Consumers are surrounded by artificial intelligence, from the iPhone’s Siri to voice-first user-interface devices such as Amazon’s Echo to the emergence of autonomous vehicles through Google’s Waymo. The same is true for brands, which are bombarded by marketing technology platforms that tout how their latest artificial intelligence capabilities will give brands an edge in the marketplace and, above all, solve their marketing challenges.
Attend any marketing conference, and you’ll see the terminology everywhere. Phrases like ‘AI-powered,’ ‘deep learning,’ and ‘machine learning’ are incorporated into product descriptions with nary a marketing technology solution positioning itself outside this trend. Often, though, this language brings a lot of flare without a lot of substance, making this constantly evolving landscape of new technologies, tools, and cognitive-marketing tactics difficult for brands to parse through. Which platforms deliver real value versus which simply attempt to stay relevant in a competitive marketplace?
Brands understand the need to embrace technology to compete. The threat posed by the likes of Amazon and its growing reach, enabled by constant technology innovation, is top of mind across leading companies. And the growing importance of incorporating innovation and technology into corporate road maps is evidenced by the free-flowing use of tech terminology in corporate earnings calls. In order for brands to make use of these technologies in a meaningful way, though, they need to cut through the hype and work the core principles of technology into their overall strategic marketing and product development initiatives.
Why the current emphasis on cognitive technologies?
Because the initial hype cycle is part of nearly every cognitive marketing technology, companies too often rush into the market in an era when innovation is a critical aspect of corporate strategy. Inevitably, though, some of this hype is short-lived. A decade ago, for example, everyone was touting Second Life as the beginning of a shift toward virtual communities. While virtual communities certainly changed some aspects of users’ worlds, though, they didn’t live up to their transformative hype for brands and marketers.
Still, cognitive technology is not a new field of study. It’s been an academic field of study for more than 50 years, but its mainstream marketplace use has recently ticked upward. Of particular note, the availability and exponential increase in computing power coupled with decreasing costs have driven real cost-benefits to intensive AI applications. This shift has, in turn, led to the availability of SaaS services offering streamlined implementation across technology stacks with consumption-based costs, rather than more traditional infrastructures that required significant hardware capital expenditure in order to utilize cognitive technology.
Consequently, mainstream marketing by the likes of IBM’s Watson has driven this terminology into the lexicon of both the consumer and corporate world. The ongoing trend to drive efficiency and focus on customer experience gives rise to the application of cognitive technology to meet these objectives. Combine that trend with the exponential growth of data available for analysis, and it’s no surprise that the number of companies investing in AI, as well as the number of startups or incubators fostering it, are growing at the fastest rate among new technologies.
What can cognitive marketing technologies offer?
Nearly every marketing brief now includes the objective of driving efficiency, reducing manual touchpoints, and connecting consumers across the journey in an integrated omnichannel ecosystem. When implemented correctly, cognitive marketing technology delivers against these objectives by improving effectiveness, decision-making, and audience identification. The common theme in achieving these objectives lies in driving efficiency — how marketing budgets can work harder by doing more for less — through automation. With proper marketing technology stacks and training systems for machine learning and automated optimization, brands can reduce manual processing and intervention across channels and tactics, from acquisition to customer services and retention.
AI is not a panacea, though. Often marketed as an add-on to a brand’s existing tools or as an out-of-the-box, natively intelligent solution, its power ultimately lies in the investment and rigor placed on training AI models with robust data sets against desired outcomes. These solutions require well-defined use cases that solve organizational and marketing challenges, which is why plugging in an AI solution without a robust learning model, a training and optimization plan, and clear objectives and success metrics will not result in meaningful outcomes.
How should you approach marketing technologies?
Placing too much confidence in new technology is all too common, but it’s important for brands to keep feasibility and limits in check by treating initiatives as learning and optimization exercises toward a larger goal.
Consider automated language translation services, for instance. Despite thousands of years of written prose available for analysis by machine-learning models, seamlessly translating static text from one language to another is still a challenge and is only recently seeing gains from neural networks. In many cases, a platform’s specific AI functionality is nothing more than a business rules management system, which has been part of software platforms for decades. Thus, organizations that want to incorporate these tools into their marketing platforms more comprehensively need to understand the role they’ll play, carefully vet their capabilities, create a blueprint for implementation, and thoroughly test the output.
In order to do this, brands must first identify the marketing problems they need to solve: What insights are they looking to derive? How can they leverage these insights to programmatically address marketing challenges? What manual processes and friction points would be replaced? What end value does this bring to the customer?
Brands can then more thoroughly evaluate platform capabilities by asking specific questions, such as how the AI features actually function and how they were developed, and by requesting case studies and demonstrations that reveal value matched to their marketing goals. Strong products and platforms aren’t easily dismantled when taking a deeper dive into the underlying algorithms, data analysis capabilities, and integration tools that are available with them. Moreover, they are backed by proof points derived from analysis of extensive data sets, adapting learnings and automating optimizations, which clearly demonstrate where they can render an automation fit against a brand objective.
Why is a good fit more about your marketing plan than the tools?
When first incorporating cognitive marketing technologies, brands can become overwhelmed by data and algorithmic approaches if they do not have a clearly articulated outcome. For this reason, they should note where these technologies could actually add true business value, focusing on how the platform’s features solve an identified marketing problem, such as reducing the time and expense to adapt creative across regions, improving the personalization and relevance of marketing messages to individuals, or reducing friction points to improve the customer experience while reducing cost. This forethought helps organizations understand the technology’s limitations and how to measure and refine the platform once it’s in place.
Despite the best laid plans, though, brands cannot always know the extent of these limitations or the pitfalls of AI automation before actual implementation. Netflix, Facebook, and Twitter, for example, have all had experimental AI applications that went awry or ended up costing more than expected. Thus, brands that set failure limits or investment amounts before implementation can better gauge whether a certain application merits the time and money put into it.
Ultimately, by looking for opportunities to bolster their current marketing campaigns and reach customers in a highly personalized way, brands can utilize cognitive marketing technologies to deliver more effective, efficient, and targeted campaigns while avoiding the typical mistakes that come with early adoption of new technology capabilities. Considering the more practical aspects of cognitive marketing technologies — with an ever-present focus on delivering end-consumer value through the pragmatic analysis of connected data sources — allows brands to skirt wasteful investments and unfinished automation projects.
Cognitive marketing technologies offer extraordinary benefits to marketers and consumers alike, and organizations that set clear objectives and success criteria, mapped to an implementation and test plan tempered by realistic expectations, will find that tomorrow’s technology may be exactly what they’re looking for today.