How To Integrate with Metamarkets in Under 72 Hours
April 26th, 2016 Eric Albanese
Integrating a service with a new tech vendor can be a long and painful process — but it doesn’t have to be. When potential customers are evaluating Metamarkets, we are often asked about the integration process and how long it takes. The answer is easy: we can put data at your fingertips within 72 hours.
Metamarkets runs the largest Druid analytics stack in the world, enabling us to process hundreds of billions of events per day and deliver real-time results through an intuitive, point-and-click interface. As a result, our customers can quickly and accurately understand their business through data while internal engineering resources continue to be focused on developing products that drive their core business.
If you’re reading this, then you’re likely considering working with us. To help further solidify your decision, we’ve put together a few recommendations to ensure your integration with Metamarkets is efficient as possible:
0. Commit a Resource
Before getting to any actual work, it’s very important to ensure that you have proper resources prepared and ready for the upcoming integration. During the integration, you will have a dedicated Metamarkets Data Engineer whose highest priority will be to help you get up and running. We have found that the single most significant variable affecting integration time is preparedness; most of the work required for integration comes at the very start of the project. We find that that having a dedicated tech lead on your end will help ensure a speedy onboarding.
1. Nail Down Core Functionality First
As you’re building your integration, you may run into some data content that is difficult to format and stream to Metamarkets. If the amount of work needed for some dimensions or metrics looks like it may cause your team to miss integration deadlines, skip it! The Metamarkets platform can meet several different use-cases for users, and some of them can be addressed without every single data point you may have. Dashboard building is an iterative process, and your dedicated Metamarkets Data Engineer will be available and ready to engage for any subsequent phases and additions!
2. Follow OpenRTB When Applicable
OpenRTB is a protocol created by buyers and sellers of digital media. Metamarkets is a big proponent of OpenRTB and we highly encourage all to adopt this standard. That said, Druid and the Metamarkets interactive analytics suite is highly customizable– we can support non-OpenRTB formats– but pre-defined OpenRTB 2.3/2.4 pipeline templates will greatly speed up your integration. In fact, our integration SDK was designed specifically for ad-tech companies operating on OpenRTB and we find that our fastest and smoothest integrations are with clients who use the tools we’ve developed for them (more on that below).
3. Send Data Samples Early
As we work together to build the integration, the sooner you send over sample records and the associated business logic to Metamarkets, the faster we can start building the data pipeline. Even with just a few records, your designated Metamarkets Data Engineer can begin setup and unit testing against the samples. Once you’re at a point where you’re able to stream data, a working data pipeline and dashboard will be ready for you.
4. Leverage the SDK
Metamarkets has two SDKs that we strongly suggest clients leverage in order to get up and running quickly with the Metamarkets platform. First, we have developed RadTech Datatypes , which will transform your OpenRTB formatted data into a properly formatted event record ready to be processed by Metamarkets. Second, we created the RDI Client, which will assist you in posting data to our endpoints.
5. Ensure Quality Data
Before creating any dashboards, all data should be accurate and matching to your internal systems. Any data sent to Metamarkets should be properly formatted in JSON format. ID’s used to join various event types should be unique for each individual record. Events sent should follow a standard “funnel”; for example, any Bid Responses that have an associated Click should also have an associated Impression. As the popular saying goes, “Garbage in, garbage out.”
The Data Engineers at Metamarkets are the foremost experts of the Druid database and spin up new instances with astounding speed. With some upfront prep work and planning from your team, you can leverage our expertise to bring the industry standard in reporting to your business. We look forward to working with you and helping you instantly see what’s happening in your business, discover trends, and take action on your data.
If you enjoyed this and have ideas for how to achieve greater speed/accuracy/flexibility, we encourage you to join us at Metamarkets.
Eric Albanese is Director of Data Engineering at Metamarkets. He’s responsible for ensuring the stability, scalability and performance of Metamarkets’ big data ecosystem – including data collection, pipelines, ETLs and the data warehouse.