With a little help from some algorithms and machine learning, marketers may soon be able to tap into the power of Hadoop, a programming framework that supports the processing of large data sets.
Hadoop can not only quickly identify relationships and activate rules, but can easily enhance user experiences and deliver more relevant content to the right user on the right device at the right moment.
I recently spoke to a colleague, Christopher Harmon, principal at theBatstudio, about how exactly he thinks this will play out.
For his part, Harmon called Hadoop an enabling database technology that allows users to store information in a way that they can action off of it very quickly.
So what does this mean for marketers?
At any given moment, marketers have multiple programs running concurrently across various platforms, like emails firing off through marketing automation programs and display or paid media programs firing off through various DSPs, as well as ongoing social media campaigns and websites where they are trying to convert visitors, Harmon said.
“Right now, the ability to coordinate between those programs is limited. Many of those mediums have point-to-point connections and if someone responds to an email through the marketing automation system, you might be able to alert the display/retarget program to send that person ads around whatever they responded to.
You might have the ability to customize content to this person if they visit my website. But what I don’t have is a program that sits on top of all of my tools – social, email, website, etc. – so that I can coordinate one user experience across all my different channels of communication.”
And that’s where Hadoop comes in.
In other words, Harmon said marketers can theoretically feed user data from all of their various systems into this database. They can then have an algorithm look for connections to identify users who, say, have indicated interest in specific content, so it can go ahead and activate certain rules and push out specified follow-ups to various channels.
And, in turn, those users will theoretically be more interested in this content and more likely to consume it.
“That gets really interesting. You can automate communication paths and where Hadoop is interesting and valuable is in its ability to store that data we care about at a high level and to quickly run rule sets to identify opportunities to take action.”
And it may only get more nuanced from there…
Further, this ability to take large data sets, quickly identify relationships and activate rules could enable marketers to better automate their interaction with consumers and deliver more relevant content – even on a more customized basis.
What you do with that activation depends entirely upon whatever set of rules you have in place. What I mean by that is we could say, ‘Hey – if three individuals from Company X interact with the same kind of content anywhere in our systems – opening an email, clicking a display ad, visiting a web page, responding to a tweet – if three individuals from a company do that, take all the people in the company we’ve identified and send them an email nurture and retarget them in display campaigns, or tweet them with an interesting article.
So the ability to use what some people refer to as digital body language, or signals, to activate some set of programs that you have sitting on the shelf, ready to go, [could be enhanced]. Right now, today, again, there are some point-to-point connectors that enable you to do this in an automated way, but few vendors have done this across the entire marketing architecture.”
The future of targeting with Hadoop
What’s more, Syntasa is developing systems to expose relationships and attribution in data sets and eventually enabling their clients to act on them even more strategically.
Instead of a simplistic rules-based architecture that is static and stays in place until marketers make changes manually, Syntasa has a deep understanding of algorithms and machine learning and has built an application layer.
This layer sits on top of Hadoop and can actually learn and see if tweaking a rule one way or another impacts overall effectiveness – and then automatically update those rule sets based on actual user behavior.
Further, this means companies don’t need to rely exclusively on internal resources to create and manage rule sets and evaluate which ones are working and adjust them accordingly. And that should be a welcome relief to already understaffed marketing departments.
“If a system can automate the process of trial and error without a human having to go through setting it up, that’s really interesting,” Harmon added.
While ad fraud has become part of every marketer’s vocabulary, attribution fraud—the practice of gaming outdated attribution models to justify self-serving means—has ... read more
When you’re just starting out as a business owner it’s easy to become wrapped up in the seemingly endless number of metrics ... read more
Earlier this year, a report from Consumer Intelligence Research partners (CIRP) pegged the sales of Amazon’s Echo at more than three million ... read more
One of people’s biggest fears about artificial intelligence is that machines will rise up and humanity will be obsolete, that we will ... read more