IBM’s first Chief Brand Officer, Jon Iwata, really nailed one of marketers’ greatest of-the-moment paradoxes when he opened his speech at the ANA Business Marketing 2017 Masters of B2B Marketing conference in Chicago:
Iwata’s talk went on to show how customers’ willingness to share data about themselves, combined with the latest advances in digital interaction technologies – increasingly automated by artificial intelligence (AI) – heralds a future in which marketers will custom-tailor brand experiences for each individual customer or prospect, at the moment of interaction, instantaneously. Even down to the emotional triggers most likely to move that individual to act.
Really? How near can that future be in world where, shockingly, not one respondent to a September 2017 study of marketing and media professionals could say they fully understand their own marketing technology stack? And only 8% deemed their understanding “good,” while everyone else ranged from “some” to “little.”[i] That survey was fielded in the U.K., but the results exemplify study after study, in every region of the world.
The paradox, you see, is that marketers must master the digital technologies of the Modern Marketing tech stack before they can attain the Post-Modern Marketing future Iwata and others envision – and they have a long road still to go. As Stein IAS Chief Innovation Officer Marc Keating puts it, "The biggest challenge for the future of marketing technology in the Post-Modern Marketing age is that there's still not a lot of companies doing the now."
And yet … a small number of pioneers are, in fact, leading the way toward Post-Modern Marketing’s individualized, tech-enabled intelligent experiences that will immerse customers in creative/content that emotionally connects with them – propelling them to explore the brand and share their experience. We saw that in chapter 3’s studies of “Fearless Girl,” “The Field Trip to Mars,” and “VERYX 360 Experience” – though none of those were individualized one-to-one.
So, what are the key characteristics of the Modern Marketing stack that marketers have yet to master and, most importantly, what does it take to master them? How will that stack evolve in the Post-Modern Marketing age, and how can marketers prepare for it? And – existentially – will marketers who don’t move quickly enough to master modern martech and embrace Post-Modern Marketing ever catch up with competitors who do?
Ironically, the fastest path to success in marketing automation (MA) is to start small. Note: that does NOT mean go slow. But hitting an existing business or marketing culture with MA’s maximum potential in one fell swoop is certain to cause total paralysis.
“When I mention marketing automation to a client for the first time, I can see the fear instantly light up their eyes,” says Stein IAS Chief Innovation Officer Marc Keating. “They’re thinking, ‘Oh my god, this means I have to coordinate marketing with sales, and maybe customer service, so I have to get senior management on board, bring in HR, impart training, get the web team involved, get IT involved because I’m capturing data, and now that means GDPR privacy considerations.’ The only path through that fear is to start with a small pilot project.”
Another absolute requisite is organizational aspiration to achieve total digital transformation, preferably embodied in a CMO looking to be a game changer. For many large enterprises – the main organizations with the scale and resources to fully realize MA’s potential – that may mean millennial CMOs, or younger. Someone who’s main understanding of the world comes through advanced digital communications.
Here is the step-by-step high-level plan for Modern Marketing automation:
Step 1: Pick a small pilot campaign, in one region, with a small team of people.
Step 2: Align management and sales stakeholder goals with marketers’, making sure each is clear on the short- and long-term MA benefits to expect. For management, MA brings lower cost, higher productivity and, most importantly, superior visibility into the pipeline which results in far better forecasting and decision-making. For marketing, it’s about much better effectiveness through automation and the ability to prove ROI and revenue generation from marketing investment. For sales, it’s about better leads that drive their success faster and higher.
Step 3: Define the outcomes you seek, in terms of pipeline or sales increase, loyalty, retention or growth.
Step 4: Develop the blueprint for achieving those goals, defined in terms of the personas that influence or decide purchase; the different buyer journeys each of those personas pursue; the content that will influence them, rationally and/or emotionally, at each stage of each persona’s journey; and the trigger events that indicate transitions between journey stages. This is the essence of the “nurture stream.”
Step 5: Test your assumptions about personas, their journeys, and their content needs via digital body language (DBL) analysis and predictive modeling, using the results to refine your personas and enhance all levels of the blueprint. (See the Predicting Qualified Leads sidebar for further detail.)
Step 6: Get consensus from all stakeholders on the initial lead-scoring approach, and what characteristics constitute a qualified lead. Plan to make this a “living document” that evolves during execution to embrace new insights into prospect behavior.
Step 7: Begin execution, i.e., program the blueprint into the MA system you’ve chosen, commission the content, build the initial audience database and/or launch awareness campaigns that draw people into the nurture stream.
Step 8: In 2-to-4 weeks, plan a round of campaign optimization based on immediate results. At this stage you can assess email open rates, click-through rates, subject line performance, performance of key content assets, etc., enough to make adjustments to tweak performance
Step 9: After six months or so (actual timing is based on the duration of your target’s expected buyer journey), analyze results against stakeholders’ goals to see what benefits accrued, and what learning you can use to improve/optimize the nurture stream. If you’ve integrated your CRM, you can evaluate stats like pipeline value and revenue closed.
Step 10: Based on the attributes of your successful pilot, build a business case for expanding MA to more business units and regions.
Step 11: Consider connecting your MA system to data management platforms (DMPs) such as Oracle’s Bluekai, which opens a new world of opportunity to use intelligence from the MA platform to trigger custom banner ad construction for different personas, in real time; consider connecting to your website for the same opportunity.
Keating says to expect it’ll take at least two or three years to “go to 11” – i.e., reach a Step 11 maturity level, which allows the full potential of MA to be realized. Meanwhile, though, the business will benefit at every level through better quality leads, more customer conversion, and increased customer and pipeline visibility that supports superior strategy development and management decision-making.
Marketers have a simple goal for martech: get the right message or content/offer to the right person, in the right context, at the right moment.
Today, that means segmenting buyer audiences into personas representing broad groups; many personalize further, perhaps by digital body language, by named account or account profile, by purchase intent, by geography, etc. Then, buyer journey paths for each segment must be pre-defined and embedded into a marketing automation or programmatic advertising system using if-then rules that can’t change much once they’re programmed in.
Iwata called this crude, but most marketing organizations are not doing “crude” all that well – or at all – by their own admission. In Act-On and Econsultancy’s 2017 State of B2B Marketing Automation study of 355 B2B marketing professionals, only 7% rated their use “very effective.” Similarly, Stein IAS’ Digital Marketing Maturity Index concluded that only 6% of the nearly 400 marketers surveyed could be considered “highly mature modern marketers.”
Ironically, the key hurdles to highly effective use of martech don’t include the tech; they have more to do with marketing mindset and state of organizational transformation. Even though marketers may buy an Oracle Eloqua system, for example, they typically still think with a “campaign cycle mentality.” That means they manually execute a linear series of touches, get the results and factor any learning into their next cycle. Eloqua, however, makes possible automated, multistage, multidimensional nurture programs in which all stages and dimensions can execute simultaneously, in response to the digital body language of each audience member. But few have wrapped their heads around the process transformation this requires for marketing operations.
The organizational transformation hurdle is that even today, most large enterprises remain siloed – whereas effective use of Modern Marketing technology increasingly demands close collaboration among marketing, sales, customer service, and even product development. One implication, notes Scott Brinker, editor of the Chief Marketing Technologist Blog and program chair of the MarTech conferences, is that the martech stack must expand and integrate with a variety of other enterprise technology systems.
Yet, we can already see tomorrow. John Ellett, author of The CMO Manifesto, described it this way at a February 2018 ANA Business Marketing meeting in New York:
Stein IAS’ Keating shares that vision, but acknowledges that it presents B2B marketers with a daunting question. “If you’re still struggling with the modern marketing transition from linear thinking to a couple of nurture stages and a handful of dimensional attributes, what happens when you’re called to design an orchestrated AI system and feed it several orders of magnitude more variables and data?”
Further complicating the looming impact of AI-orchestrated Post-Modern Marketing is that AI is more than a single solution you can just buy, like marketing automation or CRM. “It’s a whole new layer that marketers must figure out, and then integrate into all the other tools in their stack,” says Keating. “You can buy the AI layer from a vendor – for example, IBM Watson, Salesforce Einstein or Adobe Sensei – or you can build it up yourself with point-products that have AI at the core, like Lattice Engines’ predictive models. Either way, it has to be integrated into everything else to be effective.”
The point being that, at least in the early years, AI-orchestrated marketing will require marketers to design and integrate systems themselves – although IT teams will need to collaborate and support. Only later will more turnkey solutions emerge.
Martech experts all agree the move to Post-Modern Marketing technology will happen even faster than the Modern Marketing wave did, and faster than most people can comprehend.
Keating believes five characteristics of marketing technology are key to the rapid evolution toward Post-Modern Marketing tech – and to delivering what excites Iwata and Ellett: one-to-one intelligent experiences across the brand-demand continuum.
Recombination. Today, creative and longer-form content are mostly built as complete executions, sometimes with multiple versions developed for each persona or segment. But a handful of marketing AI players have illuminated “tomorrow” by beginning to “atomize” creative: breaking complete executions down into smaller objects, like headlines, images and offers. These can be mixed and matched – i.e., recombined – by automated systems to create a greater number of executions differentiated by any number of rational and emotional triggers.
Part and parcel of the Post-Modern Marketing movement to integrate more emotional messaging into modern marketing, Adobe uses deep learning from its AI and machine learning platform to identify and detect approximately 35,000 unique tags/keywords in images (e.g., family, holiday, motion, age, mood), as well as some higher-level emotional concepts and themes like “happy,” “love,” “anger” – all based on the context of the image in its roughly 100 million Adobe Stock image assets.
Chris Duffey, Strategic Development Manager for Creative Cloud at Adobe, notes that, “Adobe Sensei is continuously learning to have a deeper contextual understanding of the emotions happening in an image. Within Adobe Stock, Sensei is actively working on the prediction of emotions for people and images. In addition to a broad set of object identification features, it can assess a picture of two people hugging, for example, which might have a tag of love since other similar images have been tagged with love. These AI vision recognition capabilities are ultimately enabling the orchestration of micro-targeted personalized digital experiences at scale.”
“Tomorrow’s automated tools will need to adapt the buyer journey path at every stage and turn. This will be driven by profile data and user engagement – their digital body language. Instead of being limited to a single channel, there will be a new kind of ‘adaptive campaign canvas’ that operates across all channels and adjusts in real-time as a buyer interacts with brand content anywhere in the world.”
Automation. “Today we have marketing automation, but very little is automated apart from pre-planned steps in the buyer journey,” says Keating. “Near-future marketing automation platforms will have adaptable campaign canvases and experiences that build themselves. As with recombinant content and creative, marketers will define ‘experience blocks’ that contain messages/content/creative and are tagged by buyer stage, profile data, and digital body language. A self-learning AI platform will compile disparate elements together to create the ultimate real-time, adaptive, optimized experience for each user, curated via analysis of big data captured within programmatic ad-buying platforms and marketing automation platforms.”
Self-learning. Today’s pay-per-click and programmatic display systems can optimize a message or a budget to increase conversion; a web page or call to action can be shown more if it converts. But these are siloed optimizations using data usually captured from within that channel. The Ingredion system referenced above went further, integrating web and email channels with sales data. That lets the marketing team perform manual analysis of data, enabling them to go back and optimize marketing efforts – about as state-of-the-art as you can get with today’s tools.
An implication of all five of these characteristics is scale. For the foreseeable future, there will be a sweet spot for AI in larger enterprises that have the data and the global scale to justify the investment needed to stitch together all the different pieces required to deliver the return.
Looking further out, standardized interfaces, open source software, and public data libraries will eventually make AI-orchestrated Post-Modern Marketing available to all business sizes – assuming they survive the initial competitive onslaught. That more egalitarian future will likely be a decade or more in gestation.
Predicting Qualified Leads
Of the many ways marketers are failing to get the most out of their Modern Marketing tech, the biggest may be failure to exploit the digital body language (DBL) their marketing automation systems gather about customers and prospects.
Doing so will get you a double win: it’s the easiest way to gain firsthand insight into the potential value on offer from AI/machine learning. Feeding customer DBL into new predictive AI modeling engines (such as Lattice Engines) can yield insights into the true size of your buyer universe, what organizations are most likely to buy from you – and even which of those organizations is actively seeking solutions like yours, right now.
Predictive modeling engines start by combining hundreds of different data sources. Their resulting models of a buyer universe can identify all the standard stuff – industry, geography, workforce size, revenue, number of locations, etc. – but also go much farther. When such an engine analyzes your customers’ DBL, it identifies consistent patterns associated with the types of organizations that become your best customers (rising or falling stock price or profit? Female CEO? HQs in Silicon Valley, Boston and London?), as well as the buyer journey that brought them to purchase. Those patterns can then be used to identify the most likely buyers who are already in your marketing automation system – PQLs, or predictive qualified leads.
But then those patterns can also be applied to the universe model to identify the organizations most likely to want your solution. Looking deeper, the model can determine whether workers at those organizations are downloading content relevant to your buyer journey, thus identifying those organizations actively seeking your solution.
The advances described are anticipated over the next three to five years, so it’s clear that the martech stack will be changing rapidly. Keeping pace with martech’s advancement will remain a major challenge for marketers.
Up until now, advancements occurred primarily through martech start-up waves, but each new wave usually isn’t out-of-the-box compatible with what came before. Consequently, while the state of martech advances rapidly, marketers may wind up relatively frozen in time at the moment they commit to a vendor or an approach.
Acquisition sprees by the mega players – Oracle, Salesforce, Adobe, IBM, Google and Facebook – has begun to organize much martech capability into those vendors’ marketing clouds. That may eventually solve the problem of backward compatibility for new capabilities. But for now, notes Keating, “Even within the big boys’ clouds, the legacy of bolt-on acquisitions leaves the level of integration, when it does exist, woefully inadequate for the near future of Post-Modern Marketing.”
One approach for keeping up is through what Keating calls “app clouds” (in a slightly different metaphor from “cloud computing”). In his thinking, app clouds are collections of dozens – or more likely, hundreds – of small martech tools narrowly focused on individual functions, but “plug-in” compatible with a major vendor’s marketing cloud, in general, or with a particular marketing automation system such as Eloqua, Marketo or Pardot. In the app cloud scenario, marketers keep pace with the ever-increasing velocity of innovation by planning an architecture that can embrace such plug-and-play tools as they emerge. Even this approach is hard; but, “If it weren’t hard, there wouldn’t be winners and losers,” Keating points out.
Still, marketing clouds and app clouds aren’t the only major martech adaptations on the horizon. Several major technology challenges remain. A combination of martech vendor innovation and marketers’ savvy will be necessary to resolve these challenges before Post-Modern Marketing’s one-to-one potential can be fully realized.
Data collection and synchronization. There must be far better connectivity among different vendors’ offerings, as well as among the components of major vendors’ marketing clouds. We’re talking comprehensive integration among data management platforms (DMPs), customer relationship management (CRM), and marketing automation, as well as myriad app cloud tools. Customer Data Platforms (CDPs) have begun to emerge as a repository for all these, but have yet to even start climbing the hype-cycle curve.
Embedded AI. Artificial intelligence/machine learning must become embedded in all parts of the business, i.e., sales, support, service, (in addition to brand-and-demand marketing). This adds emphasis to the preceding challenge for big data systems that can feed, and connect, all this AI, because data is the not-so-secret key to winning with AI. Perhaps more importantly, though, this emphasizes the need for enterprise integration well beyond the marketing department – and the required organizational transformation we already discussed.
Orchestration. Major advances in marketing automation orchestration engines will be needed to support embedded AI and the ‘advanced campaign canvases’ it makes possible. Personalization/individualization tools capable of delivering one-to-one experiences/messages across all channels will be key.
Experience apps. Software tools that take content experiences to the next level are needed – and those that exist must get easier to use. Think webinar tools that create “presence” via a Second Life-like virtual experience, or “professorial” AI agents embedded in future “white papers” capable of answering questions and conversing deeply on the paper’s topic.
Experience layer. Brands will need to invest in experience tech as a whole new layer riding on the martech stack. So far, most martech stack capabilities have focused on finding and differentiating audience members, not delivering experiences. Elements of this emerging martech experience layer include augmented, virtual and mixed realities (AR/VR/MR), conversational voice interfaces, and AI virtual assistants.
If the complexity described herein, compounding the complexity you’re already trying to accommodate, is starting to make your head throb, fear not. Or, at least, fear a little less.
Simplification is top of mind for everyone. As MarTech Conferences’ Brinker says, “Most martech software is incredibly painful to use. Much of the innovation I see in the next big wave will be in interfaces and solutions that make it easier and more intuitive for marketers to work with all these tools at a conceptual and activation level.”
HSBC, the global banking and financial services organization, can’t wait. HSBC commissioned Stein IAS to develop a custom app overlaying its marketing automation system to simplify and standardize campaign initiation from around the globe. The app also is being designed to expose sophisticated marketing automation capabilities, such as multistage nurturing, to marketers who aren’t yet fully trained and conditioned to consider such capabilities.
Whether through the slower-moving tech vendors or expert agency partners like Stein IAS, HSBC and other marketers are on a mission to master Modern Marketing, in part, by simplifying it.
At the climax of “Star Trek: The Wrath of Khan,” Kirk knew exactly what to do when Spock said,
Technology is accelerating B2B toward a Post-Modern Marketing world of hyper dimensionality. The return to pre-modern-like deep emotional resonance has already emerged as one of the most essential post-modern dimensions. This Post-Modern Marketing world will stand on the shoulders of the Modern Marketing revolution, requiring mastery of modern martech techniques but increasingly dependent on AI orchestration.
Of paramount importance will be the need to articulate business and marketing objectives, strategies and tactics with far greater clarity, depth and precision than ever before. Greater precision is demanded because the human judgement of the “corporate brand police” will give way to programming embedded in AI-orchestrated systems, notes John Ellett.
“How do you codify what a brand looks like and sounds like when it’s a machine that’s delivering that experience?” he asks. “We [Ellett and Tom Stein, Chairman and Chief Client Officer of Stein IAS] came up with this concept that there will be a new job in marketing that’s going to be called the brand encoder, who is going to actually train AI systems on how to behave according to a brand’s purpose and values. I don’t know what that job looks like, but I know it’s needed.”
Most marketers aren’t aware of the many challenges the Post-Modern Marketing age raises, struggling as they are to fully understand and exploit Modern Marketing toolsets and skill sets. “You can’t be post-modern unless you have your act solidly together on your modern marketing tech stack, have made it optimally actionable, and mature enough to layer on AI, voice, augmented reality, and the stream of new capabilities that will continue to emerge,” says Marc Keating.
“The clear threat is, if you don’t catch up, you will only keep falling farther and farther behind in what amounts to an arms race. The pioneers in your competitive set may leave you for dead,” he concludes.