Jul 17, 2020 | 5 min read
In the moment personalization is something that a lot of organizations (end users and vendors alike) talk a lot about, but in reality, very few are able to achieve. For many years, the pioneering Celebrus solution has been relied on by many of the world’s largest and most mature organizations to enable genuine real-time personalization, by virtue of our granular data capture capabilities combined with our millisecond data stream to enterprise decisioning solutions.
In practice any system which captures or generates a large volume of data, must provide a means of filtering this data to enable real-time use cases. Downstream real-time decisioning systems are simply not designed to handle and to process vast quantities of data. These solutions require a feed of only the most relevant interaction data which is focused on the outcomes they were implemented to solve for, so the upstream application must filter out any irrelevant behaviors. For example, a bank who aims to boost personal loan applications through real-time personalization is hosting 1 million customer sessions at a given point in time via its mobile app, but 80% of customers are simply checking their balance or doing routine banking tasks. Of the remaining 20% of visitors, half are engaging with the mortgage calculator and the other half have clicked on a banner promoting personal loans. If the bank connects the data from all 1 million sessions to a decisioning application, that solution must undertake analytics prior to decisioning taking place, which would cause latencies to increase, diminishing their real-time capabilities.
Organizations have for many years been able to configure Celebrus to instantly recognize behavioral patterns which signify opportunity and intent. Our pioneering clients have created predictive models which instantly score customer behavior to ensure that only the most relevant customer data is connected to decisioning systems for the creation of personalized content. But as the maturity of enterprises (and in particular banks) increases, the Celebrus solution has become increasingly popular. As a result, we took the decision to incorporate powerful new machine learning features into the product to reduce our customers’ time-to-value when deploying real-time decisioning. We call this pre-configured functionality Automated Marketing Signals (AMS).
AMS automates the filtering process described above courtesy of out-of-the-box, fully embedded predictive algorithms, which have been pre-configured to detect the 50 behavioral signals which are most relevant to the banking industry, such as interest in products (loans, mortgages, credit cards, etc.), subscriptions to services and changes to personal circumstances which could indicate a heightened propensity to buy. These models can easily be adapted by citizen data scientists to create new signals and of course we will be adding to this list in future releases of Celebrus.
The simple fact is, no other solution can connect highly focused marketing signals to the enterprise decisioning solutions that Celebrus does, and with the launch of AMS, organizations can now deliver in the moment personalization use cases from day one of their Celebrus deployment going live, enabling them to benefit from the dramatic uplifts* our solution delivers at an even earlier stage. In addition, the out-of-the-box functionality of AMS saves significant development costs and frees data science teams to focus their resources on other cash generating activities.
We use necessary cookies to make our site work and analytics cookies to help us improve it. We will not collect any personal identifiable information unless you enable us to do so. For more detailed information about the cookies we use, see our cookies policy & settings page.