Why Automation Is Your Strategic Advantage in the Personalization Wars
With 47,000 rooms and 20 million guests staying with them annually, MGM Resorts — located on the world-renowned Las Vegas strip — had to make a shift from pushing products on guests to offering them the personalized service they wanted. Regardless of the method, they needed their customers to constantly be doing something — downloading an app, consuming a product, clicking on this push notification — for the brand to reach their engagement and conversion goals. This is undoubtedly a daunting task to accomplish in a landscape where millions of customers already had their ideas of the perfect vacation in mind — and often planned — before ever stepping foot on the strip.
So, using automated personalization, MGM Resorts wanted to put the control back into their customers’ hands by asking — not telling — customers what they wanted to do. But, with 20 million guests annually, how is this type of one-on-one personalized experience even possible? In 2013, the resort looked to Adobe to find their solution. With the right analytics and personalization tools to automate and, therefore, scale their efforts, they were able to not only ask guests what they wanted to do, but also then offer them the resources to follow through via rich imagery and up-to-date information — and all at their fingertips, whether through their apps or their website. As a result, engagement and conversions skyrocketed and so did MGM Resort’s ROI.
Suffice to say, the marketing umbrella has widened quite a bit since digital burst on to the scene. From optimization, testing, email, and mobile to social and cross-platform extensions like beacons and geofencing, the opportunities are endless. But, despite the incredibly diverse landscape, at the end of the day, the goal is the same. Like MGM Resorts, you want customers to do something — to take action, to convert, to opt in, to hit some type of benchmark — and you want them do whatever that is with you and not with your competitors. Pretty straightforward.
And, most recently, that ‘doing something’ is being fueled by the one-two punch of data science and machine learning. By tapping into artificial intelligence and machine learning to create spot-on relevant experiences and automation to scale, brands can curate unparalleled customer journeys that far exceed what human marketers can deliver.
The Next Stage of Customer-Driven Relevance
None of this is new either; but now, a higher level of access, intel, and integration is enabling companies to gain more, do more, and deliver more for every consumer every time. Those old-school recommendation engines definitely did the job and were better — hands down — than no personalization at all. But, these tools are growing smarter and more advanced, and today, companies are pulling out all the stops, incorporating offline data, and improving algorithms themselves to create constantly personalized experiences at scale.
One national home-improvement chain, for example, is doing some very interesting things with their recommendation engine, incorporating all sorts of offline data. They have two different data flows feeding their engine, adding new dimensionality to the entire experience — specifically, the data consumed by their recommendation engine creates more granular relevance. The organization creates more things for the engine to work with so it can make better decisions thanks to finer, sharper algorithms created and refined specifically for this brand. That level of granularity allows the company to deliver much more spot-on relevance and to reap increased conversions and other key performance indicators (KPIs) as a result.
Here’s another prime example. We work with a major US-based bank that’s really doing it right. The bank’s website has numerous slots on their homepage, as well as all across their site, that show different permutations and user-specific content based on each visitor’s product consumption and behavior. A human marketer could not possibly sit down and create all the potential content combinations that exist based on click throughs, account status, credit card usage, and other relevant details. But, by allowing machine learning to take the reins, the bank’s robust automation engine can cook up ideal combinations of creative, content, and onsite ads. The result is a highly customized experience that draws customers and prospects further down the path toward engagement, activation, and long-term loyalty.
What It Means for Brands, Marketers, and Everyone in Between
Done right, the impact of processes and workflows driven by machine learning like this can be dramatic since the customer is much more likely to experience relevance. Done right, experiences are smoother, increasingly meaningful, and in line with each customer and what he or she is trying to accomplish. But, despite delivering all this relevance, the best machine-led personalization is totally seamless and behind the scenes. To the customer, it just happens. It’s the marketing equivalent of, “If a tree falls in an empty forest, …” — you’re delivering spot-on relevance in a completely frictionless way. It’s a massive improvement but one at which no one will ever bat an eye.
In my experience, this is something that — oddly — agitates marketers. They don’t want to go the automation route, because they don’t want to give up control — and I understand that. No one wants to feel obsolete or as if they’re being left behind, and companies that aren’t culturally prepared for this kind of seismic shift can be tough sells. However, adopters are completely convinced that machine learning works magic. They’ve seen the proof and are now believers.
Until you have something to compare it all to, it’s difficult to really wrap your mind around what you’re missing. Many brands we’ve worked with start realizing what the curated or random experiences were generating — compared to the bigger, better, algorithmically generated ones — but not until they took leaps of faith and began leaning on the machine and benchmarking every step.
The Future of Marketers (Admit It: You Want to Know)
None of this means that marketers are obsolete — far from it, actually. With an automated system in place, marketers do lose some control but more than make up for it in other ways. Sometimes, they literally intervene — the marketing calendar says this needs to happen this month; thus, regardless of what the algorithm says, the marketer is going to intervene. And that’s fine so long as they don’t go overboard. But, if they start applying too many rules, the machine can’t do its job effectively. They’ll, essentially, tie its virtual hands behind its back and limit its ability to deliver those spot-on experiences — and that serves no one. Sure, the marketer maintained control, but at what cost?
The alternative is to give marketers control at the other end of it all, when outputs are apparent and results are ready to be viewed. At this stage, marketers can really understand what’s happened, so it’s not just a black box. Here, marketers can look at the full picture and determine the dominant variable — maybe it was this audience that’s made up of people who are this age, live in this market, and have these behavioral traits. Maybe, together, that’s what moves the needle and makes this segment a top performer.
Armed with data-driven intel, the marketer can call the shots — maybe do more of this, test something new based on the results, or up the ante for this vital segment. By digging in, making decisions, and informing future activities, marketers wind up with some very high-value, high-level control. They can shape the decisions made and inform future activities, and that’s the kind of control that really matters. Through that lens, it’s clear that marketers aren’t giving anything up; instead, they’re gaining greater clarity and potential. Risks drop, returns soar, and everyone’s happy. What could be better?
Where Machine Learning Is Headed
Hands down, I think the theme of machine learning, automated personalization, and data science is more, more, more. We’ll see more adoption, which in many ways, I credit to the cross-pollination of the academic world and the tech and innovation industry. People are leaving global universities with well-articulated senses of machine learning and data science as well as a variety of smart, strategic industry applications and use cases right out of the gate.
Further, a variety of shifts and enhancements accompany that. For starters, we need to have tools in place that make machine learning, data science, and automated personalization incredibly user-friendly and more democratized than ever before. Some companies will be unable to carve out the dedicated talent and resources needed to make all these now-essentials a reality. By simplifying and streamlining everything and layering in sophisticated, programmatic systems, companies can share the wealth.
As that happens, companies in regulated industries — healthcare and financial services, for instance — will have to figure out everything from appropriate recordkeeping to archiving and privacy. It sounds simple, sure. But realistically, it can be quite challenging, especially when you’re talking about millions of unique experiences — millions of experiences that must be combined to create spot-on relevant touchpoints for countless consumers. We’ll figure it out, but until then, it’s a lot to digest.
Finally, we must come together to resolve this whole cross-device, cross-interaction, Internet of Things (IoT) world in which we’re currently living. To be successful, companies need to look at the probabilities and propensities of consumers to do this or that based on previous interactions, expressed wants and needs, and third-party data. Right now, organizations can take advantage of a variety of data inputs — the volume and complexity of data as well as its sources — and create really interesting opportunities on the machine-learning front. And, while that will benefit both businesses and their customers, it will also make the need for machine learning that much greater.
There’s so much more on the horizon — and I, for one, can’t wait! Adobe Marketing Cloud and Adobe Target are both positioned to facilitate and advance this next wave of marketing, and at its core, encourage consumers to “do something” — something bigger, something of higher value, and above all, something with OUR companies. Because, at the end of the day, that’s the goal no matter your industry, your niche, and your target audience.