With Tynt buy, 33Across wants to be a user-data powerhouse

Ad-targeting company 33Across is acquiring link-tracking specialist Tynt Multimedia in an attempt to give publishers a leg up as they compete for advertising revenue against platform providers such as Google (s goog), Yahoo (s yhoo) and AOL (s aol). One of 33Across’ greatest strengths in doing so will be the enormous amount of user data it now has at its disposal.

It is difficult to say exactly how much data the combined company will have, but with a combined social and interest graph of 1.25 billion users — or three-fourths of the online world, more than Facebook’s estimated 800 million users — it is certainly one of the more impressive data-centric operations on the web.

33Across, a marketing platform that lets companies target potential customers based on those companies’ social graphs, came into the acquisition with 200 million of those users and already had a major big data operation under way. As CEO Eric Wheeler explained to me over the summer, “We constantly re-score the brand graph to understand who are the best targets for that ad right now. We use social connections of that brand to know whom we should next target.” In order to do this, 33Across maintains a “massive Hadoop implementation” complemented by machine learning and predictive-analytics algorithms that it has developed in-house.

According to Wheeler, 33Across is processing 40,000 transactions per second.

Source: Tynt

Wheeler said the data implications of the combined 33Across-Tynt graph will be enormous. It is looking at billions of data points every hour in order to drive maximum value, he added. Tynt has an ecosystem of 500,000 publishers for whom it tracks the copying, pasting and sharing of links, and the company has said that people still copy and paste 50 links for every 1 they share via buttons such as those from ShareThis. As of April, Tynt said it was collecting and analyzing 1 terabyte of data per day on its Hadoop platform, although Wheeler isn’t certain of the total size of the companies’ combined data stores.

Going forward, the task will be integrating the existing 33Across and Tynt environments so the latter’s publishing can benefit more organically from 33Across’ Brand Graph. Wheeler said it will let it act even faster and more intelligently based on what its readers are doing, which will help it chip away at the disproportionate amount of online advertising revenue that currently goes to Google, Yahoo, AOL and other platform providers. Those companies are able to sell lots of ad space because they have so many users and can target ads by determining users’ interests from activity across their expansive platforms.