Experimentation for Predictive Optimization

by | Oct 26, 2018

Whether they were hosting a dinner party or sending a wedding gift, the etiquette columnist Emily Post often advised her audience to add “a personal touch.” Emily Post authored most of her famous columns at the turn of the last century, so it might surprise her that the people most in need of her advice today aren’t society hostesses and housewives, but rather modern consumer savvy businesses. Personalization is no longer just an aspirational goal for modern marketers. With each passing day, customized experiences are becoming more common-place, whether they’re in our streaming queues and search results or on the retail floor. These customized experiences have the cumulative effect of raising the bar on consumer expectations. Brands that can’t clear that bar risk not only missing an opportunity but also looking a bit rude by comparison.

In truth, most brands know that personalization across all their channels is critical to success. What’s more, brands have access to more consumer data now than at any point in history. Most digitally progressive brands interact with customers across multiple digital and physical touchpoints, all of which yield signals and customers data points that can be used to fuel a personalization strategy. However, despite this abundance of data and a definite will to succeed, personalization efforts have run into some fairly persistent stumbling blocks. Here are the three biggest barriers preventing brands from fully hitting the mark in personalization:

Organizational Silos

One of the biggest challenges for brands looking to implement fully realized one-to-one personalization is the inability to access data. Recent years have seen many forward-thinking businesses drastically amp up their data collection efforts. As a result, much of the data required to execute a scaled personalization strategy is already available, but the siloed nature of many marketing teams makes it difficult to access let alone build a unified view of the customer.

Today’s marketing teams are often assembled in a piecemeal fashion, over a period of years, to accommodate new channels and objectives. As a result, their data is collected and stored in a commensurately piecemeal fashion, distributed across different systems, tied to specific channels, and often managed by a variety of different teams. This siloing of data makes challenging to deliver consistent messaging across channels which can, in turn, lead to inconsistencies in communication both within teams and between brands and their customers. Since continuity between marketing channels, customer support, and on-site experience is increasingly key to customer acquisition and retention it’s critical for savvy brands to unsilo their data and bridge gaps between teams. This is one of the primary challenges that customer data platforms have emerged to solve. By uniting data from disparate sources a CDP can build a unified customer view of the customer enabling a more integrated approach to marketing.

Scale

While breaking down data silos is a significant step in the right direction, marketers do face other challenges on the road to personalization. Getting data into marketing systems is still tremendously difficult, even today. Many marketing systems were not conceived to handle the scale of data needed to implement a one-to-one personalization strategy and therefore can’t accommodate the scale of enterprise data nor the complex forms in which it is represented. These technical limitations significantly hamper the way brands are able to leverage the data they already own.

Many of these challenges are endemic to businesses managing digital transformation. As data-driven growth strategies become paramount to the survival of businesses it’s important to remember that there’s only so much one can achieve with a point solutions based on cookie-cutter one-size-fits-all strategies. The most effective road to greatness is one that builds towards long-term strategies while also showing immediate ROI.

Technology and Innovation

The final hurdle standing between us and true one-to-one personalization is a little harder to clear. Personalization technology has evolved rapidly in recent years, producing breakthroughs that have made a more unified customer view possible for many brands. But going all the way is going to require another technological leap. Specifically, we’ll need the kind of next-generation AI technology that we’ve just begun to scratch the surface of. From today’s vantage we can see the beginning of the path that will ultimately lead us to the personalization promised land. Just seven years ago deep learning experienced a huge emergence, and today those same algorithms can drive cars and defeat human chess masters. We may be just one major innovation away from a similar AI explosion.

The marketing technology of the future will need to be scalable and flexible enough to account for rapid growth and to anticipate new phases of digital transformation. We expect these next generation solutions will focus more on actionable insights built on a wider range of data inputs including real-time behavioral, historical, product, and deep access to content.

What’s next?

Despite the stumbling blocks, true scalable one-to-one personalization is close at hand. Brands increasingly understand what their organizational and technological challenges are and as a result they’re moving to solve them. By restructuring marketing teams to reflect a more integrated approach to customer interactions the old channel-based silos are starting to come down. Similarly, a new generation of marketing technology has emerged to help scale in-house data operations and manage the ever expanding array of data sources and inputs needed to create a complete unified customer view. Meanwhile, a string of promising advances in AI combined with the huge investment of resources the technology has attracted, promise to soon crack the type of next generation technology that will tie all these threads together.

Ready to get started? Get in touch for a demo