A Human Writes This Blog … For Now

“Our machines should be nothing more than tools for extending the powers of the human beings who use them,” Thomas Watson Jr. That quote, from IBM’s founder, is on the site of IBM Watson. I’ve spent this past week admiring artificial intelligence (AI) — or “cognitive business” as IBM positions itself…

Our machines should be nothing more than tools for extending the powers of the human beings who use them.
Thomas Watson Jr.

That quote, from IBM’s founder, is on the site of IBM Watson.

I’ve spent this past week admiring artificial intelligence (AI) — or “cognitive business” as IBM positions itself — and the achievements happening in the world of machine learning. IBM Watson Chief Marketing Officer Stephen Gold presented at Marketing Idea eXchange, telling us of the marvels of computing today. If we think our current daily data output of 2.5 quintillion bytes is a flood, wait until the 44 zetabytes data tsunami that’s just around the corner gets here, every day.

Now, 88 percent of this data is unstructured — speech, video, social, according to Gold, much of it beyond the realm of most present-day database analyses and analysts. We know from McKinsey, IBM and even DMA studies that’s there’s not enough talent in the world, never mind the U.S., to analyze it all — to find the patterns and make sense of it, and then to apply that knowledge in faster and faster time.

Enter, AI. IBM is not alone in this quest. Watson may have had its early fun (and success) on Jeopardy, but IBM Watson today has learned a lot since then — how to converse, how to discover, how to optimize decisions, how to personalize and how to analyze. Google DeepMind’s GoAlpha is making its own statement with a Go human challenge — chalk another one four up for the machine. Most certainly Amazon, Oracle and others — anyone with a cloud — are making their way into your mind.

There are implications for society and for marketing. (Is Siri really after my job?)

A former colleague of mine at Harte Hanks used to tell me that automated analytics software is like dynamite — very useful, but only in the right hands. But like Mr. Watson says, it’s only a tool.

We will still need a marketing discipline that is sure-footed, astute-crafted and red-blooded — with young men and women who need to be smart with data, and even smarter with data tools. Marketing success in our business has always been about data, but wow, how that data has changed in volume, velocity and variability!

So I leave with three questions to ponder, for comment and to keep my job a little longer:

Question 1: Can Artificial intelligence fill the talent gap in the world of marketing? I believe the answer is yes.

Question 2: Is AI indeed like dynamite — needing to be handled with care, only in the hands of professionals? (Or is it a democratizing tool, best used in the hands of everyone?) The verdict is still out for me here.

Questions 3: How will AI enter the marketing suite? And which C-level officer will be the first to introduce it in the C-suite? It’s always via the CFO, isn’t it?

Underscoring all of this are ethical implications, too. Much of what we know about risk and data governance comes from a more structured world, but what will we find when we collect immense amounts of unstructured data, and start finding and applying patterns there? Let’s plan for the positive, because there are so many tremendously socially valuable needs which AI can serve (and is serving). Let’s fence the negative, because individual respect, democracy and universality must be preserved, too. And let’s keep humanity in control of the process – because that’s how machines learn in the first place.

The Data-Content Continuum: A Marketing Virtuous Circle

To transform an organization to customer-centricity, to provide prospects and customers with relevant content, and to achieve sustainable marketing ROI all depends on data. We all get that. It may be hard to break down data silos and create a whole customer view — to prompt smarter marketing triggers — but increasingly brands and their agency and ad tech partners are finding better ways to do just that.

Data DrivenTo transform an organization to customer-centricity, to provide prospects and customers with relevant content, and to achieve sustainable marketing ROI all depends on data. We all get that. It may be hard to break down data silos and create a whole customer view — to prompt smarter marketing triggers — but increasingly brands and their agency and ad tech partners are finding better ways to do just that.

We’re data-driven marketers. We develop and depend on data strategies that incorporate data governance and data quality at every information spigot. We strive for analytics to predict behavior, to increase ad (and audience) buy efficiency, and to gain insight in response. We refine continuously. Increasingly, that response isn’t about measuring the impact of this marketing tactic or that ad campaign result, but rather marketing’s overall contribution to enterprise objectives.

Data, check. But what about content?

As channels proliferate, as paths to purchase become increasingly complex — heck consumer wander as they self-select their paths – a tough question remains: Is there enough branded (and non-branded) content available, across channels, to scale 1:1 at a mass level — and still be compelling, relevant and timely?

Well, in a word or two, that’s the very direction brands are going (or trying to).

A recent gathering of Marketing Idea eXchange (MIX, formerly Direct Marketing Idea eXchange) explored the relationship between data and content — led by Velocidi CEO David Dunne and Unified.Agency CEO George Wiedemann.  I share some interesting observation that may inform any data-driven content strategy:

  • Personalization matters — but personas matters more. After all, personalizing irrelevant content is counterproductive, said kindly. Since the funnel can have many points of entry, and many paths to conversion, all the better to have accessible content based on buyer and influencer personas (B-to-B and B-to-C).
  • Big data, small data, which data is most important? The data that provides context. There could be hundreds (even thousands) of data elements across the marketing ecosystem that can be assigned to any one individual. Chances are only a precious few serve to identify context (a locale-based search, for example, a prolonged Web site visit with active engagement there, or transaction history). When trigger-based marketing is based on context, it serves to help define the right types of content to serve the customer’s wants and needs in that moment.For example, luxury spa and personal care products manufacturer used a mix of traditional marketing research (in-store surveys) and market testing to transform its digital presence as a “pampered” customer experience. Innovation resulted: Online beauty consultants on demand, chat features, appointment booking by Web, digital promotions to introduce new services — all working to extend the brand’s exclusivity without sacrificing accessibility to its customers.
  • Can creatives learn to love data — best that they do! “I never met a creative person who did not want their work to resonate [with an intended audience],” Dunne remarked. “Data provides the insight that can unlock their creative capabilities.”
  • How do we create content at scale? As media budgets are optimized, less waste, theoretically at least, frees resources for both persona- and context-driven media buying, and persona- and context-driven content development. Content does not always need to be paid, earned or owned — it can be curated from second- and third-party sources.
  • “Content” also comes in many forms — some of which may or may not be in a marketing budget at all: The call center conversation, the online chat transcript, media placements, analyst reports, social content, customer reviews, direct mail — is all of this being organized smartly around context and personas?

And perhaps the largest observation of all: Marketing is less and less campaign-based — it is more and more continuous. The relationship between data and content is a continuum — always improving each other. Marketing budgets need to reflect dialogue, not spray and pray. We used to jump on the bandwagon. It’s now time to jump in the circle.