Written By

Mark Taylor VP, Customer Insight Group LinkedIn
Bio

With Contributors

Marc Sanford, PhD Director, Customer Insight Group LinkedIn
Bio
Pradeep Ananthapadmanabhan Chief Technology Officer, VivaKi LinkedIn
Email
Bio

Follow the Conversation

#bigdata #orv10

It’s Time for Big Data to Improve Customer Experience

Channel-based marketing is dead. The increased amount of data available at the individual consumer level, combined with the proliferation of cloud computing, have allowed savvy analysts and marketers to create a truly singular view of the consumer, regardless of touch point. This single view enables a truly enhanced consumer experience and more efficient use of client and agency resources for decision making. All the customer data out there is worthless if you can’t process it and turn it into actionable intelligence.

Unfortunately, older data processing technologies (such as Relational Database Management Systems, or RDBMS) are simply not capable of processing data in volumes that the industry has collectively coined “Big Data” — volumes that are in terabytes/petabytes. As such, we position the consumer as the only real appreciating asset and we tie everything together through the use of Big Data.

Awareness of the challenges of a multi-channel world is nothing new, but each channel touch point represents an immense opportunity for insight. An average Razorfish client has billions of customer interactions a year across paid, earned and owned channels. With so many opportunities for insight and learning, we create a 360-degree view of each individual in the database.

Each touch point is an identifiable interaction and an opportunity to build value.

Using integrated Big Data approaches, we are now informing the holistic data view to gain the fullest understanding of consumer interactions, intent and value possible. This current shift centers on how customer intelligence across channels is not just used for insights, but actioned at great velocity to power multi-channel targeting and personalization, made real through dynamic digital messaging. From insight to action, we’re now finally implementing consistent and relevant messaging approaches that provide cohesive consumer experiences.

In our experience, each client using Razorfish’s Big Data-led performance marketing approach takes a different path. Ultimately, a client’s path is based on business priorities and what information can be leveraged from the available tagging and data strategy. Working with different clients has enabled us to determine realistic roadmaps.

Modular approach to platform and services, by fully integrating an organization’s owned, paid and earned channels for insights and targeting.

Holistic integration benefits:

  • Common Data Marketing Platform (DMP) for reporting, analytics, targeting and media integration.
  • A channel and customer view of success.
  • Metrics that measure end to end, not just in parts.
  • Decision-making through actionable insights.
  • A common language for performance across different teams, brands and markets.
  • Better use of the team’s time to focus on what matters most to their business.

Data enables us to understand customers and to manage contact and content strategy. Data is a core component of integrated marketing and, via an integrated approach, we can speak with a single voice across channels and lines of business. However, to succeed in a meaningful way at that level of customer centricity, we have to manage all that data in a way that holistically fuels customer engagement and experiences. That effort requires a whole ecosystem of people, processes and technology.

Even the most sophisticated and modern businesses today are surprisingly ill equipped to manage even the most basic digital marketing standards and activities, let alone jettison forward into the new world of Big Data techniques.

Through a series of in-depth interviews and client experiences, Razorfish found a common set of fundamental challenges holding back meaningful data integration:

  • Fragmentation of efforts between different teams, tools and data sources across multiple channels, brands and regions.
  • Political and fiscal turf protection.
  • Multiple sales funnel constructs.
  • Inability to identify the customer.
  • Inability to quantify the value of customer experiences.

Let’s take a closer look at how our approach to Big Data, using that razorfishOPEN™ framework, can remedy these issues.

Fragmentation

We’re in an era where intelligent use of Big Data pays huge dividends. Implementing solutions that improve integration of data is very challenging and complex but not for the reasons you might think. Much of the technical and analytical challenges for tapping Big Data have been solved — but failures today often stem from attempting to use legacy small data solutions, internal politics, effort fragmentation and failure to manage the true value of Big Data-based solutions. While a lot of niche players using Big Data approaches have stepped up to solve parts of this challenge, building incremental capabilities in a silo can by default push you further into a silo-based culture and limit your understanding of the customer.

Example of a siloed view of data management and reporting.

Any holistic Big Data solution requires a scalable measurement plan and tagging strategy at its foundation so you can take into account performance marketing efforts across channels, tactics and disciplines, with a shared strategy of measurement and tracking that is scalable across international regions and markets.

The end solution provides a subtle and intelligent approach that can evolve by integrating and building upon other assets, data sources and capabilities already in place. This approach enables a modular and organic ability to evolve and grow, but with a standardized core. These qualities are not always the prerequisite in Big Data techniques, but without this there is no foundation for growth.

Turf wars

Crossing organizational units can be tricky. Often clients are not set up internally for a path to success based on complete integration and use of available data. Organizations are formed around channels — one unit owns the Web site and its data, another owns CRM and email, another may own Web media and yet another may be in charge of social media. Worse yet, each silo may have its own analytics arm. The only way to be part of this organizational conversation is to think big. We have gained phenomenal success by leveraging Big Data-based techniques as part of a modular, digital roadmap that directs current and future business investment in the next 100 days/12 months/3 years. Be prepared to think big even while starting small, and determine your starting point and roadmap — no matter how audacious your goals.

A Razorfish global technology client decided their initial priority was to gain cross-channel insights before embarking on targeting and deep analytics. This was the foundation starting point for their organization and it ensured they gained political capital across their business model through an evidence-driven, customer-centric approach that enabled financial modeling of return on investment. Their next phase focus is on actioning that data for targeting across display advertising and the Web site.

Another Razorfish client, a major global retailer, recognized that they had a wealth of underutilized offline and digital data. They decided to leverage Big Data approaches to integrate multiple channels and power media, dynamic re-messaging, analytics and more. The ultimate purpose is to enhance the value of those relationships by aggregating information about the customer and communicating with them in the most relevant and engaging way. Previous iterations of this approach resulted in a three- to five-time increase in return on ad spend, and a significant decrease (about 65-70 percent) in cost per acquisition. Each phase typically pays for itself in weeks, while providing the funding for the next incremental phase. This becomes a sound position to be in when convincing your peers of the rationale and business case to fund such solutions.

razorfishOPEN™ targeting roadmap.

Big Data can be organized without a major disruption or re-architecture of existing structures, internal teams, vendors, agencies, platforms or focus. Instead, our approach to Big Data utilizes an open standard designed to exploit existing assets and fit the best custom solution for business environments. This approach has an evolving set of modular relationships managed as a single solution, resulting in a single and holistic view of the customer based on all available data. Our clients are using this common view to engage and encourage their different teams to speak in the same language.

Multiple sales funnel constructs

Funnel management is where people are getting clever with Big Data, however it runs the risk of solving only one part, rather than the whole. We know that leveraging a single view of the consumer drives value at all levels of the funnel. So why do many continue to approach client problems and challenges as one-offs or focus on just one area of the funnel?

Too many distributed engagements will lead to:

  1. Single point-in-time solutions that require rebuild with every new engagement.
  2. An additional data silo that requires more time and effort to manage and process.

For example, the illustration below shows how an effective re-messaging program will grow the bottom of the funnel. However, if this becomes a one-off without integrated implementation and access to the data, the solution becomes a very clever silo at the expense of the broader opportunity.

Conversion and remarketing is only part of the answer.

The reality is that the rules and the data to enable an integrated view and management of the funnel would need to come from first-party data via a DMP solution and the organization's data assets, rather than a third-party data provider. Third-party data intelligence can provide these larger insights into what’s working and where there’s opportunity for more scale. Data providers can be joined to first-party data, not the other way around.

Within Razorfish’s framework for integrating data and services (described below as razorfishOPEN™), targeted, dynamic ads are combined with a Demand Side Platform (DSP), such as Audience on Demand (AOD), to match impressions to users identified in real time. This allows you to only reach users that have been already “qualified,” and avoid upfront agreements and negotiation by paying the market price for users meeting criteria defined in the audience segmentation. By reaching the right audience at the right price and allowing the ability to control bids at a cookie level provides a great deal of efficiency and relevance. This integration also enables the ability to bring a wide array of data at the bottom of the funnel to the audience at the top of the funnel.

Behavioral data captured by the razorfishOPEN™ first-party DMP on client-owned assets integrates with Media DSPs third-party data to help build more precise audience segments and add to our clients’ audience buying capabilities.

Inability to identify the customer

Razorfish implements a customer-centric approach through an organized framework of measurement and tagging that tracks all digital business activity and harnesses the full stream of data as the core basis of the single view of the customer. From day one we leverage all existing assets, people, agencies and platforms, without a big, disruptive overhaul. Chances are these existing components are there for a reason and are providing value, but getting that cross-functional view and line of sight is the first objective.

razorfishOPEN™ tagging framework that tracks cross-channel business activity.

Inability to quantify the value of customer experiences

Organizations are increasingly demanding faster value return on their marketing investments. Razorfish has found that businesses now more than ever need a true, meaningful understanding of what drives customer value. Rather than using Big Data to improve one area of the customer experience, we need to build toward meaningful interactions at a customer level and progress the value of their brand relationships over time.

We have seen our clients quickly moving toward a culture that understands customer data as one of its most valued assets.

Focusing on customer value helps companies move away from channel performance and toward greater customer-centricity. But to calculate customer value, companies must fully utilize the recency of interactions, along with the required behavior, revenue and relationship metrics. A key challenge businesses struggle with is finding advanced analytical skill sets and analytics-based approaches that can leverage and interpret that data to determine the key levers that drive value within their organization, or at least within a specific team’s control.

We define and value the segments of customers we want to engage with and create Big Data-applied algorithms to create a new type of value segmentation model that can operationalize differentiated experiences at high velocity. We not only deliver unique and consistent experiences to these customers, but also leverage our knowledge about them to bid for the opportunity to deliver those experiences. From the beginning, we value the revenue and other business impact of the opportunity and in the end we prove it.