2,000+ external clients accessing data through self-service portal
300—400 columns of raw data processed through the scalability of Looker
GumGum implemented Looker as its primary BI tool for managing KPIs and data modeling, centralizing data, giving external clients self-service capabilities, and enabling conversational analytics and AI.
GumGum implemented Looker as its primary BI tool for managing KPIs and data modeling, centralizing data, giving external clients self-service capabilities, and enabling conversational analytics and AI.
Searching for a unified BI platform
GumGum is a leading ad tech company that processes billions of real-time contextual, creative, environmental, and historical signals to match every ad with the most receptive audience. With a global footprint and more than 450 employees, GumGum facilitates ad placements through direct deals and programmatic bidding, serving a diverse clientele that includes advertisers, agencies, demand-side platforms, and publishers. At the heart of its operations lies a massive data ecosystem, where transparency, performance tracking, and real-time insights don’t just add value—they're business imperatives. The company’s data is critical for internal key performance indicators (KPIs) and must be shared with business partners for transparency and to report on targets. As Ruchi Singh, Manager of Analytics Engineering at GumGum, puts it, “The ad exchange is the brain of our business, and data is the nervous system. Everything flows through it.”
Before adopting Looker, GumGum relied on a patchwork of databases, which created silos and increased maintenance overhead. The company first consolidated its reporting data in Snowflake to reduce overhead, but it still lacked a unified business intelligence (BI) platform, as teams used disparate tools and manual processes to calculate KPIs. Additionally, as internal and external clients depended heavily on analysts for data access, bottlenecks slowed decision-making. With a growing customer base and increasing data volume, GumGum needed a solution that could scale without compromising performance. “We deal with massive amounts of data, including open market bidding and near real-time streaming data,” Singh says. The raw data has 300—400 columns, and the scale involves collecting historical and current bidding data, including scanned page URLs and timestamps in near real time.
LookML is our superpower. It gives us full control to implement business logic, formatting, and calculations directly in the platform, which would be time-consuming in other BI tools.
Ruchi Singh
Manager of Analytics Engineering, GumGum
Solving data challenges with LookML
GumGum selected Looker as its enterprise BI platform for managing KPIs and data modeling, drawn by its robust modeling capabilities and the power of LookML. GumGum particularly liked the solution’s semantic layer control. “LookML is our superpower,” says Singh. “It gives us full control to implement business logic, formatting, and calculations directly in the platform, which would be time-consuming in other BI tools.” With LookML, the company can also quickly implement front-end logic, use table calculations, and integrate easily with SQL and HTML, empowering the team.
GumGum also took advantage of the solution’s ability to embed dashboards with row-level data access, ultimately building a self-serve portal for more than 2,000 external clients. In addition, Looker’s seamless compatibility with Snowflake, GumGum’s cloud data warehouse, ensured high performance and flexibility.
We make sure execs never wait for data. It’s always ready when they are.
Ruchi Singh
Manager of Analytics Engineering, GumGum
Singh’s team leveraged LookML to define metrics for both the demand and supply sides of the business, enabling nuanced reporting for different stakeholders. “We use Looker to define guardrails around metrics,” Singh explains. “For example, what we call ‘publisher revenue’ externally is considered a ‘cost’ internally. LookML lets us manage those perspectives cleanly.” GumGum also implemented continuous integration and deployment (CI/CD) practices using tools like Looker Continuous Integration to validate LookML code, ensuring quality and scalability as more engineers joined the analytics team.
Ensuring a single source of truth and empowering clients
Looker’s modeling layer allowed GumGum to ensure a single source of truth for data by centralizing business logic and eliminating discrepancies in KPI calculations across teams. With a central place for business metrics, the company can ensure people from all parts of the business are performing calculations with the same measures. Also, by embedding Looker dashboards into a self-serve web portal, GumGum gives more than 2,000 external clients simple access to their own campaign data securely and independently. Clients’ specific IDs are used to implement role-level security, ensuring they only see their own data.
The portal offers read-only access with features that facilitate changing filters, viewing performance over time, and downloading data, and has eliminated practically all support tickets for data requests to GumGum, as clients use a custom-built solution that’s intuitive, highly secure, and fast. With fewer ad hoc requests, GumGum’s analysts can focus on innovative and revenue-generating projects. “We went from being a service desk to being innovators,” says Singh.
GumGum executives use Looker to monitor performance in near real-time, with alerts set up for slow queries and compute prioritization to ensure a seamless experience. This increase in visibility is leading to improved executive engagement with data and faster strategic pivots. Singh says, “We make sure execs never wait for data. It’s always ready when they are.”
KPI consistency and trust is another benefit for GumGum, because Looker’s semantic layer ensures that all teams—from sales to finance—use the same definitions for key metrics. The result? Improved cross-functional alignment and faster decision-making. “There are no more debates about which number is right. Everyone’s on the same page,” Singh confirms.
Additionally, by consolidating onto Snowflake and integrating with Looker, GumGum achieved a scalable, cost-efficient data stack, with better control over compute usage and faster query performance. “Snowflake gives us the flexibility to allocate resources where they’re needed most,” says Singh.
Enabling conversational analytics and AI
GumGum is now actively exploring Looker’s AI features, including Gemini in Looker and Model Context Protocol (MCP), to drive operational efficiencies. For example, the company sees conversational analytics as an effective way to address the frequent rotation of new team members, particularly in sales, by allowing them to ask questions in natural language and pull data without extensive training. “Sales teams rotate frequently. Conversational Analytics lets them ask questions in plain English and get answers fast,” Singh says. Early tests show promise in reducing onboarding time and increasing data accessibility.
GumGum is excited about the future of Looker, particularly the potential of agentic AI and automated optimization. “We envision a world where data becomes invisible, where agents talk to each other, make decisions, and optimize the business without human intervention,” says Singh. With Looker as a foundational pillar of its data strategy, GumGum is well-positioned to lead the next wave of innovation in ad tech.
GumGum, headquartered in Santa Monica, California, is the mindset company transforming advertising. Using their AI-driven data engine, the Mindset Graph™, GumGum delivers advertising that drives meaningful outcomes for advertisers and publishers, and is more relevant for consumers.
Industry: Technology
Location: United States
