Adobe sensei artificial intelligence

The Future of Marketing In An AI-Powered World

Artificial intelligence (AI) already has had a huge impact on business, and marketing in particular. With predictive analytics, natural language processing, and computer vision, companies can better understand their customers.

But, according to Gerry Murray, a research director with IDC's Marketing and Sales Technology service, companies have yet to scratch the surface of what AI can do. 

In this exclusive interview with CMO by Adobe, Murray talks about the current state of AI, how it fits best within marketing, and what we can expect in the future. 

Power incredible experiences with AI.
Adobe Sensei Artificial intelligence technology
Power incredible experiences with AI.

Designing and delivering the perfect customer experience can be complicated. Adobe Sensei uses AI and machine learning to make it easier. Learn more.

  • How are companies using AI today in marketing?

    AI is proliferating in every major marketing cloud in nearly every marketing point solution, but most organizations are still experimenting with it.  There are many use cases for AI that span every role in marketing from advertising to social, web, content, analytics and more. How you apply AI to marketing should reflect how you do your marketing. If you do a lot of outbound email, there are AI engines to enhance list selection and segmentation, subject lines and dynamic content. If you do a lot of social, there are AI engines to analyze sentiment, identify influencers, and post content. If you do a lot of web marketing, there are chatbots and recommendation engines to assist with site navigation. And many more.

    The difficulty is most marketers do all of those things at scale and each requires a different kind of AI engine. This creates the potential for technology proliferation, so marketers should be careful to avoid silos in the AI layer of their infrastructure. The goal should be for every interaction to enhance every other interaction regardless of whether a bot does it. So the AI engines need to be connected.

    It's important to start down the learning curve with AI as there are new technologies, practices, governance and skills needed. The primary way most marketers will interact with AI is by feeding data or content into the engines. Marketers should think of AI as a new way to cook, and their expanding datasets as ingredients they can use to create exciting new experiences for themselves and their audiences. As with everything in the kitchen it can be a little daunting or even dangerous if you don't know what you're doing. You don't need to know exactly how the microwave works, but you do need to know that you shouldn't put metal in it and that some things like eggs tend to explode. Once you get the basics down its about how creatively you can use your data to not only do the same old marketing better, faster, cheaper, but also to create new value and experiences for your customers. 

  • How do you expect AI to evolve over the next 12 to 24 months?

    The two big trends for AI right now are conversational interfaces and proactive AI assistants. Conversational UIs enable natural language interactions with AI both via text and increasingly voice. We're already starting to expect voice-enabled interactions based on our experiences with Alexa, Siri, and other devices that are going into our homes. It's something marketers have to put deep thought into because AI will literally be the voice of your brand. The success of conversational interfaces is not just about what AI does. It's also about how it’s communicating with people and what it can do to embellish the relationship.

    Ultimately, AI assistants should evolve beyond task level automation and become proactive lifestyle companions that curate and expand our experiences. They should suggest a broader set of tasks they can do for you with approval. The basic design principle AI assistants should adopt is to boil down all the complexity into a simple question: "Would you like me to _______?" That is friendly, intuitive, and proactive. It simplifies the whole process for people and enables the AI to present capabilities and recommendations that users wouldn't think to ask of it.

    There's going to be a lot of competition based on which bots bond best with which customers. There will be a certain amount of emotive engineering and experimenting required to answer new questions like: How much do we support conversations that are non-transactional, maybe even non-commercial? Some of the most popular AI applications in the world have been ones where people are just chatting about their lives.

    The whole concept of branding is to create an emotional bond with the customer. If you put a boring AI bot out there it will fail against one that does the exact same things with more personality. People will gravitate to AI that is as useful but also entertaining and makes them feel better.

    However, marketers must keep in mind that AI-enabled interactions with customers are recorded and how they treat that data needs to be thought of as a core brand value, in addition to a compliance requirement.

  • So that’s the near term. What about 10 years from now? Do you see a use case where AI inspires new ways of operating or growing a business?

    Today's AI is rapidly improving its ability to automate analysis, tasks, and communications. It's really good at looking at vast data sets and identifying relationships between all kinds of behaviors that we just wouldn't pick up on. AI is not so good at innovation yet. It can't look at those same datasets and invent new things to do. Over the next 10 years, AI may be doing more and more of that creative work. It may also become a formal part of org charts. Most likely, people will manage, collaborate, and report to or be evaluated by AI—at all levels across all departments. That will require new skills, attitudes and culture in the workplace.

    It will also proliferate on both sides of the marketplace. Buy-side AI will handle increasing volumes of interactions and transactions. Bot-to-bot commerce will come with radically different approaches to content, data, pricing, branding and pretty much everything marketing does. 

  • I'm interested in hearing your thoughts around how AI is helping brands become more human—ironically, because it's technology! Any thoughts about how AI capabilities like emotional detection are helping brands better connect with people?

    Emotional capabilities will be key to making consumer AI assistants lifestyle companions instead of task-based worker bees. Emotive AI is key to getting the customer to bond to whatever customer-facing AI you put out there. Once it goes beyond simple content recommendations, how far afield do we take the conversation from book my flight, buy this product, and alert me to this discount?

    To the extent marketers can include more positive emotional support for the customer, the customer will feel better about the interaction. One of the things we commonly hear from people interacting with bots is that it's simpler. It's the difference between having a phone conversation versus just texting. With texting, you cut to the chase. You take a lot of the emotion out of the interaction, and it's more efficient.

    Now you want to figure out how to fill in some of that emotional bonding for the human part of the conversation. It's going to be tricky, and it's going to vary based on what kind of customer you're dealing with in what kind of interaction. Is this someone responding to an ad or clicking on your website? Or is this person complaining about a bill or scheduling a field service call? There are very different levels of information disclosure and opportunities for expanding the conversation beyond the transactional element. Regardless of how deep the capabilities are, it must intuitively adapt to the preferences of each customer. 

  • AR is often discussed as the future of customer interactions online, but it hasn’t truly broken through into our everyday digital experience. Will AI be the thing that makes AR break through to mainstream?

    There's a lot of urgency, particularly in retail businesses, to make a destination an experience. Going out shopping is not just trying on clothes to see if they fit. There's a lot of potential to add more to the engagement in stores due to augmented reality and potentially virtual reality as well. IDC expects strong adoption of AR. Particularly in retail, sports, entertainment and hospitality applications. Because AI is so good at personalizing content, it's a really great method for creating a rich AR environment. In a sporting context, it can potentially scale for each and every person in a 50,000-seat arena. Everybody's got a different favorite player. Everybody's got different parts of the game that they are tracking. Fantasy stats they want to track, etc. There are a lot of ways to incorporate AR into those real-world experiences.

  • Can you provide our readers a word of advice on how they can prepare for the future of AI in marketing?

    Everyone in marketing and in many other roles in organizations are going to have AI assistants. Key questions abound about: How much of which jobs will be delegated to AI? How do we retain and retrain the intellectual and experiential capital of our human workforce? How do we make AI plus human better than either on their own? How do we work for, with, and manage all these bots? How do we express our brand values in our customer-facing AI?

    As a marketer, you're going to start having to interface more with data sets and data science. Data scientists don't necessarily understand marketing and the marketer might not understand data science. But the two need to come together to understand the parameters of use cases, the data that is required to optimize them, and how that data can be acquired, governed, activated and enhanced, There are also a lot of issues with bias in datasets and marketers need to ramp up in all of these areas.

    In addition, AI is not set-it-and-forget-it. The world changes in unexpected ways and the AI may become less effective over time. You've got to continuously monitor it and understand that many of the assumptions used to deploy AI will change over time. Your audience will change. The products will change. The competition will change. Market regulations may change. So the AI is going to require constant retraining and updating. In other words, adult supervision is mandatory.

    In my opinion, the areas that will see the most innovation and result in the greatest competitive advantage are how brands manage their data relationships with customers; and how they use AI to give customers an equal voice in the marketplace. 

Check out the recent "IDC MarketScape: Worldwide AI in Enterprise Marketing Clouds 2019–2020 Vendor Assessment." (Note: You must be an IDC client to gain access to the report.)

Up Next

Trends in Asia Pacific

Emotional Analytics: From Personalisation To Individualisation

Insights from Adobe

Mobile Personalization: Top Trends For 2020

From the Blog

Dig Deeper