Community Spotlight: smart advertising from Sage+Archer + Apache Druid

Out-of-home advertising has changed. Gone are static, uncompromisingly homogenous posters, replaced instead with bright and fluid installations. Installations that make smart decisions about what and when to show content to reach the right people with the right message. Advertisers have unleashed a whirlwind of creativity in physical and digital spaces, assured their efforts – and investments – will be rewarded with targeted visibility and higher engagement.

Sage+Archer meets Apache Druid

In Germany, the Netherlands and the United Kingdom, 70% of all digital screens for “out-of-home” advertising are connected to a single product: Sage+Archer. They reach millions of people through JCDecaux, Axel Springer, the Guardian, and Bauer (to name but a few), partnering with supply-side platforms that include Google and Rubicon (a fellow Apache Druid user).

So the Apache Druid Community Team at Imply were grateful to spend time with Edwin Metselaar, the co-founder and technology lead at Sage+Archer to find out more about their use of Apache Druid.

“We were one of the earliest adopters of Apache Druid. Since we picked it up in 2015, all data analysis has the Druid database as its source – both mobile and digital out-of-home advertising.

“It’s the single source of truth.”

Demand-side platform customers are renowned for being themselves demanding. Sage+Archer have spent significant effort creating the perfect user experience, not just the look and feel, but the accuracy and freshness of the data it presents.

“That’s what our customers expect!”

And no wonder: Sage+Archer’s client list includes Disney, Ben & Jerry’s and citizenM.

The Sage+Archer deployment

Behind-the-scenes, Druid isn’t just fed high-reliability data in bulk overnight (Hadoop), it’s also constantly fed real-time data from event streams (Apache Kafka) – a touch of data magic that keeps customers – and Sage+Archer themselves – ahead of their competitors.

“It was very easy to connect Kafka and Druid. It’s very easy to say, ‘okay, let’s connect to Kafka and read out this topic from this point’. And off it goes!”

Apache Druid is hosted in a hybrid way – a combination of local and cloud infrastructure. There is a mix of OpenStack resources and Google Cloud storage. Despite the latency you may well expect this to introduce, Druid’s behind-the-scenes processes ensure query performance is unaffected.

So critical is Druid to their operations – and to the success of their own customers – that all of the Apache Druid database is fully protected against disaster.

“It’s where we base our campaigns. It’s where we optimise pricing. It’s the data behind dashboards. Everything. All the reporting we do is powered by the Druid database.”

Up front lies a custom interface together with Metabase, giving customers a hyper-fresh and flexible data experience that modern decision makers need. Meanwhile, engineers have the freedom to use Druid’s own in-built SQL interface and APIs when more technical work demands it.

Edwin remains an active participant in the community and, like many adoptees, continues to look for ways to optimise and configure their deployment to squeeze out every possible ounce of performance. We thank the team at Sage+Archer for letting us know about their journey!

*      *      *

Email the community team at community@imply.io to schedule your own Community Spotlight, to get the latest information on community events across the world, to hear about speaking opportunities, and to find out how to get your company’s name in lights on Apache Druid’s Powered By page!