It's true that companies like Spotify, Amazon, and Netflix have set an incredible bar for customer experiences with AI. But unlike the past, when lofty advertising goals, could only be achieved by mammoth companies with deep pockets, today, marketing departments of all sizes can rely on the input and efficiency of machine-driven AI applications.

Thanks to the increasing availability of AI to smaller businesses, marketers everywhere can now engage with each customer in meaningful and personalized ways.

This doesn’t mean humans rely on AI to do all the work. Instead, the performance of AI-based algorithms does best when reviewed, sanitized, and operated by real people.

With AI-based programming, new marketing platforms are appearing almost daily. Most of them can be classified into the following categories: vision, language, insight extraction, and anticipatory predictions.

Vision-driven AI applications

AI technology can do a lot for physical stores, especially with analysis and analytics. For instance, machine-learning algorithms can identify behavioral patterns of repeat customers. They can use facial recognition technology based on CCTV footage inside stores and see patterns of efficiency in various product layouts. These can then be leveraged to optimize store design and operations.

But visual-driven AI isn’t confined to in-store experiences. Vision-based AI can also be used to recognize license plates for passing cars. Since every license plate is registered to an individual, retailers can partner with third-party data collectors to get more information on the owner of a vehicle. For brick-and-mortar establishments, this allows them greater insight into potential customers who drive past their place of business on a daily basis. These companies can then target marketing communications specifically to that vehicle owner, with a message designed to draw them in the next time they're driving by. However, retailers must be aware and compliant with customer privacy when applying both in and out of store AI technologies.

Language-driven AI technology

When it comes to the speech-enabled side of AI technology, Alexa, Siri, and Watson from Amazon, Apple, and IBM, respectively, are undoubtedly the leaders. They demonstrate the maturation of language understanding and processing for other elements of consumer interaction  In the past, communicating with customers was restricted to email or other far less personal experiences. Thanks to the new frontier of language technology, customers can now literally speak to their device and end up interacting with a business or brand. AI speech technology is bringing a new level of personalization to customer service and outreach. 

For more direct and natural conversations with customers, chatbots and conversational UIs are providing a great deal of opportunities. Successful adoption of language-driven AI technology can be used to support and even partially replace call-centers for the customer service needs of companies across industry verticals.

Even more, AI can now automatically detect sentiments, meaning certain degrees of frustration apparent in a customer’s voice is recognized by the technology. This is extremely helpful for companies, because the technology understands when it's appropriate to  transfer the customer from automated voice systems to human representatives.

Another application of language-driven AI already being leveraged is the generation of marketing messages using machine learning software. Communication templates based on general vocabularies are integrated with customer preferences. These and observed behaviors enable systems to interact with customers through the right channels, at the right times, in the right tone, and with the most relevant content.

For example, ZenDesk implemented a machine learning strategy to bring down the costs of their PPC campaign. Utilizing an advanced social media engagement software, they were able to compile a list of contacts based on their behavioral patterns. Then, they segmented them into personas in order to best target those ready to purchase their product. ZenDesk claims that this earned them 4 times greater lead generation volume and reduced their cost-per-lead.

Insight Extraction

AI developments have made the analysis and mining of big data possible in order to extract actionable insights. The three common uses for this type of AI are programmatic advertising, lookalike audience modeling, and algorithmic real-time personalization.

Programmatic advertising allows marketers to optimize decision-making strategies. When it comes to the purchase of advertising space in relation to audiences, demographics, and keywords to target, there are various strategic ways AI technology can be used. This is considered a must-have for businesses wishing to optimize their online media spend and campaign performance.

Lookalike modelling is often part of data management platform toolkits. It allows businesses to collate first, second, and third-party data to determine and manage their target segments and consolidate user profile information.

Algorithmic real-time personalization today is usually driven by manually-created rules that look for particular contextual data points. Details like user location, customer status, or estimated household income are all available as areas of focus. The AI program then delivers content based on an assessment of relevance to the campaign’s goals. This can include personalizing websites for a user’s browsing session and dynamic offers of discounts based on established probability models.

Decision-making AI applications

Product recommendations in the e-commerce space have been around for some time. However, AI-powered engines are able to avoid cold-start problems by considering a broader set of data. They analyze customer purchase data along with third party information on relevant associated topics. 

For instance, Infinite Analytics is one such platform that identifies and finds products straight from pictures. With their Infinite Search feature, once a customer snaps an image from their mobile device or shares a picture from a social media feed, the software is able to provide recommendations of similar offerings from your own brand's catalog of products. Additionally, they can personalize online shopping websites, and communicate search results through Amazon Echo.

Predictive analytics and anticipatory design essentially provides predictions about the future. It thereby expands the traditional digital analytics approach from reporting on historical data to predicting and alerting marketers of likely business-critical moments. Predictions are based on mountains of data points spanning from general market trends to personal data.

AI technology is no longer an abstract toolkit for global tech companies. Today, the tools previously only available to enterprise-level companies have become affordable and accessible to medium and small businesses too. If you’re a marketing leader and not already leveraging AI technology for your day-to-day operations, you should start identifying and prioritizing areas within the company that these programs can be implemented.