AI and Big Data: B2B Personalization Leaps Forward
By Praveen Balla
Vice President, Production, True Influence
AI is making the classic notion of the marketing funnel obsolete.
That may seem a bit dramatic, but it’s true – Artificial Intelligence (AI) is revolutionizing every aspect of B2B marketing. And perhaps the most significant revolution happening today is the ability to hyper-personalize content to the preferences of a single contact.
With the classic funnel, marketers created core customer personas and made educated assumptions about the kinds of content that would resonate with groups of contacts at different stages of interest and engagement. But they were still sending out basically the same content to sizable groups of prospects. Inserting a name and logo into an email template or sending action-based behavioral response emails is a start toward personalization, but just barely.
AI-driven Personalization Across B2B Funnel
With an AI-driven funnel, you can analyze thousands of data points to determine a prospect’s propensity to respond to certain messaging details across apps, websites, web, etc. Big Data AI and machine learning algorithms, like those in our Relevance Engine intent intelligence, can pinpoint a contact’s exact stage in their on-again, off-again purchase journey. And the data that fuels these AI insights comes from all your prospect’s online activities, not just interactions with your own content.
The resulting ecosystem of data that surrounds every individual makes delivering creative that’s limited to standard personas and funnel stages fairly pointless. Note that I said “limited” – obviously, you have to start somewhere, and your standard personas remain the baseline.
But if your AI-driven ecosystem tells you the CEO you’re reaching out to shares traits and behaviors with outliers who tend to respond to heavily technical content, then that’s what to send that individual CEO. AI lets you find the precise point in the three-dimensional matrix of purchase intent, content preference and buying persona to deliver highly personalized messaging. Simply put, an ad shouldn’t sound like an ad but a message, an addition to the points known by the target user/persona.
Precise Programmatic Ad Messaging
A highly effective application of AI content personalization right now is dynamic content advertisements, which tend to be powered by AI on the DSP side. Again, digital is a highly quantifiable channel, and DSPs serve extraordinary volumes of ads – lots of data for AI to analyze. And ad tech has evolved to the point that personalization options are incredibly precise.
Let’s take the example of the color of a car being displayed in a banner ad. In what I’ll call V2 of AI display personalization, an advertiser could provide 20 or so options for creative. AI would have determined that the viewer has traits in common with people who respond to the color white, and so that is the color car that person will see.
In AI V3, which we are entering now, ad serving tech can simply modify attributes like color on the fly. So, based on what AI can discover or predict about a user, they might see an off-white car on a light blue background. There’s simply no set number of options – the car might be a convertible. Or it might not be a car at all (more on this in a bit).
So, AI can swap out entire messaging points; nuanced messaging points; or small personal preference details, like background color. The combinations are boundless.
Site Content Helper
AI technology not only helps understand and structure content, it finds other content that’s similar and presents it to users, based on the context. This sort of AI, often called semantic, is essential to our Relevance Engine intent data analytics.
It’s also the backbone of “smart” content management systems that tailor site content to individual visitors. If you log into LinkedIn, for example, the content presented to you from Microsoft and other partners is a result of AI finding interest patterns in your reading habits, both on LinkedIn and other sites, and then pushing content to you based on those interests.
Content publishers have eagerly pursued this application of AI for years. The New York Times, notably, introduced a fully personalized e-newsletter in 2018. The publisher boasted it was generated by both human editors and “machine-learning algorithms designed by The New York Times.” It added a personalized site channel in 2019, and is now experimenting with personalized vertical or category pages.
Continuity from Email Push to Product Page
B2B sellers typically apply AI-driven content personalization on product pages that users navigate to, or sometimes on organic / SEO landing pages. If I visit a product category page at IBM or Microsoft, for example, AI will tell it to use versions of content blocks tailored to a Demand Generation Marketer, not to an IT professional.
You can personalize outbound campaign landing pages in similar fashion, but you can accomplish the same goals with successful segmentation and targeting. And there’s something to be said about maintaining continuity of voice and offer across an email or other touch.
Outbound AI-driven Personalization Gaining Ground
At the moment, most personalization of content in emails and other outbound channels is best accomplished via list segmentation. How finely you slice lists is up to you, but marketers certainly should incorporate more than just a basic persona and assumed funnel stage in email offer personalization.
However, the AI revolution is coming to outbound, particularly for sales. This article details how Magic Sales Bot can write five personalized sales emails for reps to choose from, based on some limited criteria they enter about the contact. Obviously, this is at limited scale, and most of the AI “intelligence” is devoted to writing the mail, not building that data ecosystem around a single prospect. But you can see the potential for integrating these technologies.
Having any ideas for AI in your organization? Here’s one: Register for “Accelerating Revenue with AI” on April 14.