The Role of Personalization in US Consumer Engagement

Personalisation has become the norm for digital interactions. Seventy-one percent of consumers now want companies to provide personalized interactions. Most firms have sought to meet this expectation by investing in this competence, but there are indications that it is not being met. According to a Twilio Segment survey, 85% of businesses believe they provide personalized experiences, while just 60% of consumers agree. The technology that enables customization is always evolving. At FT Strategies, we think that all firms should invest in the ongoing development of this capability. In this essay, we look at the technological maturity curve for customization. We believe that all of the strategies outlined here are valuable, regardless matter how mature your technology capabilities are. Starting out: Active Personalizatio Active personalization is a simple method to provide customers a more engaging experience. Active customisation occurs when users contribute input, resulting in a personalized experience. For example, an app or service may request that you identify or share your location in exchange for obtaining content personalized to that information.

MyFT is one of the publishing industry's most effective examples of active customisation.

This Financial Times product feature enables subscribers to track subjects of interest. Readers can access a personalized portion of the FT website and applications based on their input. They also receive daily email digests based on their specified preferences. Even this simple application of customization can have a significant business impact. MyFT remains the FT's primary engagement engine. The feature increased an individual's RFV score by an average of 86% when compared to a control group (for more details on RFV, see our blog post How the FT moved from no digital presence to over a million paying digital subscribers). Other publishers have experienced similar results. Active customisation increased interaction by 60% at Gannett and 23% at Mediahuis. MyFT, the largest engagement driver at the Financial Times. Source: MyFT, the largest engagement generator at the Financial Times. Active personalisation is also an effective complement to more advanced forms of personalisation. This strategy is increasingly being employed in onboarding flows. For example, when you join up for Spotify, you are prompted to select from a choice of musicians. This information enables Spotify to immediately give personalized playlists and recommendations. Spotify gradually accumulates listening data to support more advanced 'passive personalization', which should result in more accurate recommendations. Spotify's onboarding experience includes active personalization. Source: The App Fuel - Spotify's onboarding flow offers active personalization

Intermediate: passive personalization based on segments.

This method eliminates the hassle of active personalization by predicting client preferences based on usage data. Passive personalization uses an alternative technique, needing no customer interaction. Instead, a customer's experience is personalized based on conclusions drawn from actual behavioral and interaction data. Whereas active personalisation is generally simple (e.g., by tagging content and presenting it based on stored user preferences), passive personalisation usually necessitates more extensive modelling. When developing their personalised content recommendations function, the South China Morning Post (SCMP) investigated a number of different methods. SCMP's Vice President of Data, Romain Rouquier, has described the distinctions between the primary options they considered: Content-based filtering (assumption: if a person reads a given article, they are likely to like similar stories) Demographic-based filtering (assumption: individuals with similar demographics are likely to prefer comparable content) Collaborative filtering (assumption: readers with similar reading habits are likely to love similar content). Each of these models has advantages and disadvantages. For example, using content-based filtering, you can't assume that just because someone engaged with a piece of information means they enjoyed it. The second model is based on data that may be difficult to get (for example, if someone uses a VPN, their location cannot be precisely determined). Finally, while collaborative filtering outperforms the other two methods, it requires considerable browser history data from a userbase. SCMP, like other successful implementations of passive personalisation, has taken a hybrid approach that draws on the characteristics of all three models.

This modeling enabled SCMP to identify over 20 user segments. Rouquier adds that these categories refer to key user personas, which include

China watchers interested in in-depth policy and economic research. Lifestyle aficionados seeking fashion and wellness information. Each segment has a completely different experience with SCMP, from the information they see to their conversion journey. With 'Lifestyle aficionados', the publication focuses on increasing serendipity through recommendations. 'Hong Kong readers' have a more immediate need for news, thus they are encouraged to subscribe far earlier in their connection with SCMP. Hyperpersonalisation is often built on hybrid datasets, similar to the approach indicated by SCMP above. However, it goes far further in distinguishing between individuals, treating each as a'segment-of-one'. To accomplish this, it employs advanced machine learning algorithms rather than the more straightforward rules-based approach. This has the ability to give each user a totally unique experience throughout the consumer journey, from marketing and pricing to customer service and loyalty. Netflix, Spotify, and Amazon are regarded as pioneers of hyperpersonalization. Each of these services places a high value on personalization as an essential component of the client experience. For example, each Amazon user has a unique homepage. In addition to bespoke widgets such as 'Keep shopping for', 'Buy again', and 'Pick up where you left off', each user receives personalized advertising banners depending on the services and items they have previously used, purchased, or clicked. Even the primary menu adjusts to you as a person.