Author Archive

Rachna Manwani

Rachna Manwani is Vice President of Customer & Marketing Analytics at Manthan. Rachna has 15 years of experience in helping Fortune 500 companies across various industries such as retail, CPG, healthcare, telecom and software. At Manthan, Rachna is focused on helping retail and CPG businesses leverage analytics and Big Data technology to improve understanding of its customers and provide an omnichannel marketing experience. Prior to Manthan, Rachna served in analytic leadership positions at IRI, ZS Associates and other startup organizations.

What goes into a Customer Data Platform Implementation

Milestones to track, success criteria and key stakeholders

This article is part of our CDP series. To read other articles in this series, click here.

Customer Data Platforms are in fashion, but not a trend that will fade away soon. As a pre-packaged, marketer managed system that unifies a company’s customer data to enable use cases such as personalization and contextual marketing, a CDP delivers high value to B2C businesses.

A successful CDP program requires cross-functional team effort where marketing plays a critical role in defining the program objectives and success criteria. Given the nascent stages of the technology, there is little guidance available to businesses investing in a CDP.

This implementation approach note will serve as a handy guide for B2C enterprises looking to ‘buy’ (and not build) a CDP. Read this for our view on build vs. buy.

Broadly, a CDP implementation program has three phases:

Phase 1:  Business and Data Discovery

This phase involves understanding business needs, customer data and the technology landscape – along with buy-in from key stakeholders on how the unified customer record will be created and maintained in the system.

Typical activities include:

    1. Planning and requirement setting

Marketing (CRM/ loyalty and marketing operations) leads this phase where the ask is to define four key program requirements

        1. Marketing objectives – this could be improving loyalty engagement, customer retention , customer acquisition or customer engagement across touchpoints
        2. Challenges with current data – is it gaps in the data, poor quality or lack of trust in available data, or a combination of all
        3. Success criteria – metrics that business will track to validate the quality of customer data against marketing objectives
        4. Marketing campaigns – the types of campaigns that will be executed once transformed customer data is available, and how the data will be fed to marketing systems such as campaign execution tools
    1. Identify data sources

This session is primarily led by data and IT engineering teams who manage customer databases. Marketing technology and analysts may also be involved if they capture customer data via surveys or campaigns. Other aspects covered in this phase require understanding the primary source for customer master, data collection, integration and verification process

    1. Establish business rules for customer data unification

Based on available data and customer marketing priorities, CDP vendors work with marketing and data/ engineering to establish rules for processing and creating a comprehensive view of customer. Some of these rules could be around:

      1. Standardization and verification of captured customer data
      2. Prioritization of data sources for identifying customer attributes, for example, in cases where more than one value/ inconsistent data is present for a given attribute (such as more than one home store for a customer or multiple phone numbers for a customer)
      3. Establishing mechanism for de-duplicating customer records. This would require agreement on approach to run deterministic and probabilistic matches to create the golden record of a customer
    1. Customer data enrichment

This consists of not only adding customer data from third party sources but also enriching customer data with analytical insights using the ‘as-is’ customer data. Examples of such enrichments are – deriving customer behavior such as a ‘discount buyer’ vs. ‘full price’ buyer – derived from historical purchases and behavior. ‘Active’ visitor or an ‘engaged’ customer based on the customers interactions on e-commerce and digital campaigns. The analytical tags maintained for every customer vary based on marketing objectives, and form critical inputs for segmenting customers to drive engagement. Marketing plays a critical role in identifying data gaps and determine how the data should be enriched.

Phase 2: Deployment

This is where the rubber meets the road. Deployment phase is vendor led (since we are talking buy option) – once critical data (such as customer master) is loaded and processed in the system, marketing is given access to the data. Other customer data activities such as enrichment of ‘nice to have’ customer attributes and cleansing of low priority customers (as defined by business) are executed in an agile manner. This ensures business can use customer data early in the cycle – they can test and validate the results and provide inputs, without having to wait till the end of implementation when it’s too late.

  1. Data onboarding and ingestion: Getting the process and mechanisms to incorporate data sources and rules agreed upon in Phase 1 in place, to create the comprehensive customer data record
  2. Configure data feed into downstream marketing systems: This is to ensure marketing teams have access to create customer segments, extract customer lists and data for analytical and personalization purposes
  3. Set-up process for tracking customer metrics: This stage requires setting up a tracking and measurement processA critical but often overlooked component of a CDP implementation is to track the health of customer data and business metrics. Information such as total customers vs. active/ inactive customers, customer retention, acquisition and conversion rates are key metrics to track to ensure the program objectives are being met
  4. (Optional) Set-up customer engagement and campaign integrations: Not all CDPs include ‘engagement features’ – such as real-time interactions or recommendation engines that provide personalized product and offer recommendations to customers using preferred channels such as email, mobile app, online, SMS etc. B2C marketers see higher value from these optional capabilities, making it a prized component of a CDP

Phase 3: Hand over and training

The final phase in a CDP program is to ensure marketing and other customer-facing functions are trained and have access to customer data for their requirements. In addition, data/ engineering teams are trained on monitoring data ingestion and serving from the CDP on an ongoing basis.

The success of a CDP program requires establishing a governance team which comprises of marketing, IT and data engineering. Collectively, they ensure ongoing collection of new data, and verifying/ tweaking the applied rules such that they stay current and continue to meet business needs. By tracking the health of customer data and marketing metrics such as loyalty, acquisition, retention, conversion and engagement consistently, an organization can become truly customer-centric. A well-executed CDP program delivers on business goals quickly, be it controlling customer churn, increasing share of wallet or growing customer value.

Download CDP Handbook Now!

Retail customer loyalty programs

How Big Data Improves Loyalty Programs

Until a few years ago, most retail loyalty programs were all about handing a card with a Rewards number with primary focus on rewarding shoppers based on their spend. Reward typically implied reduced pricing achieved in the form of (i) redemption of earned points towards a purchase, or (ii) a deeper discount for being part of the loyalty club, or (iii) simply access to ‘special’ member pricing. With increasing competition, introduction of new channels of shopping (brick and mortar vs. online), increasing number of millennials driving consumption; retailers are being pushed to rethink their loyalty program. While retailers may think loyalty is dead and shoppers are simply seeking for the best price, several studies indicate that the new age shopper tends to be loyal to brands/retailers – they are not loyal to just ONE SINGLE retailer for their shopping needs. Today’s shoppers do not equate loyalty program to a discount alone. They seek a loyalty program which offers value. This value implies:

Personalized benefits:

Receiving product or offer recommendations that’s relevant to his/her purchase needs. Shoppers want discounts, but also want personalized service or attention which improves their shopping experience. Examples of personalized services could be (i) free valet parking, (ii) free delivery service, (iii) access to a personal shopper or stylist, (iv) express checkouts for high spenders.


Ability to redeem rewards earned in any format. Gone are the days when shoppers carry printed or digital reward coupons to be able to redeem the rewards. Shoppers demand that retailers know them and are able to pull up and apply the relevant rewards/offers earned towards the purchase.

Timely rewards:

Ability to redeem rewards timed to shoppers’ purchase cycle. While shoppers have become more demanding, they’re also more open to sharing their information in exchange for value.Big Data platforms, solutions and services can help marketers redesign their loyalty program that can create a win-win offering for both the shopper and the retailer. Here’s how:

Data Capture:

Big data platforms give retailers the ability to gather and store large volumes and different kinds of data. Big data solutions can store not only structured data such as shoppers’ loyalty rewards earn/burn and purchase transaction history but also unstructured data such as online browsing history, feedback/survey, shopper call logs and engagement on social media brand assets. These different data elements help retailers get a comprehensive view of their shoppers.

Data Analytics:

Big data solutions can execute analytics at scale and quickly generate different shopper personas with richer insights. Based on shopper personas, marketers can define loyalty programs to achieve the desired behavior – i.e.: (i) increase repeats, (ii) increase spend; (iii) acquire; (iv) retain at-risk of churn/churned shoppers; (v) increase profits. Simultaneously, marketers can identify shopper personas to map relevant incentives to drive behavior – (i) offer discounts and/or (ii) offer service perks.

Process automation:

Big data technologies have made it possible (i) to automate the data collection and storage process, (ii) run analytics to deliver shopper insights quickly, (iii) deliver personalized product or reward to each shopper based on his/her channel of engagement , purchase history, intent to purchase and timing, and (iv) to track metrics and change in customer behavior to measure the success of the loyalty program. How are you using Big Data to improve your loyalty program? What challenges do you face? Share your thoughts …….. Related Solution: Transform your Customer Marketing with Manthan’s Customer Analytics Solution