For generations consumer-facing brands have been perfecting their in-store customer experiences. This entails everything from retail store design (did you know 90% of customers turn right upon first entering a store?), to hiring knowledgeable salespeople and training them on the art of the cross-sale, to playing the right music (FYI: slow music is more likely to lead to sales), and much more. All of these lessons were learned from experimentation, measurement, and adoption of best practices.
The same care can and should be applied to the online customer experience — the question is how?
Until recently, it was difficult if not impossible to replicate the personalized experience consumers typically found in stores. According to a June 2017 eMarketer report, 52% of shoppers, just a short time ago, had difficulty finding the exact product they sought online. But today, AI-powered recommendation engines are solving for this and transforming the way brands interact with online customers. Recommendation engines analyze vast amounts of customer data to predict a customer’s interests, and surface the right items, content, or messaging on the brand’s website, app, or other platforms.
And companies are hungry for this technology. According to data from 451 Research’s Voice of the Connected User Landscape 2017 Corporate Mobility and Digital Transformation survey, 82% of businesses say that machine learning for automated contextual recommendations is important to creating customer experiences.
Data-driven personalized recommendations are directly
impactful for a company’s bottom line: they increase conversions for customers
considering a purchase and drive revenue by increasing average order value. They
also increase customer engagement and improve the online shopping experience.
Customers receive a consistent experience across all channels — for example, if
they search for an item on their phone, they might receive an email with a
relevant promotion or see the item pop up again when they visit the company’s
website on a desktop computer.
Most recommendation engines in use today are not personalized – meaning they do not take each individual’s online behavior into account when making a recommendation. Instead they use a technique called “basket analysis” to see which products are typically purchased together by other customers. This is the technique that yielded the apocryphal example of beer an diapers together on Friday nights.
Unlike basket analysis, personalized recommendations look at an individual’s behavior across multiple platforms and devices, in addition to the tendencies of other customers. However, this can be difficult – many companies struggle with processing the volume and complexity of the data coming in from multiple sources simultaneously. This is where Syntasa’s platform really shines.
Syntasa ingests behavioral (or clickstream) data from Adobe or Google Analytics on an ongoing basis. The platform synthesizes all datasets continuously for the most up-to-date unified data, which reflects both recent and historical interactions.
Syntasa’s recommendation engine then generates
recommendations in real-time. These recommendations can be delivered in two
- A brand’s website can make dynamic calls (via REST API) to get recommended content, messaging, and/or products; this works well for website and app personalization where personalized recommendations are needed for each individual visitor as they interact
- Syntasa’s engine can generate recommendations as a batch and send it to an activation end point; this works well for list-driven activities, such as generating an audience for advertising, email, or app notifications.
Electronic components company RS Components used Syntasa to build a recommendation model identifying which customers are likely to purchase their professional line of products, resulting in a £50,000 per month increase in revenue. Electrical and mobile retailer Dixons Carphone saw a 300% uplift in attachment rate to shopping baskets after they implemented personalized, AI-driven product recommendations with Syntasa.
For these brands and others, the initial uptick in purchases and revenue is just the beginning. Syntasa’s engine can continue to hone recommendations over time, developing strategies to engage new and returning customers with the most effective messaging, content, and product recommendations possible. And because Syntasa is a platform (not a SaaS point solution) it can support additional use cases, such as customer journey optimization, subscriber churn prediction, fraud detection, audience segmentation, algorithmic attribution, call center reflection, and more.
Next, discover what AI ASSISTED MERCHANDISING can do for you