When the concept of digital transformation in marketing moved beyond early adoption and hit critical mass earlier this decade, marketers shared a vision of seamless digital technologies that would replace their cumbersome manual processes.
Their thinking was simple: “What I’m handling manually right now, technology will digitise for me.”
So they reviewed their channels and tactics from top to bottom to determine where technologies could create efficiencies, and what it would look like operationally and culturally to get their teams to adopt these new systems. Because hindsight is 20/20, we now know that they ultimately swapped their cumbersome manual processes for a series of cumbersome digital systems.
Marketers didn’t know it then, but we’ve since learned that digital marketing technology should not be organised into channel silos the way we’ve traditionally organised teams. We now know that data collected from one channel needs to inform efforts in every other channel and that technologies that were introduced as channel-specific tools now need to work across entire organisations — something even the marketing clouds have trouble with.
Because of the way marketing technology has evolved, marketers are left managing very complicated tech stacks comprised of multiple technologies, stitched together to complete what should be seamless and interconnected marketing processes. It’s no wonder that even though companies have more technology at their disposal than at any other point in history, only 39% of executives today say they feel they have the digital capabilities they need to compete.
The team at Albert have spent the last decade reimagining how to process, analyse and act on audience, channel and tactic data at scale. We believe the introduction of artificial intelligence into this game will be the final tipping point for marketing’s digital transformation — despite the many challenges that remain. Here’s how.
1. Digitisation will become intelligent
One fatal obstacle along the digital transformation journey – to uncomplicate interconnected business processes like marketing – is the idea that simply digitising systems or going electronic would be transformational. The goal posts have since shifted to reflect reality: Transformation doesn’t result from merely digitising manual tasks; it comes from automating entire processes and leaving humans to guide strategy rather than execution.
This requires handing over data processing, analysis and pattern discovery to intelligent machines that can autonomously and instantly act on insights. At the same time, organisations must recognise that these remarkable digital systems can only go so far without talented human guidance.
2. Humans will apply more of their intelligence
The introduction of AI to digital transformation initiatives will result in a dynamic where employees no longer “use” technology, but collaborate with it. Successful AI transformation will be characterised by a symbiosis between man and machine, where each does what they do best and uses their individual strengths to heighten overall levels of performance. Humans, for example, are needed to guide the AI on matters of strategy, brand and customer experience. By doing this, these systems function as true collaborative partners – amplifying their capabilities beyond the limits of a single human or simple automation. When these hybrid human/machine teams learn to interact and experiment, they create entirely new possibilities and outcomes for companies.
3. Technology will become cross-linked into full processes
Digital transformation efforts have resulted in massive amounts of valuable data, and there are now technologies smart enough to take this data, learn from it and orchestrate campaigns across channels using existing technology stacks. I believe we’ll see many companies fully realise digital transformation by doing exactly this.
AI will enable marketers to bring together disparate technologies using the ability to ingest and process massive amounts of data, and find patterns in the noise that yield unexpected insights and results at lightning speed.
The challenges ahead
Each of these three steps is, of course, a massive undertaking – and AI transformation certainly won’t happen overnight.
There are a few common challenges we’re seeing that will inevitably delay companies, across all industries and use cases, from realising full AI transformation.
One such challenge is the seemingly innate human need to tightly control AI systems while getting comfortable with them. Giving only half of the control to an AI, however, will only yield half of the learnings and results. Any AI system requires a certain amount of testing and data accumulation in order to learn and ultimately perform. Though this period of learning can be relatively short, humans tend to become impatient and step in to manipulate the process, rather than letting the machine run. Other times, this is less a matter of lack of patience and more a response to seeing the machine tackle problems in a different way than a human would. That tends to make humans uncomfortable and seek ways to control its process.
AI developers must bear some of the responsibility for both of these responses by guiding users through the upfront adoption process. We learned fairly on, for instance, that we cannot be passive technologists — we must be educators, too. Companies’ reaction to this upfront learning curve will set the tone for the rest of their experience, so they must be coached to let the AI go through its process without interruption, no matter how tempting it is to step in and guide it.
On the complete opposite side of the spectrum are companies that give the AI too much control without giving it a strategy. When companies view AI as a magic bullet, they often make the mistake of sitting back and not setting parameters that will guide it toward their desired outcomes. Make no mistake: artificial intelligence is dependent on — and better because of — human intelligence.
Both of these scenarios point to issues that arise when companies don’t recognise that there’s a clear division of labor between man and machine. The first example illustrates a situation where humans are too involved, and the second illustrates a situation where humans aren’t involved at all.
Mastering the fine line of giving up control on everyday execution and retaining control of strategy will be critical to organisation-wide AI transformation.
Article previously featured in Forbes, Why AI Transformation Is Digital Transformation, Fully Realized, February 11, 2019