Skip to Content

Migration to cloud

Blog | September 5, 2022 | By Vijay Gowtham, Lokesh Bansal

The Advertising Council or more widely referred to as the Ad Council, is a preeminent non-profit organization that works with partners to promote social causes. With a history of more than 80 years, they are one of most recognised brands in US. They produce, distribute, and promote public service announcements on behalf of sponsors’ donated media assets. Ad Council’s campaigns are distributed thru a network of media outlets like TV broadcast, print, outdoor (i.e., billboards, bus stops), and Internet.

Ad Council’s objective is to inspire action, create impact, bring about positive change and improve lives. A critical component of each campaign is evaluation of a campaign’s effectiveness. Rigorous research goes into assessing each campaign’s effectiveness through a dashboard of indicators around exposure, awareness, engagement, and impact.

All campaign assets are monitored using the latest, cutting-edge tools to track media out-reach, placement, impressions, and coverage.

Thru continuous tracking of a campaign’s media assets like websites, social media or other digital properties and point-in-time surveys, a campaign’s awareness among its target audience, changes in attitude & behaviour is assessed.

For a campaign to be deemed success, it’s equally important to supplement in-house data trends with external data sources from other agencies or sponsors.

Given the nature of operations and priorities, Ad Council is a lean organization. With their limited means, they are unable to dedicate resources to own, support & maintain expensive hardware.

The emergence of cloud as a reliable and trusted service offering opened a new avenue for Ad Council. With ANSI SQL literate data analysts and easy integration to existing systems and applications, our client opted for Microsoft’s SQL Server hosted on AWS EC2, a fully managed platform service (PaaS) from Amazon Web Services.

This solution had various benefits that not only eliminated the need for upfront costs, with a pay as you go/use model, the costs were always easy to manage.

How this helped Ad Council

  • removed the drudgery of time-consuming administrative tasks without having to provision infrastructure or even maintain software. consistent and reliable performance.
  • consistent and reliable performance. 
  • scalable compute at the click of a button with zero downline.
  • high availability with near zero downtime. Backups were managed by AWS.
  • best in industry security for data using AWS Key Management Service.

AWS EC2 helped Ad Council pivot from routine, time consuming and resource intensive administrative tasks to managing campaigns effectively.

While this solution was effective for Ad Council, as data volumes increased the need for additional storage couldn’t be effectively catered to while using AWS EC2. Additional costs could no longer be absorbed. Also with newer data sources, the analytics team concluded the existing infrastructure was no longer able to sufficient and had reached a point of diminishing returns.

USEReady’s consultants were tasked to help identify a scalable, future proof & easy to administer data platform that could meet Ad Council’s requirements. The need for a modern cloud-based data platform that could handle scale and complexity of current and future data feeds.

A detailed and comprehensive study of the available options helped shortlist Snowflake – a cloud native data platform. Below is a quick comparison of features between the two systems.

Microsoft SQL Server Snowflake Computing 
Requires hardware provisioning Cloud native, SaaS model with no hardware requirements 
Needs periodic upgradation to latest version Always up to date 
Compute directly linked to hardware config Unlimited processing. Elastic computing 
Storage provisioned on AWS Low-cost storage on-the-go, independent of cloud service 
Parallel processing leads to contention No impact on performance due to parallel processing 
Server size is static, cost remains same No cost when no application is not in use 
additional storage required for data archival No need to separate storage for data archival 

Further, USEReady’s consultants reviewed the choice of Snowflake for specific use cases at The Ad Council.

Microsoft SQL Server Snowflake Computing 
separate servers for DEV and PROD instances Dev & Pro instances can reside within a single Snowflake account 
files needed reprocessing when metadata changes No need for reprocessing of files 
data duplicated on PROD server for approved files No duplication of approved data on a separate database 
Almost 11 hours to process daily feeds processing time reduced to less than < 2 hours 
Complex design Simplified Design 
Live connection to database from Analytical applications hindered performance. ZERO impact on performance due to live connections from Analytical applications. 
Redundant storage of data required for faster processing No need of redundant storage, faster on-demand processing 

This validated the earlier studies and confirmed the choice of migration to Snowflake would immensely benefit The Ad Council. Further, this migration would also help save significantly in terms of licensing costs and on-going expenses as illustrated below.

6-month 1-year 2-years 5-years
SQL Server on EC2 AWS Cost (2 x r3.4xlarge) $37,800 $75,600 $151,200 $378,000
Software License Cost $0 $0 $0 $0
Snowflake AWS Cost (1 x r3.2xlarge) $9,450 $18,900 $37,800 $94,500
Software License Cost $5,000 $10,000 $20,000 $50,000
Savings $23,350 $46,700 $93,400 $233,500

Next in Series – How the migration to Snowflake helped The Ad Council modernize it’s data stack.

author image
About the Author
Vijay has been a data professional for more than a decade and half. Over the years, he has helped businesses scale, improve operational efficiencies and build competencies in visual analytics, cloud data engineering.
Vijay Gowtham | Practice Area Lead - Cloud Data Factory, Director | USEReady
author image
About the Author
Lokesh Bansal is a data platform modernization expert with more than 14 years of diverse consulting experience. He has worked in various capacities on engagements in data warehousing, data visualization, metadata management, cloud migrations, platform rearchitecting & refactoring.
Lokesh Bansal | Solutions Principal | USEReady
Back to top