Using big data for big wins in media
Nov 5, 2020
I recently had the pleasure of facilitating our Data Solutions Event for Media, where we shared the latest on the most impactful ways you can harness big data in the media industry. Looker customers GroupM and WarnerMedia delivered fantastic presentations, as did some of our Looker team members. We highlighted the importance of automatically bringing data when and where it’s needed to drive better decision making, and to continue the push for data democratization. Join me for a recap of the day’s insights.
GroupM’s data journey: becoming data literate, compliant, and custom at scale
Our event kicked off with a session by Polina Melamed, Marketing Director, Data and Technology at GroupM. You may not recognize the name of the company, but it’s almost guaranteed you interact with them on a daily basis. As the world’s largest media investment company, they are key partners with Google, Facebook, and Amazon, serving 121 markets and representing a third of global media billings.
GroupM is facing a changing industry, including a stringent regulatory environment with regard to customer data usage, increasing consumer concern about how information is shared, and rapidly fragmenting technologies and tools. Their client and industry stakeholders have high expectations about the company’s data literacy and expect the tools they’re using to drive campaigns to be precise and compliant.
GroupM had built an in-house reporting product to provide customized media reporting to its vast client base and leverage its large volume of metadata — but it just couldn’t scale the product and customize it the way customers wanted. Enter Looker. The company has deployed Looker in a unique way, with three macro-modeling layers:
- The first is a universal, foundational model that establishes globally applicable data definitions like “spend, impressions, month, and year.”
- The second layer is applicable to their agency customers. This includes definitions like cost-per-click.
- The final layer is where the magic happens, or, as Polina called it, “the client game changer.” This is the individual advertiser model layer. A streaming media company and a consumer packaged goods company have very different definitions of cost-per-acquisition, and they can be as granular — and as unique — as they need to be with Looker.
This three-tiered system has yielded major benefits. They have stable, configurable data models for all clients and can update them in a fraction of the time and effort it took previously. They onboard new clients rapidly. And finally, all internal users write their own queries in Looker. That includes business users — 50% of GroupM’s analytics are run by non-analysts. They run queries like, “How much did we invest with Partner X last year, and what results did that investment produce?” Or, “How did my cost per acquisition compare to my last campaign, across each channel?” It’s the epitome of data democratization.
Meeting WarnerMedia’s data challenges head-on with Looker
Renee Ducre, Senior Director of Data Strategy at WarnerMedia, generously hosted a live fireside chat. She shared the company’s data journey and the challenges they’ve met along the way, such as:
- Bringing together disparate data sources, reporting, and visualization tools
- Reducing ad-hoc, time-consuming, manual workflows and reporting
- Developing cross-platform reporting and insights
- Driving adoption of new Looker reports
- Adhering to data governance and security controls
With Looker, WarnerMedia has met these challenges head on as a data-driven media company. Getting data right at WarnerMedia is critical — last year, 113 billion total minutes were consumed across all its platforms, and the company had the most engaged social sports audience in the industry. WarnerMedia uses the insights they find in Looker to help inform content, subscriber engagement, and customer experience strategies.
Renee shared some reports from Looker, showing how we helped WarnerMedia bring together disparate data sources, develop cross-platform reporting, and automate reporting that was previously done manually and was extremely time-consuming. The “March Madness Daily Registration Reports” enable the marketing and operations teams to optimize registration flows and prioritize marketing activities. A report showing “March Madness” viewer engagement across different platforms — mobile and desktop — provides a single dashboard for operations and product teams.
Before Looker, the teams had to create an unwieldy report that they couldn’t drill down on. The Looker dashboard allows them to drill down on each element for further insights. And, a “Sources of Traffic” report shows the content and marketing teams how well their campaigns are driving audiences to “March Madness.” For the NCAA tournament, Looker replaced a massive static Excel document with a real-time, in-tournament analysis that summarized tournament performance at a glance.
Renee has a few key lessons on data and media from WarnerMedia’s Looker deployment to share with other companies beginning their data journeys:
- Be patient in your data journey. It takes time.
- Having a data-driven culture and buy-in are critical to success.
- Identify a few quick wins and implement them.
- Prioritization of data projects is important once you build momentum.
- Democratization of data, reporting, and insights are key to scalability.
What’s next for WarnerMedia? They’re increasing self-service analytics to free up more cycles for research analysts. They plan to integrate artificial intelligence (AI) into data analytics to drive deeper insights. And, they’re planning to integrate data across all WarnerMedia properties to drive amazing data experiences for fans.
Making data-driven experiences easy—announcing the new Extension Framework
As we saw from both GroupM and WarnerMedia, bringing data to business users and using it to automate critical business operations is the key to building a data culture. Steve refers to this as having data “meet us where we work.” The new Extension Framework is a great way to get started with that. With custom data applications, you can truly make data operational. And when it’s baked into people’s actual daily workflows, you start reaching beyond business intelligence and into building rich, data-driven experiences.
Think of the framework like a sandbox on top of the platform: you can leverage the Looker application programming interfaces (APIs), user interface (UI) components, authentication, and permissions. This enables you to focus on the data experience instead of the infrastructure and technology underpinning it, so you can build internal platform applications or external platforms for your customers (like customer portals for Powered by Looker applications).
“It took one developer one day to stand up an application using the extension framework! I’ve seen a lot of great Looker features built over the years. This has the potential to be the most ground-breaking.”
— Jawad Laraqui, CEO, Data Driven
Steve also announced the general availability of the Looker Data Dictionary Extension, the first of several extensions we’ll be releasing this year. The data dictionary is a way of investigating your data model without making changes to the model itself. For example, you can see actual definitions for fields in your model (distribution, heuristics, and minimum or maximum values) that are stored in the data warehouse.
The goal of our Extension Framework is to reduce friction for your developers, operationalize your data, and expand your data culture. You can start developing data experiences with the framework quickly, creating a common language and experience—and know that everything “under the hood” is dialed in.
Wrapping it all up: data-driven experiences in media with Looker
The final session of our event tied all the others together by diving deep into what really creates a data-driven event in advertising. Connor Sparkman and Tom Yeager, Customer Marketing Analytics Managers at Looker, delivered a great presentation showing how media companies can leverage and create data-driven experiences.
We use the term data-driven a lot, so it was great that Connor and Tom took the time to really explain what we mean by it. Every decision needs to be backed by data. Every workflow in your organization needs to have data embedded. Everyone — internal or external — needs to have the right data delivered to them, at the right time.
Connor and Tom encouraged listeners to think of it in three core categories:
- When? Make sure information is delivered to people when they need it. Many of our customers use Looker during meetings with prospects or clients, showing what they can expect when they sign on. Or, they send out information on a regular cadence with the scheduling feature, so no one has to search for it.
- Where? Ensure information is delivered where someone needs it. Integrate it right where they are — into daily or operational workflows. This increases the value they can derive from it and the decisions they can make. Leverage Looker APIs — with Marketo, Slack, and Salesforce, for example — to deliver reports right where your users are.
- Who? Remember, you have internal and external data users. Data democratization and a data culture are key to your internal customers. You can also provide data as a product to your customers, utilizing Powered-by-Looker (PBL).
There are four ways you can get data-driven with Looker:
- Integrating with modern business intelligence and analytics products by serving up real-time, relevant reports and dashboards that act as a starting point for more in-depth analytics
- Integrating insights and relevant information into the tools and products people already use
- Supercharging operational workflows with complete, near real-time data
- Building custom applications to deliver data in an experience tailored to the job
Connor walked us through the first example in detail, taking us through an integration with business intelligence and analytics products to track a website visitor. Using Google Analytics, we followed the journey of a site visitor to determine whether the individual is a unique or a repeat visitor; and whether there are URL parameters, cookies, marketing campaign or referral links associated with the visitor. Google Analytics maps the unique visitor ID to help us track web behavior. We can capture leads or marketing automation through a Marketo integration, and store prospect information in Salesforce. With these integrations, data is already modeled, and dashboards are already built to explore metrics like spend, revenue, clicks, and conversation. ROI can be broken down by cohorts, campaigns, and keywords.
It’s important to integrate the analysis of this journey with tools that people are already using — for example, automating data from Looker back to Salesforce or Marketo. This makes it easy and automatic for business users to interact and engage with insights.
Wrapping It All Up
I hope you enjoyed attending our Data Solutions Event for Media or catching up on the recap above. You can find more resources at our Media Content Hub or our Media Solutions page. We’re here to provide resources that will help you automate when it makes sense, create data-driven experiences for your team and your customers, and continue making data as accessible as possible for everyone.