Real-Time Analytics and the Path to “Right Offer-Right Time”

I nearly began to swear recently while listening to a radio interview because an overconfident hype-slinger promised we’d all have location-aware, personalized marketing pushed to our smart phones in a week or two. The audacity of his assurances irked me; they are easy to make to lay audiences that can’t refute them. Those in the know realize we’re still a few steps away from true “right offer, right time” capabilities. Comptel’s Antti Koskela and I had a spirited chat about this issue in the context of Comptel’s acquisition of Finnish real-time analytics developer Xtract.

An analytics play makes sense for a company like Comptel because, says Koskela, the company was basically missing an analysis, or analytics capability in its portfolio. But before we get to complex personalized marketing, analysis can improve churn prediction and prevention. That’s proven and practical. We all know that small reductions in churn metrics equate to big bottom line contributions. Koskela says the idea is to “bring a much higher degree of analytics based automation in business processes” where the data coming out of a mediation process can be used “to predict churn from the bottom up.”

To achieve “right offer, right time,” however, several pieces need to evolve. Koskela says that as more OTT players enter the market “the complexity increases” and operators need multiple SLAs; excellent catalog based order management so that services can be assembled quickly from readily-available components; strong business policies that govern how those components are used; and analytics that determine the right courses of action in everything from QoS and billing to timely promotional offers.

But analytics itself is a loosely defined term. It’s become a catch-all phrase for anything that crunches data. But a big data warehouse that spits out reports is not the same thing as an application that diagnoses inefficiencies in highly complex business processes. And neither of those is the same thing as the real-time, customer-facing analytics needed for “right offer, right time.” As an industry, we’re doing these sub-segments injustice by calling them all “analytics” and otherwise failing to define their taxonomy.

Koskela offers the term “Intelligent Customer Interaction” to describe the real-time analytics needed for “right offer, right time” and says that many analytics solutions process “stuff that is somewhat static.” There’s a place for each in the world, but the static data-crunchers are fairly common today whereas the real-time engines, like Comptel feels it is acquiring with Xtract, are far less common.

Now, all of this heavy engineering talk gets me thinking about IMS. When IMS was hot – before it stroked out – proponents insisted it was needed to do a lot of the orchestrated, over-the-top kinds of services we see today. But then things like Skype, YouTube, Hulu, and thousands of apps blew that assumption away.

So – is “right offer, right time” any different? Was the radio prognosticator right? Do we really need all of this complex gear to make it work? Well, if we can’t identify and correlate some very specific data about users, location, time of day, network quality, and service availability in real-time, then we’ll probably just end up with irritating smart phone spam instead of useful, timely information. Koskela notes this is already a problem in Asia where “people stop reading SMS because they are being spammed with marketing; it’s not a ‘right time, right offer’ kind of thing.”

But, he adds that while IMS is not particularly relevant for end user services, it did drive “intelligence into the packet core layer” along with “policy control and online charging.” He says that “having service awareness in the underlying plumbing” enables a lot of what’s happening in the “higher layers of the business architecture on top of it.” And ultimately, this is what operators do well.

If operators can tie the key pieces together – real-time fulfillment; mediation; policy-control; and real-time analytics – it should set the stage for innovators in the business layer to create solutions for “right time, right offer” personalization (and who knows what else).

Share this:
Linkedin Twitter Email
About Edward Finegold 122 Articles
Ed is now Director, Strategy for NetCracker. Previously, for 15 years he was a reporter, analyst and consultant focused on the OSS/BSS industry and a regular contributor to BillingViews.

Be the first to comment

Leave a Reply

Your email address will not be published.


This site uses Akismet to reduce spam. Learn how your comment data is processed.