep learning can be implemented in many different ways within an organization, particularly with regards to marketing. For example, one of the biggest questions that marketers have to answer is who exactly their target audience is. By looking through purchase data, engagement metrics and so on, a deep learning algorithm can pinpoint the characteristics or affinities that make someone more likely to engage with a brand. Moreover, it can analyze consumer shopping habits to determine the conditions under which someone is more likely to make a purchase and the types of items or services that the person prefers to shop for. Deep learning is notorious for being a black box when it comes to insights, but effectively reverse engineering algorithm outputs is not impossible for a strong data science team. Having this ability allows brands to identify many more clusters of people they should be targeting, when and how best to do it, and which products that person is more likely to buy. In other words, deep learning can help marketers do their jobs better than ever before.

In addition, deep learning is a hugely useful tool for any organization looking to optimize their business. For instance, it can be used to try to predict future demand for certain products, which in turn would determine how the budget is allocated between divisions. It might also identify opportunities and partners that are currently underserved, thereby giving brands the opportunity to tap into additional revenue streams.

However, as Karen Hao, a reporter for MIT Technology Reviewnotes, there are potential negative consequences for brands that blindly rely on deep learning to optimize personalized advertisements. In a podcast episode I did with Hao in 2019, she pointed out that the algorithms underpinning Facebook’s advertising platform “started to discriminate [by] showing different people different types of housing or employment opportunities,” which ultimately resulted in the algorithm showing more ads for secretarial jobs to women and more ads for jobs in the lumber industry to men, to give just one example.

The same study also found that ads that contained content stereotypically associated with black users, such as rap, were much more likely to be shown to black audiences, whereas content stereotypically associated with white audiences (country music) was delivered to a largely white audience, despite the fact that the target audiences as set by the advertiser were identical.

While businesses that use Facebook’s advertising platform have to trust the company is doing its best to weed out such harmful stereotypes, they can fully control the algorithms they develop themselves. For marketers looking to deploy algorithms for targeted advertising, it’s advised that they try to work through the possible consequences beforehand by thinking about what the algorithm might learn, what data is used to train it, and what stereotypes it might end up reinforcing, and decide whether the potential outcome is in alignment with the company’s values. Having this level of intent ensures that the algorithm takes into account the nuances of human life and society, and can prevent the propagation of harmful stereotypes.

One of the major advantages of deep learning is its ability to take information from many different sources and process it effectively. Not only is it able to do this on an unparalleled scale, but it is also capable of bringing together disparate types of information — images, audio, app data, clickstream data, location data — in a way that other systems cannot handle. This also means that deep learning is completely customizable depending on an organization’s needs and the data they have on hand. Deep learning is already about to help them do their work to a higher degree of accuracy than ever before, as long as they are paying attention to the opportunities already in market. To put it simply, deep learning is helping marketing take that final step from being an art to a science.

Show CommentsClose Comments

Leave a comment

News ORS © 2020. All Rights Reserved.