The biggest reveal at the Oracle Modern Customer Experience (CX) event held in Las Vegas from April 25 to April 27 was the debut of Oracle Adaptive Intelligent apps for CX. Adaptive Intelligent Offers (AI Offers) is the first app in a planned portfolio of artificial intelligence applications that will eventually span CX, ERP, human capital management, and supply chain management.
The introduction was no surprise. Oracle said a CX app would come first when it announced Oracle Adaptive Intelligent apps last September. Since then the company has been refining the behind-the-scenes automated machine learning capabilities. Executives revealed in January that work with beta customers was about to begin.
AI Offers delivers a subset of the total capabilities envisioned in the Oracle Adaptive Intelligent App for CX (see image above). AI Offers is also in limited release. For now, Oracle is working with early adopter customers. Executives from Team Sportia, a Swedish retail chain, were among two early customers appearing at the Modern CX event. Another customer, an international bank, was identified under non-disclosure terms. Broader availability is expected "soon," according to Oracle, but there's no official general release date.
Oracle Adaptive Intelligent apps complement Oracle cloud applications. AI Offers, for example, is an optional (extra cost) add-on to the Oracle Customer Experience Cloud. At launch, AI Offers is integrated with the Commerce and Marketing clouds in the CX suite, but ties will extend to the Sales and Service clouds later this year.
AI Offers taps into first-party data from a company's Commerce Cloud instance, enriches it with third-party data from the Oracle Data Cloud, and then applies Oracle's decision science and machine learning capabilities. The resulting contextual insights into individual customer behaviors and purchase propensities drive best-fit offers as they browse a Commerce Cloud-powered website. The same AI Offers contextual analysis and triggers real-time personalized content recommendations in the Oracle Marketing Cloud. Pricing for AI Offers is based on cost-per-thousand offers or content recommendations -- the same model used to pay for Oracle Data Cloud enrichment.
AI Offers is an advance over older, rules-based technologies (including Oracle's own ATG personalization engine) in that it is both adaptive and real time. AI Offers applies and reapplies a recommendation engine and predictive algorithms with each new click and navigation step. The design point is 100ms to 150ms response time -- on par with state-of-the-art ecommerce and advertising systems. Even in marketing scenarios, fresh email promotions are triggered at open time, reflecting up-to-the-second analyses of the latest transactions and online behaviors.
Oracle executives insists that AI App data-collection and data-usage practices adhere to global privacy laws and principles. Personally identifiable information remains in each company's instance of the Commerce Cloud and is not copied, stored, or shared. Third-party from Oracle Data Cloud is temporarily joined to the first-party data and the combination is then run against a combination of collaborative filtering, online learning, and predictive algorithms at run time to trigger optimized offers and content recommendations.
AI apps will learn over time, and Oracle plans to connect these insights across clouds and domains. For example, AI Offers tracks outcomes in the Commerce Cloud, such as purchases by category, brand, and product, along with aggregated demographics, such as age, gender, marital status, and education level, and psychographics, such as interest in sports, fitness, hobbies, and so on, drawn from Oracle Data Cloud. Oracle plans to bring these aggregated, outcome-based insights into the Marketing Cloud, where they will inform targeting, and thus, smarter content recommendations and promotions. Similarly, once Oracle brings Adaptive Intelligent App capabilities into the sales and service areas of CX -- a move expected later this year -- the company expects to be able to inform next-best sales steps by connecting insights into recent service activities to drive predictive account health insights and predictive recommendation capabilities.
Next steps for Oracle include building up the base of customer success stories, adding connectors for popular third-party commerce systems, and adding domain-specific features for sales and service.
Analysis: No front runner in the AI war (yet)
Adaptive Intelligent Offers is just the first release in a rollout of AI apps that will surely extend into next year and beyond. Constellation expect to see more CX capabilities over the summer and apps in other clouds announced by Oracle Open World in early October.
I like the fact that Oracle has incorporated controls and provisions for business rules in AI Offers so humans can tweak and otherwise override "smart" offers and recommendations, as required by business requirements. Marketing agreements, for example, might dictate that competing brands cannot be shown in the same promotion while overstocks or profitability requirements might need to sway cross-sell and up-sell objectives.
Oracle's closest competitor on the AI front - specifically in the CX arena -- is Salesforce Einstein. Salesforce has already released 20 Einstein "features" and had another 25 planned for release this year. The features are focused and not akin to Oracle's more expansive AI apps. For example, Marketing Cloud Einstein includes a Predictive Content and Product Recommendation feature, while the Commerce Cloud has separate Product Recommendation and Predictive Email features. (The complete list of Einstein features is detailed in my Inside Salesforce Einstein Artificial Intelligence report.)
Oracle Adaptive Intelligent apps include multiple capabilities that roughly compare with Einstein features. The Oracle AI App for CX, for example, will eventually include at least 16 capabilities, as shown in the slide above. In the Marketing and Digital category, for example, there's Intelligent Message and Personalized Open-Time Content. In the Commerce category, there's Personalized Product Recommendations. It's easy to spot similar capabilities in each CX portfolio, but keep in mind that Oracle also plans to apply AI to the ERP, HCM and SCM arenas.
Who will win the great AI war? We have yet to hear in-depth customer testimonials from either Adaptive Intelligent App or Einstein customers who have really put the offerings to the test. Also keep in mind that you can't subscribe to either Adaptive Intelligent apps or Einstein independently; they're both optional add-on services to underlying application clouds. We've all heard examples of the tail wagging the dog, but it remains to be seen whether smarter AI capabilities will lead customers to choose or switch to different core application clouds.
You can read more about Oracle's strategy and broader plans in Inside Oracle Adaptive Intelligent apps.
AI's reach is soaring in 2017, but are businesses ready?