AdTheorent CEOインタビュー:モバイルRTBのデータマイニングについて
Mobile ad network AdTheorent has unveiled real-time, predictive modeling software that aims to address the inefficiency inherent in current mobile data mining methods. Called Real Time Learning Machine (RTLM), it can analyze 50,000 mobile bid requests per second.
“There is a problem with the way data mining works,” CEO Anthony Iacovone tells AdExchanger. “First, you feed that system a bunch of information and then run a bunch of different models. The system then helps you home in on what the best model is. However, once you’ve decided on a data mining model, updating it and refining it are difficult. Basically, if you want to lay additional data on top of that existing structure, you have to start all over again from scratch. It’s a complete waste of time, money and resources.”
To illustrate his point about his issues with current data mining methods, Iacovone offers this analogy: Let’s say a person has read chapters one through 10 in a book, but in order to progress to chapter 11, the person has to start all over from the beginning. Considering the low click through rates that mobile already gets, this kind of constant “do-over” is what keeps the mobile RTB space from greater acceleration. In addition to making mobile data mining more streamlined, Iacovone says that the primary selling point is that it can raise click-through rates.
He makes some pretty bold claims, suggesting that the average CTR in mobile verges on 0.4%. “When we had the last version of the RTLM, we saw CTRs go from 0.4 to 1%, and have even seen CTRs as high as 1.4% for some campaigns.”
Now that RTLM is out of beta, it will be up to AdTheorent’s agency partners to register the ultimate verdict as to its efficacy. So far, the company has been working closely with Omnicom media shop PHD and Publicis’ MediaVest early on.
“The value of data in media is immense and intelligent technology companies are now using that data across mobile devices,” says Sal Candela, mobile director at PHD. “This is enabling real-time learning and prediction, which is presenting a great opportunity to move marketers toward data-driven results.
The data mining technology was developed by Dr. Saed Sayad, AdTheorent’s chief data scientist. In terms of the basics of how RTLM works, Iacovone says the application analyzes 50,000 bid requests per-second on a single server, filtering out bids with a low probability of click, conversion or awareness lift. The RTLM system “learns” from incoming bid requests and builds and modifies predictive models as it learns, applying such models in live campaigns to match each mobile advertisement with the optimum mobile impression.
Apart from signing up more agency clients to the RTLM system, which is being positioned as part of AdTheorent’s core product offerings, Iacovone is planning to hire more sales people in major markets such as Chicago, Los Angeles, Dallas and Atlanta. He’s also in the process of deciding whether now is the right time to build a UK business. That decision will be made within the next three months.