In this digitally conscious world, business models are undergoing a paradigm shift triggered by the changing consumer behavior and rapidly evolving market dynamics. This behavioral change, across multiple dimensions, is the driving force for enterprises to look at data, both internal and external, from a lens that encourages careful adoption and sharing of data assets for delivering ecosystem-wide impact. Data has come a long way – from being a closely guarded internal asset, to now being recognized as a shared asset. Every business is actively exploring external sources to understand the customer landscape and to drive growth.
The usefulness of data is measured based on its relevance (context) and the actionable insights it can possibly deliver. In its natural journey of transformation, a data landscape assumes a structure similar to that of a water system with lakes, rivers, check dams, canals and river farmers (hunters). Let’s take a quick look at some of the terminologies before we dive any further.
- River farmers (hunters) are the data hunters, who are constantly on the lookout for fresh consumable sources of data that can bring value to your internal data repository
- Rivers are the external data marketplaces that could be categorized as product data marketplaces (AWS, Snowflake/Oracle), syndicated data providers (Nielsen, IRI) and, business partnerships across verticals. These rivers give you a constant feed of fresh and accessible data from a plethora of sources.
- Canals are the tech architecture pipelines that pull and leverage data from marketplaces to ensure that your data lake’s health is maintained
- Check dams are your data governance and quality control towers where data hunters, along with stewards and tech partners, act as custodians of the data flow
Recommended approach to them work in tandem
Though enterprises interested in this landscape are keenly acquiring more data, showcasing value from these can only come through incremental insights. These incremental insights enable cost justifications and further drive enterprise-wide awareness, adoption, and consumption.
In a conversation with a leading retailer in South America, we realized that even internal channels such as superstores, e-commerce, and convenience stores within the retailer, were hesitant to push or leverage data from a common data lake. The lack of understanding of how a data set from one channel/category could impact and facilitate the other, led to dry data lakes or sometimes even forced drain data (irrelevant data) into active data lakes. This further exacerbated with flow of external data – leading to no tangible outcomes. Thus, we recommended an insight to data map – similar to that of a play store app, that showcases data asset value to the larger ecosystem.
An insight-model-datamap helps your data hunters to not only map requests to data, but also encourages the larger team to learn and leverage from insights of others – a data insight play store of sorts. Tools such as Collibra, Alation or PowerMe are addressing the need of building connected data lineage and tracking consumption patterns. This, in combination with internal customization to address specific business-level use cases, can go a long way in driving the data-to-value culture.
Further to this, it is crucial to build the right data hunting strategy. While each organization has its unique data flow, here is a recommended approach that can be tweaked to needs.
These steps are intuitive and flexible. Let us assume a new use case is being built and the request for data has reached the data hunting team. Once the available data from step 1 is tallied against a business use case, the data hunting team can start scouting marketplaces or partners for the right data fit. Evaluation parameters today can vary from granularity, coverage, recency-frequency and compliance to local regulations. Sometimes governance and ownership cost take precedence for assessing value drivers. With a myriad of available data partners, determining which partner and for what purpose, is a demanding exercise. This process requires a man-and-machine ecosystem, needing human touch to evaluate and machine to automate data flows.
For a leading tech giant that wanted to capture upcoming trends and identify potential opportunities for its collaboration platform, the internal data alone failed to give any actionable insights. We recommended building a growth analytic framework. A data hunting team was mobilized (externally) and a number of platforms for intent, purchase behavior, geo location and, social sentiment was evaluated. The team narrowed it down to Bambora, Mintigo, HG Insights, social sentiment data from LinkedIn and worked on integrating these external data sources to internal product purchase data. This process, enriched with automated data flows and ML ops, helped the tech giant identify product penetration, technical implementation and post-deployment customer success, across geographies. The end product was an analytics-driven cockpit that enabled unlocking a potential worth of $80 million.
Furthermore, data efforts that are more nuanced to a particular business – like fraud detection in insurance or revenue management in CPG, can benefit from B2B partner ecosystems. Fraud detection and prevention is quite an investment for insurance firms. We are observing trends, where specific insurance data marketplaces and B2B partnerships with credit card firms like VISA/Mastercard are driving richer data-driven insights with extensive coverage. While CPG firms rely heavily on Nielsen or IRI for syndicated data, partnerships with retailers have their unique vantage points of making granular level information easily accessible. When choosing a B2B partnership, you must keep a close watch on agreeing to co-shared outcomes, which in some cases, might not be beneficial for either party. While a fraud detection pattern identification benefits both the credit card company and an insurer, the same doesn’t apply for a CPG firm looking to optimize its margin, while the retailer pushes for Everyday Low Price (EDLP).
Advanced teams can choose to create an insight marketplace rather than a data marketplace. However, the external insight and data marketplace today are still questionable with respect to actionability, applicability and trust of insights (based on data used and model performance). These concerns prevent immediate acceptance and usability of an insight-based marketplace in place of a data marketplace.
The best foot forward today is to move towards external data management. While many customers are leveraging sustainable benefits from external data sources, it is advisable to be careful when ingesting such an external asset into the business ecosystem after proper validation, appropriating privacy laws and preparing an internal architecture, ready to consume its impact and deliver positive outcomes.
Links to further reads
As a digital growth consultant, Vigneshwar leads consulting interactions and delivery of data-driven strategy for the Fortune 500s. He is highly skilled and experienced in leading analytics partnerships to success. He is always up for conversations around tech trends, problem solving and loves to spend his weekends idling by the beaches.
Subscribe to blog