DPI & analytics: Creating competitive advantages for advertisers

DPI & analytics: Creating competitive advantages for advertisers

By Jim Crawford, President of Crawford Public Relations

Monday 9 a.m. An agency representative checks her smartphone to see a detailed record of a client’s favorite websites, how often he visits them, what he likes and what he recommends to friends. Probing deeper she can review heuristic, historic data on past buying habits, then switch to a second app that provides real-time insights on the customer’s mindset in that moment. Monday 9:01 a.m. The app then tells her the “next best action” for selling more of the right kind of product to that individual, in the customer’s preferred channel and at the exact level of promotional investment required to close the sale.

Welcome to the world of deep packet inspection and real-time predictive analytics – technologies that work hand-in-hand to create competitive advantage for smart users. Developed independently over the past decade, DPI and RTPA are coming together as essential tools for winning new business, retaining customers and increasing the lifetime value of accounts.

Not Your Father’s DPI

DPI arose from 1990s “packet sniffing” technologies that enabled Internet Service Providers to see the types of traffic transiting their networks. Today, increasingly, DPI is used by advertisers who want to understand customers’ preferences in order to better serve their needs. As rules-based software, DPI can be programmed to provide valuable insights without violating customer privacy.

Just one little hitch: Given today’s broadband explosion, the cost of crunching every bit of data available on a customer can add up quickly. Nobody wants to spend a fortune on the massive parallel processing required to perform detailed DPI on millions or billions of transactions.

“Real-Time” Insights

This is where real-time analytics kicks in. Analytics are the natural offshoot of work done by big data platforms – a separate technology development that evolved to help derive value from massive amounts of data collected. There are two types to be dealt with: data at rest and data in motion.

Data at rest is what you know about a customer. Data in motion is what you’re learning in “continuously updated” mode, a constant mind-boggling flow of information that might include recent purchases, social media activity, plus data from the many sensors we use every day in smartphones, connected automobiles and even vending machines.

How companies are now learning to drive value from data in motion: by incorporating data from DPI in the stream of information processed in real time, for a snapshot of the customer in “real time.”

We’re placing “real time” in quotes because it is a relative term. The speed with which a system delivers real-time analytics and next best actions depend on where the users set its latency – in other words, the amount of time between each analysis of a new data feed. Need to keep costs under control? Simply set system latency at a level that provides essential insights without breaking the bank.

Go Virtual and Mobile

While many RTPA systems still require software licensing, there’s a growing trend for getting away from fixed capex on IT. Increasingly, companies are taking their analytics needs “to the cloud” and paying either a usage-based fee or nominal monthly charge. 

Get it “on the go.”  Some cloud-based providers of analytics offer smartphone apps that put real-time insights in the palm of your hand. When you think about it, that only makes sense. Today you not only need to see what customers think in the moment, but have the power to act on it instantly.

With DPI-fed real-time analytics sent straight to your smartphone, the customer in essence selects the ad they want to see. In programmatic advertising, can it be any better than that?


Jim Crawford is president of Crawford Public Relations. Crawford’s clients include companies that provide advanced analytics solutions to business and government.

Leave a Reply

Your email address will not be published. Required fields are marked *