By Barry Flynn, Contributing Editor
ThinkAnalytics, whose content and recommendations engine services over 150 million subscribers worldwide across more than 60 video service providers, is unveiling a new advanced data analytics platform at the CES show, which opens in Las Vegas on January 6.
For Peter Docherty, ThinkAnalyticsâ€™ Founder and CTO, the move represents a natural development for the company, since it leverages existing data assets: â€œWe obviously are experts in the programme metadata side of things, weâ€™ve got a good understanding of customer behaviour, weâ€™ve got all the data on what theyâ€™re watching, recording, and purchasing, and weâ€™ve got subscriber data about what products and packages they have. So without even considering any other data, the engine actually has a lot of data which is quite fundamental to [â€¦] understanding whatâ€™s going on from a business perspective and a marketing perspective.â€
However, argues Docherty, not every customer is currently able to take advantage of such data, even when itâ€™s available to them. â€œThey’ve got a lot of data but they havenâ€™t got a lot of insight and they don’t know what to do with it in order to take any actions that are going to affect the business,â€ he says. â€œTo say that we know what billion recommendations were made and we know what hundreds of millions of programmes were watched â€“ that data is there, but what does it mean?â€
The companyâ€™s new Big Data analytics platform is designed to address that gap, and plugs straight into its existing Recommendations Engine, being available either on-premise, or as a software-as-a-service (SaaS) product hosted in the cloud.
Docherty stresses that while â€œthereâ€™s a huge amount of analysis and insight you can get out of the data the recommendations engine has on its own,â€ the new analytics platform is designed to easily ingest data-sets from other sources the customer might have access to, such as network statistics â€“ drawing conclusions, for instance, about the levels of network congestion that can cause customers to abandon viewing on-demand content.
Such insights can be generated on-the-fly if required. Docherty notes that ThinkAnalyticsâ€™ background, before recommendations, was actually as a real-time data-mining company. â€œWe understand analysis, and what challenges there are whether doing it in batches or in real-time. You can see it from both perspectives. You might want to react very quickly to something thatâ€™s happening, or you might want to look back and say, â€˜right, weâ€™re seeing these patterns and weâ€™re seeing them over days and months, so weâ€™re actually going to change something from a business or operational perspective â€“ because we now understand better whatâ€™s happening.â€™â€
Docherty believes his most likely prospects for the new Big Data product will be existing customers: â€œAll the customers we have on the recommendation engine could benefit from this additional insight, so theyâ€™re certainly the initial targets. We think theyâ€™re going to like this product and get a lot of benefit from it.â€
Indeed, he reveals, â€œthe customers that are thinking about this the most, weâ€™ve already been talking to them for quite a long time about the kinds of things that they would like to see in this kind of product.â€
Besides, as he points out, â€œif you didnâ€™t have the engine and were trying to do this from scratch, you would have to go and put a whole programme in place to go and get all this data â€“ the viewing data and the metadata and the subscriber data â€“ in order to create the equivalent big data analysis tool.â€
Instead, by adding the Big Data product to the existing recommendations infrastructure, â€œyou can get going really fast and you can get a lot of value out of the system, very quickly,â€ while still being able to â€œadd other data, or develop other KPIs and metrics and reports that you want to haveâ€ in the future.