8 billion requests per day and zero slowdowns: How SensioLabs scaled a Symfony-powered ad platform with Blackfire
How do you keep a Symfony app blazing fast under 8 billion daily requests? In this success story, discover how SensioLabs helped a leading ad-tech platform push the limits of performance and stay at the top of the ad tech game.
Note from the editor: This article was written by the team at SensioLabs and published here on the Blackfire blog as a celebration of a shared success. Together, we helped one of the most demanding platforms in the ad tech space achieve world-class performance. We’re proud to highlight SensioLabs’ deep expertise and showcase how Blackfire empowers performance-first engineering teams.
When scale becomes a battlefield: performance as a high-stakes mission
Imagine you’re in the final lap of a Formula One Grand Prix. You’re leading, but only by a fraction of a second. In the pits, your team executes a flawless stop, just 2.2 seconds. Your closest competitor? 2.4 seconds. A few milliseconds difference. But even in a 2-hour race, that gap could mean the difference between spraying champagne on the podium… or watching from the sidelines.
That’s the world this global ad-serving solution operates in. In the ultra-competitive ad tech space, they aren’t reacting to performance issues; they’re proactively chasing perfection.
Their mission-critical platform handles a staggering 8 billion ad requests per day across 2,000+ websites. At this scale, every millisecond shaved off response time isn’t just a nice-to-have; it’s a decisive edge.
Like elite athletes at the top of their game, this fast-growing player in programmatic advertising knows that marginal gains matter. A few milliseconds faster can mean smoother real-time auctions, happier advertisers, and a healthier bottom line. And in this high-speed arena, standing still is the same as falling behind.
To stay ahead, they brought us at SensioLabs, the creator of Symfony (and Blackfire!) and a team of long-time experts in PHP performance. Together with Blackfire, the observability solution built by and for developers, they formed a performance-first dream team.
The goal? Not just “good enough” performance. But continuous, measurable improvement. The kind of performance tuning that only top-tier engineering teams dare to pursue. The kind that separates contenders from champions.
Uncovering the invisible: Diagnosing performance under real-world extreme pressure
Our customer invited us on-site for two multi-day performance workshops. From the start, we initiated our consulting solution, Performance-Oriented “Symfony Coaching”, a proven method to help development projects attain peak performance. This wasn’t just about chasing numbers; it was about keeping a top-tier ad platform at the very top of its game.
Some applications have sprawling architectures, with countless endpoints and unpredictable usage patterns. In those cases, production-level Monitoring or Continuous Profiling helps pinpoint hotspots. But this project was different.
This global ad-serving solution’s system had a compact, highly targeted footprint: just a few key endpoints, each under immense pressure and packed with sophisticated business logic. That meant every slight inefficiency had the potential to snowball.
We started by running deterministic profiles on those endpoints. Blackfire’s call graphs are brilliant at visualizing execution paths, revealing where resources were truly being allocated, and identifying the “critical path”, the chain of function calls that consumed disproportionate amounts of time or memory.
The process always starts deep: we dive into the low-level dependencies at the bottom of the graph, then trace our way up the call tree until we find the triggering code in the application layer.
In our first workshop, we hit gold. A Twig extension, used to obfuscate and protect JavaScript code, was quietly consuming a massive chunk of compute resources. Refactoring it led to a jaw-dropping 43% reduction in response time on a critical endpoint.
That kind of gain isn’t luck. It’s the result of knowing where to look, having the right tools to see clearly, and the experience to take action.
Because when you’re fighting for milliseconds, staying ahead means never settling. That’s how our customer, a mission-critical ad-tech solution, keeps leading the pack in the high-speed world of ad tech.
Expert moves under pressure: Engineering the turnaround
The second workshop followed a similar approach: deterministic profiles, real traffic, and a relentless hunt of hidden inefficiencies. And once again, the process delivered.
This time, our focus turned to a critical part of the application logic, the advertiser eligibility computation. In online advertising, not all ads are suitable for every website. Publishers establish precise guidelines to ensure relevance and maintain brand integrity. It makes sense from a business perspective, but technically, it creates a tough challenge.
Behind the scenes, the platform needs to compute a fine-grained matrix of applicable advertisers for every website it serves. This isn’t just a lookup. It’s a multi-dimensional calculation performed constantly and at scale. The faster you can eliminate ineligible advertisers from the pool, the quicker you can generate results.
By running deterministic profiles in production, we quickly identified that this process was hogging resources more than expected. With that insight, we collaborated with this performance-driven ad platform team to rework the underlying logic.
The result? A 24% reduction in execution time for this key computation. That’s not just an optimization, it’s a competitive advantage.
Each of these improvements was verified with Blackfire before being shipped to production. The visual diffs made the gains measurable and gave the team confidence to push changes without fear, even in high-stakes areas of the codebase.
In a performance-first culture, iteration isn’t just tolerated. It’s engineered.
Achieve performance that sets you apart
At SensioLabs, performance isn’t just something we fix. It’s something we relentlessly pursue. It’s built into our DNA. With deep expertise in PHP and Symfony, and the right tools like Blackfire at our side, we help engineering teams see what others can’t, act where others hesitate, and iterate with confidence.
Because great software doesn’t happen by accident. It takes focus, precision, and a willingness to push beyond “good enough.”
We’re proud to have partnered with this leading ad-tech platform to help them stay at the top of their game. And we’re even prouder to say: these kinds of results aren’t one-off miracles. They’re what happens when the right tools meet the right mindset.
If you’re ready to join the very small club of world-class, performance-first applications, we’d love to help you get there.
Let’s raise the bar. Together.