en

BCLC: Optimizing Media Spend and Driving Marketing ROI

BCLC transformed its marketing strategy by developing in-house Marketing Mix Models, saving costs, optimizing media spend and more.

$1 Million

saved in consulting fees

40x

Higher returns on ad spend

Thousands saved

removing external vendors and automating the MMM process for more timely refreshes

 

The British Columbia Lottery Corporation — BCLC — operates 36 casinos, 3,400 lottery retail locations, and its online gambling platform, PlayNow.com. In the last fiscal year, BCLC generated $1.5 billion in net income for the province, supporting vital public services.

Post-COVID, BCLC, like many organizations, faced fierce competition for entertainment dollars. To tackle this, BCLC sought to optimize its marketing advertising budget and turned to Marketing Mix Models (MMM) – statistical analysis techniques that estimate the impact of marketing efforts on sales, to guide spending and improve return on investment (ROI).

Previously, BCLC had relied on external agencies for these insights, but these models covered only 40 percent of their marketing spend leaving key questions about the real impact of their advertising and its ROI unanswered.

In this 2024 Dataiku Product Days fireside chat, Bindu Joopally — Senior Analyst, Analytics Delivery at BCLC — talked about the challenges and triumphs of connecting to disparate data sources and collaborating closely with the business as well as what projects the team plans to tackle next.

Watch Video
How BCLC moved from using third-party consultants to bringing marketing mix work in house with Dataiku.

Watch More 2024 Everyday AI Sessions

DISCOVER NOW

From Manual Efforts to Dataiku

BCLC’s internal team actively developed their own MMMs for flagship lottery products, Lotto Max and Lotto 6/49, despite limited infrastructure (working off local laptops) and a labour-intensive system for data processing, which covered first-party transactional data, agency data, and macroeconomic third-party data. While this process took over a year to produce its first model, its success sparked further demand, but scaling up the existing process was not sustainable. This paved the way for the adoption of Dataiku to automate and streamline BCLC’s MMM initiative.

Streamlining Processes, Integration, & Automation

At the same time BCLC began working with Dataiku, it also modernized its data architecture by migrating to Snowflake and integrating AWS and Azure services. The result being, BCLC was able to deliver three new MMMs efficiently with limited resources. The key steps were:

  1. Learning the Ropes: The team obtained the Core Designer Certification through the Dataiku Academy and attended hands-on coaching sessions with Dataiku experts. What once took months on local laptops was now replicable in just a few sessions on Dataiku.
  2. Data Integration and Simplification: Internal sales, media spend and external data from sources like Statistics Canada and Environics Analytics were downloaded into Dataiku, breaking down data silos and simplifying the process.
  3. Data Preparation: Using Dataiku’s visual recipes (such as prepare, join, pivot and stack) alongside Python code, the team cleaned and merged multiple data streams into a single dataset for analysis.
  4. Running Code in a Robust Environment: Python code was moved into Dataiku, enabling seamless integration with the necessary libraries. This shift provided a significant performance boost and cut processing times from hours to under 30 minutes, eliminating memory constraints.
  5. Visualization and Collaboration: Visual insights were created and published via dashboards, and documentation was shared across teams, streamlining collaboration and enhancing visibility for stakeholders.
  6. Automation: The entire process was automated which enabled BCLC to monitor results in real-time with minimal manual intervention and ensure efficiency in future updates and model refreshes.

A Transformative Impact

Dataiku provided BCLC with the tools to build a scalable, flexible and transparent MMM process. It enabled them to orchestrate data pipelines, enforce data governance and automate complex processes — all with minimal manual intervention. 

As a result, BCLC moved from relying on external agencies to owning their marketing intelligence, saving costs, and gaining the agility needed to optimize their marketing budget. The shift to in-house MMMs using Dataiku has numerous positive impacts on BCLC’s day-to-day operations:

  • Cost Savings: By developing MMMs in-house, BCLC no longer pays third-party agency fees for these insights.
  • Data-Driven Marketing: Teams now have data-driven insights to optimize media strategies, measure ROI across channels, identify media saturation and understand consumer behavior lag effects.
  • Improved Media Planning: The model’s insights led to substantial shifts in media allocation, with some areas seeing a 70 percent decrease and others a 180 percent increase in media investment.
  • Optimized Tactics: Online marketing tactics, particularly search, outperformed traditional ones, delivering up to 40 times higher returns on ad spend compared to TV.
  • Increased Flexibility: Automating the MMM process with Dataiku has enabled BCLC to refresh models frequently and at a fraction of the cost of external vendors.

In addition to measurable business value, these insights have empowered BCLC’s internal analytics team, fostering closer collaboration with marketing and advertising agency partners.

Going forward, BCLC is well-positioned to continue driving growth and efficiency across its business lines, delivering value for the organization and the communities it serves.

 

Western Digital: Smarter Email Categorization With NLP

Western Digital built an NLP-based system in Dataiku to categorize and better understand their emails for more efficiency (100 employee hours saved per month), reduced response time and higher customer satisfaction

Read more

Bayard: From SAS to Dataiku
for a Modern Data Stack

Bayard migrated from SAS to Dataiku for greater automation, monitoring, and visualization capabilities and, ultimately, increased efficiency and productivity.

Learn More

Regeneron: Proactive Pharmaceutical Process Monitoring

Regeneron’s BioPerceptron platform is an innovative and integrated method to detect subvisible particles in formulations. It leverages deep learning, AI, end-to-end cloud orchestration, and advanced visualization capabilities to transform proprietary, unstructured data into proactive pharmaceutical process monitoring.

Learn More

Frende Forsikring: Using NLP to Automate Claims Reporting

An initial BERT model trained on 10,000 emails in Dataiku is now being used to distribute all emails in the claim center.

Learn More

Air Canada: Democratizing & Accelerating Data & AI Projects

The mighty Air Canada team has built and runs democratized processes across multiple functions, teams, and experts to achieve extraordinary ways to make work, marketing, and the customer experience better.

Learn More