As a lifelong car enthusiast, I'm amazed at the increased power that comes with each vehicle's next generation. The same can be said for analytic diagnostics, thanks to increased use of NoSQL databases and customer devices that require ever-quicker web page loading.
In the last few years, superfast page loads have become imperative in capturing the dollars and eyeballs of consumers arriving at websites via mobile devices. This means website developers must code their pages to fit the requirements of the least-capable devices -- no detection of flash or Javascript, for example. And with server traffic rising, site operators must monitor load failures and other page performance measures closely.
Enter open-source scalable databases. Some of the latest digital marketing applications and solutions have incorporated a number of NoSQL formats. Foursquare is a well-known MongoDB user, for example, and Android and Facebook applications often use Apache CouchDB. NoSQL databases are gaining in popularity because they offer the scalability required for real-time processing of complex datasets. This makes them appealing to those handling repetitive ad-hoc queries of a website performance optimization tool -- as well as those supporting corporate big-data environments.
Page load optimization tools have gained analytical capabilities. Two in particular, Yottaa and Pingdom, also create an opportunity for improved benchmarking.
Yottaa has touted its use of MongoDB for its online page load measurement tool. This fall, I attended a presentation Yottaa delivered at a New York MongoDB meetup. As a content delivery network, it loads websites from servers all over the world, and it tracks 6,000 websites with up to 300 sample URLs a day. In developing its web performance optimization process, Yottaa said it felt a relational database would be a performance bottleneck for its servers collecting and reporting data, as well as a potential operational bottleneck for sharding data (a topic we've discussed before.)
Pingdom is more established as a website monitoring service providing text alerts if, say, a site goes down. Magnus Nystedt, an analyst at Pingdom, told me it operates with a mix of databases. It uses MongoDB for its online tools, "and we're looking into how we can use MongoDB for other applications as well. Our main monitoring service, however, uses MySQL."
With website complexity increasing, developers must optimize page code to load in many devices and browsers. The databases supporting page load optimization tools can provide more flexibility in arranging data and recalling information for better insights. Pingdom and Yottaa make page load verification easy and instructive; I particularly like Pingdom's waterfall reporting.
Yottaa and Pingdom seem poised to help digital analysts link more nuanced information about website performance to business objectives. Yottaa, for example, said it has found that a one-second delay in page loads can lower conversion rates by 7 percent and customer satisfation by 16 percent. And Pingdom has become well regarded for its periodic analysis of social media community demographics.
Do you know of other companies that have harnessed more insights as a result of their database selections? Share your examples below.
Tweet This
4 
10 
