What is demand forecasting in Retail
8 June 2020
What I learnt in first two years of Agile Implementation
11 November 2020
Published by
BluePi
Data-Driven Business Transformation
Role of third party data in a customer data platform
If there is one sector that has remained buoyant despite the global pandemic, it is online commerce. Brick and mortar players are also a significant shift in revenue streams, and omnichannel presence is a necessity. The changes in the retail sector are undergoing are transformational and long-lasting.
A significant enabler in online commerce comes from customer behaviour data which results in customer intelligence. We see some tectonic shifts in data collection forms in a privacy-first world. New challenges are being in the form ad blocking, GDPR and other privacy regulations and third-party cookie elimination. All these moves make identity resolution difficult and tricky. Yet as our users become more privacy-aware, they demand more personalised and seamless experiences.
Add to this the changes in consumer demand fostered by the pandemic, and our need to adapt our customer data strategy is immediate and unavoidable. This customer data strategy needs an organisation-wide effort to break through silos to get a 360-degree view of all customer activities.
While utilising internal first-party data provides a view into the customers’ real persona, it is still a limited view of who he is. This narrow view is where data enrichment with second and third-party data sources comes into play.
Customer Identity is the Core
Every customer-centric business must adopt a coherent identity strategy across different use cases. Identity lies at the heart of delivering customer experiences as the customer interacts with the various touchpoints in the organisation the background needs to be seamless and consistent.
Usually, there are disparate systems where the customers’ data shows up be it transactional, marketing, financials, CRM & many others. Each of these systems has their customer id and attributes. To synchronise these attributes across systems to come up with a unified customer view is a challenge in its own right.
Many lives of Customer ID
The customer’s identity varies across the various systems that make up the IT landscape.
- In traditional databases and systems, it would be Customer Name, Address or a combination or a surrogate key.
- In customer-facing web-based systems, it could be the cookie.
- Newer digital systems may use email id or even the mobile number
In retail specifically, as omnichannel experiences become pervasive, it must be possible to determine that a customer who is purchasing an item in the store is the same one who added it to the cart but abandoned it later.
Hence we need a system that keeps tracking, correlating and updating the customer information across multiple different sources (first, second and third-party
There are three primary techniques for identity resolution. We could choose a deterministic, probabilistic or hybrid approach.
- A deterministic technique consists of a rules-based matching of data attributes to enrich identity data.
- A probabilistic model that uses AI-based matching
- A hybrid model that uses both of the above is a more popular third option.
The customer data thus cleaned must have all the attributes of data quality
- It should be accurate
- It should be current and up to date
- The data should be complete
- The data should be reliable – must come from trusted sources
- The data must be relevant
Customer data enrichment
Next step in the process involves customer data enrichment. We can procure customer data from various sources, including premium data providers, online sources, data owners, retailers, census and survey.
Third-party data is collected and used by DMPs (data management platforms) to create targeted profile segments. Companies can then either directly purchase these segments or enrich their first-party data to make better targeted and larger segments.
Two trusted names in this space are Lotame and Mobilewalla in India.
Use Cases for Third-Party enriched data
Armed with access to third-party data and targeted customer segments, retailers can focus on specific use cases to build and grow their business and customer acquisition strategies.
Personalisation
Delivering highly personalised products recommendations brings great business values to the brands while improving the customer experience on the site. The most significant benefits arising from third-party data enrichment stems from the ability to define the segments in a more granular manner.
Marketing Spend Optimisation
An advanced use case lies in optimising the marketing spend based on the enriched third party data. For example, underperforming segments are delivered promotions in a more suppressed manner, frequency capping and staggered delivery of messaging and AI-based channel and creative optimisations.
Omni Channel Customer View
Creating a single view of the customer across the brick and mortar and online channels provides a significant opportunity to optimise the promotion delivery and customisation. Measurement data across channels offer a substantial opportunity to marketers to better target the promotions.
Collaborative Filtering
User-User Collaborative filtering is probably the most common intelligent recommendation algorithm. A user data view with attributes enriched from third-party data sources can make the recommendations more accurate and significantly boost conversions.
Summary
In summary, to unlock the most significant value from customer data platform initiatives, we suggest a four-step methodology :
- Clean the Customer Identity data
- Build a 360-degree view of the customer across properties/channels
- Enrich customer data with trusted third-party data sources
- Deploy intelligent marketing automation solutions that include recommendations, personalisation amongst others.
