In our previous post, we described four stages of the Data Maturity Journey in companies.

As we described in the Starter / Unbundling CDP stack, it eliminates duplicate data flows with a single and unified data layer. This layer captures and integrates the user profile and event data generated by your websites and apps. Once implemented, it drives data consistency across all of the SaaS tools without painful integration work.

How should the Growth Modern Data Stack be implemented?

With this stack, you’ll be able to unlock the next phase of optimization and value with bi-directional data flows and centralized storage, compute, layer. The Growth stack is the third stage in the Data Maturity Journey.

In this stage, there are two main challenges:

1. Introducing a centralized data storage & compute layer to serve as the single source of truth for customer data.

Getting a complete view of customers and their journey by removing data silos from the stack and centralizing all customer data within the central storage & compute layer.

2. Make data available for use by every tool in the stack.

Turning data destinations into data sources, creating bi-directional data flows.

Pulling data from separate data systems (like financial systems, Sales, CRM, Customer Success or Onboarding tools) into the central storage & compute layer, making the central data store accessible to every tool in the stack.

A company should implement this growth stack if it exhibits the following symptoms:

-There are data silos between all the tools that shape the company stack. A centralized layer does not exist for data.

-Incomplete customer journey data. Because the various data are siloed in their own data system and prevent the creation of a single customer view across platforms, their capacity to evaluate user behavior across tools is constrained.

-Develop more advanced use cases requiring a centralized layer that integrates customer profile data, behavioral data, and events from SaaS tools.

 

A 5-step process for implementing a Growth Stack :

 

1. Set up a Cloud Data Warehouse and feed it with behavioral data from customers.

 Cloud data warehouse will be the center of the growth stack.

 -Customer Actions (Behavioral events).

 -Customer Profile data.

 -Account or organization profile data.

 

2. Eliminate SaaS data silos with ETL pipelines.

Centralizing your behavioral data and relational data in your warehouse is a huge step in modernizing your stack and improving your analytics. Joining tables from different data systems allows you to integrate a holistic view of customers to answer more advanced questions, such as:

What are the top pageviews for qualified leads who have made more than 5 purchases in the last year?

 

3. The event data from SaaS tools can be pulled in via streaming sources.

There are many SaaS platforms that produce event data. Event data points such as email opens and clicks are vital to the customer journey, but they are trapped in email marketing tools. Similarly, your sales team creates valuable data in your CRM related to deal cycles, opportunity stages, …, and this data remains there. For a complete view of the customer journey, it is crucial to synchronize these events with bi-directional data flows in centralized storage.

 

 4. Create a unified customer profile and customer journey in your warehouse.

 At this point you have collected all of your customer data in the warehouse:

 -Behavioral data and user traits from your websites and apps.

 -Relational data from SaaS tools and other systems across the stack.

 -Event data produced by SaaS apps.

 

This step provides you with the following information:

 

Complete customer profiles (Customer 360 or a single view of the customer).

You join every customer data point from every tool into a master table of customer profiles.

Complete customer journeys.

You aggregate all user interactions (pageviews, add-to-carts, purchases, emails clicked, etc.) associated with each user during their customer journey. This gives you a detailed overview of what each customer has done.

 

5. Data can be activated in any of your stack apps.

As part of this 5 step process, data is pulled from the central warehouse and sent to downstream tools. The pipeline used for this data flow is called a reverse ETL pipeline.

You can use reverse ETL to send that master customer table to downstream SaaS tools so that all systems have the same customer information. Every tool can use every customer profile and behavioral data from every other tool. This enables all sorts of powerful segmentation and analytics use cases in downstream tools.

With all of the data in the warehouse, you can easily join transactional data, demographic data from the CRM, behavioral events and email marketing data from the event stream to produce a table of high-value users.

With reverse ETL, you can send that computed high value tag from the warehouse to all of your downstream tools. That enables Marketing to easily create dedicated email campaigns and Product to analyze flows and usage for that specific segment in their product analytics tool.

 

The following are the outcomes of implementing the Growth Stack:

-Elimination of data silos across the stack.

-A single and centralized source of truth for every customer data point.

-A comprehensive and complete user profile is available.

-Full visibility into the customer journey.

-Data analytics can be performed on any customer data.

-Data can be sent from your warehouse to your entire stack.

 

Components required for implementing the Growth Modern Data Stack :

Bi-directional data flows.

1. Data source.

-SDK for tracking behavioral events  (APP and Website).

Technologies : Segment, Rudderstack, Amplitude, Freshpaint.

-ETL from SAAS TOOLS. 

Technologies : Fivetran.

-Event Streaming from SAAS TOOLS.

Technologies : Fivetran.

2. Integration Layer.

3. Storage Layer.

4. Transformation Layer.

Technologies : dbt

5. Data Activation (rETL).

Technologies :  Census, Hightouch

Use cases to personalize your service based on data :

-Product analytics.

-Email marketing and automation.

-CRM.

-Ad platforms.

-Customer Success.

-Account Health.

-Personalization.

-Finance.

 

Eric Dodds’ idea post inspired this piece of content.

Avatar
Posted by:Fran Castillo

@francastillo