.png)
Here at Button, data influences everything we do, from frontend UX improvements to retail media to revenue-optimized AI-powered routing models. As a company, our culture is incredibly data-driven – everyone from our software engineers to our commercial team to our C-suite reviews the data on a daily basis to ensure they’re consistently making high-quality decisions.
It should be no surprise, then, that we regularly revisit our data stack to ensure that we’re leveraging the best tools available. As we mentioned in our article about data democratization, we migrated from Tableau to Looker back in 2020 to empower our users with self-service analytics capabilities, and we remain dedicated to fostering a culture where everyone in the company is empowered to use data to guide, automate, and accelerate their work. But now, five years later, we’ve embarked on another strategic shift.
Button has experienced enormous growth in recent years. In April 2024, we shared that we surpassed $1 billion in monthly commerce for the first time, and that number continues to grow. Last month at Cannes, we announced Creator Media, a feature that enables Creators to seamlessly incorporate retail media into their campaign creation process. Given Button’s continued growth in scale and complexity, it became imperative for the Data Team to proactively ensure our tools met the company’s evolving needs.
High-performing data teams regularly revisit their stack to ensure it aligns with organizational goals. For us, the primary objectives were clear:
In short, we needed faster insights delivered when they were needed, while also carefully managing our budget in order to maximize investments in product and engineering improvements. A tall order, but we were up for the task.
In December 2024, we assessed fifteen different BI tools against criteria including development speed, data democratization, cost efficiency, performance, visual customizability, maintenance overhead, and external reporting capabilities. We evaluated a broad variety of tools: legacy tools with 20+ years of history and more modern tools with only two; tools with dedicated semantic models and tools that were SQL-only; open-source tools and proprietary ones. Importantly, we gathered input from stakeholders across product, data engineering, analytics, and data science to ensure we were meeting the needs of the entire company, not just our own.
By January, we began working on a migration strategy for our top choice: Omni. Omni stood out to us due to its flexibility – it allowed the team to build not just from a centralized, shared data model, but also directly from SQL queries and Excel-like functions. This naturally fulfilled our need for faster, more impactful BI development. We believed its functional similarity to Looker would also significantly reduce migration effort and help us avoid a steep learning curve for users. Moreover, Omni offered substantial cost savings, freeing up resources for reinvestment into R&D and product innovation.
Changes like this always require overcoming inertia, but we recognized early that the sooner we committed, the better it would be for us long-term. After thorough due diligence – consulting reference customers, leadership vetting, and negotiating favorable contract terms – we finalized our commitment to Omni in February 2025.
For as much as we prepared, we knew there would inevitably be risks, and our priority was to mitigate those risks proactively. We were initially cautious about how our stakeholders would adapt. Our commercial team, for example, relies heavily on immediate and frictionless access to data insights to build partner trust and inform strategic decisions. To facilitate a smooth transition, we organized live demos, set up a dedicated Slack channel, and scheduled regular office hours. Shearwater Data, our Omni onboarding partner, also conducted tailored training sessions for users.
In the end, our careful preparation paid off. Omni’s user interface, as we had hoped, was intuitive for users and made the transition remarkably smooth. Instead of questions about missing features, we got questions about new ones and all the incremental ways we could deliver value. The impact was immediate: faster turnaround times for critical insights, substantial cost savings redirected toward innovation, and enhanced agility in responding to evolving business needs.
Lots of people have lots of opinions about business intelligence (BI), but permit us one more: good BI is empathetic, empowering, and egoless. It’s about understanding stakeholders’ use cases as intimately as if they were your own. It’s about equipping them with the tools they need to make exceptional decisions, consistently. And it’s about investing effort in areas that will positively impact the business, its partners, and its users, regardless of what might be most exciting, interesting, or flashy.
We have a lot of pride here at Button about our ability to harness data to make exceptional decisions. Proactively evaluating data infrastructure is one of the many ways we can prove our ability to strategically shift on behalf of the business. This isn’t the first time we’ve made changes to our data stack, and it won’t be our last, but as it stands today, we’re poised and ready to deliver impactful insights faster and more effectively than ever.
Whether you’re looking to maximize conversions, drive app downloads, or access new inventory, we have a Button for that.
get started
.png)
.png)