Tanu Javeri is an 18-year digital marketing veteran with deep roots in digital analytics, search engine marketing, and SEO.
She is the Global Digital Analytics Enablement Lead at IBM.
T.J.: In my 18 years’ experience, I have had the opportunity to work with quite a few Fortune 100 companies. Irrespective of the digital analytic platform or digital martech tools, the basic need to solve business questions with data always are the same. Recognizing what decision needs to be made, allows for better data collection and reporting. It is a three-step approach, first being able to visualize data, second - being able to narrate stories with this visualized data, and lastly, turning this data into information that can be utilized to drive better decision making. Invest in developing a value-driving answer culture.
Many companies are still operating with ad hoc initiatives and one-off actions. Technology has enabled us to amass greater and greater amounts of data, and there is an accompanying growing desire to make sense out of all of this data. So this requires modern technological foundations to be placed at scale. Sprawling legacy systems, siloed databases, and sporadic automation are common obstacles. Companies have to embrace machine learning to scan through such large data sets. Automation has to empower the data collection, aggregation and analysis. Manual data processes are inefficient and strain resources. Having interdisciplinary integration allows for easy handling of multiple data points from different sources and identifying patterns.
T.J.: The core of every customer data strategy is the data quality. The level of accuracy, completeness, consistency and reliability of a data makes it the bases of good analysis. As marketer one should learn the process of how and where the data is being collected. Understand your data component parts – where and how is it collected and broken down. Where is it store?
Having an understanding how the technical implementation is done allows you to learn why a certain KPI is tracked and certain data points are not accurate. This allows the marketer to define their data points – historical, relevancy and segmentation. Marketers can then extract insights from the data and embed it back into business processes such detailed customer segments, create personalized experiences at scale and leverage predictive modeling. Storytelling is an art. Be effective, craft captivating data-driven stories. Data scientists too often fall short in articulating what they’ve tracked and observed.
T.J.: The foundation of any robust customer data strategy is the information captured. This is often done in isolated packets. I firmly believe that having interdisciplinary integrations allows for enriched customer data. Ideally, make all data shareable and accessible. In reality, however, this has to be prioritized. Identify the business decisions that you want to take or specific customer outcomes. This can then be leveraged to customer behavioral insights down the stream.
T.J.: I can’t state 3 metrics as they might be different in different business situations. But the 3 types of metrics that I really value are foundation metrics, conversion metrics, and collateral metrics. Very early in any business, it is essential to be able to track and distinguish the importance of foundation metrics and the conversion metrics.
During one of my SEO projects, I found out about the importance of collateral metrics. My team was working towards increasing Search rankings where we undertook an exercise to enhance the internal links, which led to thousands of percentage increase in internal referral traffic.
T.J.: In today’s economy, it would be maintaining the Privacy balance. On the one hand, we want to leverage data that offers a more targeted customer journey. On the other hand, we need to develop trust with the customer about having a robust data trade-off. A perfect example would be trackable devices & privacy.
T.J.: The biggest challenge for organizations is how they prepare raw data. Teams have to locate, extract, and normalize the data. I have mentioned interdisciplinary integration and how it is imperative to merge multiple data points from different sources and identify the patterns. To make the most sense of this data, Governance is important. Organizations have to recognize patterns of analysis. Are we being reactive and conducting only exploratory analysis, or are we mature in our analytic journey and being proactive by using prescriptive analysis to find what actions to take to change outcomes?
T.J.: As a digital marketer, I have been fortunate to be associated and lead some great projects. If I reflect on the common theme of all the projects that I have done over the years, I would say that I believe in “Narrating the story!”. Whether it was trying to convince senior management about a budget request or a migration project or demonstrating how the project would impact each stage of the user journey and data collection. How would this relate to business goals? Here, make data your friend. Exploring relationships between variables in your data can reveal surprising associations.
T.J.: As much as we talk about interdisciplinary integration, Data Silos still exists. Integration difficulties tend to arise due to architecture differences of the various measuring platforms.
Another very desired topic is automation. Although automation workflows do reduce manual processes and improve efficiencies and quality, they are often plagued by the scale of the operating model.
T.J.: This year has certainly been about the new economy. More than at any time in history, the demand for a data-driven strategy is really high. In an increasingly digital customer-centric world, the ability to capture and use customer insights to shape products, solutions, and the buying experience as a whole is critically important. Advanced data analytics is quintessential. A data-driven culture requires technology that can support efforts in exploiting data and analytics. The complexity of methodologies, the increasing importance of machine learning, and the sheer scale of the data sets.
