The 7-step customer centric big data plan

SAP Start-up Forum by TiE BigData SiG to be hosted by BigInsights CEO
February 10, 2014
Data scientist critical to an organisation’s success, says Teradata CTO
April 30, 2014

While big data has grabbed the attention of organisations worldwide, CMOs and CIOs are struggling to understand the full benefits it has to offer and implement that variety of purported benefits. This, complemented by an array of new technology, techniques and vendors on the big data scene, has left the c-suite with more questions than answers.

However, more than 60 per cent of Australian business leaders stated that “Improving insights about customers” are big data's key advantages in the BigInsights BigData Study 2013 survey report. Big data projects also provides an opportunity to strengthen the CMO/CIO relationship and for the CMO to show leadership across the organisation.

While some of underlying big data technology is new, the basic goals have not changed for marketers. The big data journey starts with what we at BigInsights call a 720-degree view of the customer. This 720-degree view of the customer includes all internal available information complemented with externally available information. It enables companies to gain insights into the likes, dislikes, buying patterns and behaviours of not only the customer, but also their friends and social circle.

The ultimate objective is to hyper personalise communication and products towards whatever is attractive and interesting to each of your customers.

We see seven steps to creating a customer-centric big data plan:

  1. Critical business challenges: First, identify and define the critical business challenges that need to be addressed by the project. It may be helpful to have specific customer-related questions about the customer’s insights as a desired outcome. For example, what are the signs that a customer is not going to renew and what could you do to minimise customer churn?
  2. Data inventory and quality audit: Conduct a stocktake of the customer-related data inventory. Identify data sources and types of data that you currently have within your organisation that could help answer the questions either directly or indirectly. Perform a data quality audit to determine the reliability and quality of the actual data set. This data would be in variety of CRM, data warehouse, customer service and related systems.
  3. External data sets: Research and acquire external data sets which will enhance and enrich the existing customer data sets within your organisation. This may include customer social media sentiments, click stream data from your outsourced website and other sources that may need to be collected.
  4. Analytical tools, models and environment: Create an analytical environment and build analytical models where you can load all these sources of data and start using tools to gain insights from the customer data.
  5. Refinement of hypothesis: Use an iterative process of querying and experimentation to refine the insights gained and any hypothesis in the analytical model.
  6. Test and perfect: Validate the findings by testing on small groups of customer segments before rolling it out to a broader customer base.
  7. Integrate with operational processes: Ensure that the analytical data based environment is integrated with your existing operational process. Revisit existing operational process and re-engineer them if needed to ensure that the value that the data provides is utilised to the maximum. By creating and integrating such a dynamic environment, companies are starting to realising values from big data.

A strong partnership between the CMO and CIO is another imperative to make these projects a success. Building a collaborative and cross functional team that is customer/business savvy, has ‘data science’ skills, traditional IT infrastructure and data integration skills, is critical.

While it is important for early projects to show business value, initially there is the need for experimentation to understand the insight available within the available data and how to best build predictive models based on it. A process similar to the A/B testing and iterative refinement used in modern online marketing.

There may be a temptation to outsource work to digital agencies due to a lack of IT skills in marketing organisations. However, experience has shown successful projects require deep customer/business insights, development and integration effort with internal data sources which if difficult to outsource.

The significant innovation and competitive differentiation big data can bring to the organisation from a tactical and strategic perspective make it hard to ignore. Perhaps it is time to have regular coffee meetings with the CIO.

(This article first appeared in CMO)
You're welcome to share this page: