In the marketing world Artificial Intelligence (AI) is a popular “buzz word”. If you haven’t spent much time researching AI, consider this blog your introduction to a topic you’ll be hearing more about in the future. If you are like us, you don’t want to be left behind when it comes to new technology and innovation. Thankfully – for those of you in and near the Twin Cities – you can join us at DenamiCON and explore this topic in more depth!
As a pre-conference teaser, we’re pleased to bring you this “sneak peek” of insights on Artificial Intelligence and marketing in the machine age.
Welcome to this edition of Authentic’s “virtual panel” series: a Q&A exchange with four national marketing and AI leaders.
This dynamic group of thought leaders represent the panel for the upcoming DenamiCON conference in Minneapolis on Thursday May 17th. The four-hour learning and networking event will focus on how we can become more artificially intelligent, and what AI will mean for our businesses and customer relationships in the future.
Join me in scrolling down the screen for a virtual introduction to our AI panel of experts (subsequently referred to by first name in the conversation that follows):
- Paul Roetzer, PR 20/20
- Scott Litman, Equals 3
- Kristen Findley, Ciceron
- Josh Cutler, Aftercode
Q1: There is a lot of hype around Artificial Intelligence, but a general lack of clarity about what AI is and what it means for the average business. How do you define AI in your conversations with customers?
(SCOTT) When we think about AI, we think of it as “Augmented Intelligence”, the ability for machines to work with incredibly large and complex data sets and enhance the ability of a human companion to do dramatically more in a fraction of the time. We try to reflect that in our company name, Equals 3, which is all about the idea that better then the individual, better then the machine are the two together. That we are creating a 1 + 1 = 3 for our customers.
(KRISTEN) The term AI covers a wide range of tools and technologies. Many of us are familiar with Siri and Alexa: these are forms of AI. IBM’s Watson is AI. Sometimes AI is applied to things like basic marketing automation and then it can also apply to self-driving cars. For business and marketing, I focus on the functionality of AI: using it to manage large amounts of data or using that data to create better experiences.
(JOSH) AI encapsulates a very big bucket of techniques that we use to create software. In general AI is any set of algorithms that allows a computer to make a decision. Your thermostat is an AI, in as much as it decides to heat your house when it gets too cold. Most people are conflating AI with machine learning right now. Machine learning is a type of AI that uses data to make predictions by trying to mathematically define the relationship between whatever data you have to the outcomes that you care about. In the end its all about using some set of inputs (numbers, voice, imagery, etc.) to make predictions about the world (end of quarter revenue, the user said X, there is a pedestrian in front of my car, etc.).
(PAUL) Ask 10 different experts to define AI, and you’ll likely get 10 different definitions. My favorite, in part because of its simplicity, is from Demis Hassibis, Co-Founder and CEO of Google DeepMind. Hassibis defines AI as, “the science of making machines smart.” These machines, in turn, enhance human knowledge and capabilities. Basically, artificial intelligence is the umbrella term for the algorithms, technologies and techniques that make machines smarter, and give them superhuman capabilities.
From a marketing perspective, it’s easiest to understand by looking at specific challenges and use cases of marketing automation. Humans are unable to conceive of the optimal set of instructions to guide the machine on how to personalize experiences at this scale. This is where artificial intelligence excels. It takes data-driven, complex tasks and makes them look easy. But, artificial intelligence doesn’t stop at setting up the initial rules to maximize performance, it uses machine learning to constantly evolve its actions. In other words, it learns, it gets smarter, and it creates its own algorithms. Now, imagine the potential if all the time-intensive tasks you complete and the data-driven decisions you make every day as a marketer were intelligently automated.
Q2: With any new technology or innovation, there is always the risk of buying into the big promise and vision, without clarity on how ROI will be measured or proven. How do you suggest that organizations evaluate AI and begin to build a business case? In other words, what key questions should business leaders ask and answer before moving ahead with AI?
(KRISTEN) It’s not about using AI for the sake of AI. As with most new tools and technologies, the first step is to determine what business problem you are trying to solve or what challenge you are trying to address. If you have mass amounts of data – in all different formats and locations – there is probably a good AI solution to help with that. If you are looking to automate systems or make these systems move faster, AI may be able to help. You most likely won’t get a lot of return on an effort that is dedicated to AI just because it’s AI. Figure out the business need first.
(SCOTT) Early on, we were so enamored with the technology we created that we thought it was just obvious that anyone and everyone would want this. But coming back to reality, no matter how interesting or cool a technology is, it’s only worth paying for it if it can solve real world problems. And more than that, the problems solved need to be worth far more then the cost of implementation and acquisition of the technology. So the question to ask yourself is: Do I have data problems that need solving? Do I have processes that are inefficient that can be automated? And finally: Is the pain great enough to require investment to resolve?
(PAUL) What do you struggle with daily? What strategic priorities are important for the next couple quarters? What are your performance goals (New leads? Visits? Higher sales?). What activities take you away from the work you should be doing? Make a list of these. Then begin your research by searching for these terms or problems and including the term “AI.” Start experimenting with available tools and talk to vendors about what’s possible. Most are ready and willing to talk about your individual use cases. Don’t be afraid to look at your existing tech, either. Lots of platforms are incorporating AI, and a tool you already use may have it or have it on the roadmap. From there, you’ll be able to better gauge how AI can solve your organization’s challenges.
(JOSH) When talking to businesses about how they should think about AI, I always start with one thing: Have you already identified some quantity/concept/thing which – if you could predict it with better accuracy – would improve your business? If so, then great. Let’s start figuring out whether AI can help. If not, then focus on your business. Figure out what actually moves the needle for you, where you have uncertainty, and start there. I really don’t talk a lot about AI initially because if you aren’t prepared to take action on the output of an algorithm’s decisions, then you won’t get any value from it. I am very skeptical of pitches where AI is going to help people solve “problems they didn’t know they had.” Almost all of the successful use cases I’ve seen solve very specific (narrow) problems where the business knows how to take action on the output.
Q3: What kinds of businesses are showing up as early adopters of AI, and what are some of the primary challenges they are addressing through these solutions?
(JOSH) There are really three main areas where I see this happening. The first is using AI to create or improve experiences for consumers via new interfaces. For example, Alexa with voice or the Kinect for motion. Second, there are businesses that have relied on prediction in the past and are just doing it better with AI (like banks, natural resource discovery, etc.). Finally, we see new types of businesses being built that harness previously unusable data sources. This third group represents a lot of what we are doing with Rambl and voice.
(PAUL) Enterprises are definitely poised to be early adopters of AI, since they have the budgets and the volumes of data required to make some AI solutions work. For instance, an AI-powered tool called Phrasee can write email subject lines automatically, but it needs 100,000+ emails to be able to write these subject lines. Firms with a lot of the right data are probably going to get the most value from AI first.
(SCOTT) Our view is skewed toward the types of clients we work with, which have been Fortune 1000 marketers and the agencies that serve them. That market has made sense for us, as they have a lot of data and plenty of data challenges that go with it. Within that framework of Fortune 1000 and large agencies, we are seeing the pursuit of AI across almost all verticals, at least by those companies that are interested in innovation and who are not technology laggards.
(KRISTEN) One of my favorite early versions of AI is the experience on North Face. They are using Watson and data from over 55,000 users of their product to help customers find what is right for them. It’s a simple interface with a lot of power behind it. There are so many possibilities in retail and for any company looking to improve customer experience.
Q4: How are today’s AI solutions already enhancing the end customer experience? On the flip side, how might AI disrupt or negatively affect the customer experience? (And what advice would you offer to business leaders to protect against these risks?)
(PAUL) AI provides marketers with the ability to plan, produce, personalize, promote, and perform better than ever before. This is good for marketers. AI can be even better for customers. Marketing can better deliver the products, services and experiences that consumers crave in a very personal way. In terms of potential negative ramifications to the customer experience, AI in marketing is at a preschool level right now. It’s not always going to work how it’s intended and may be a little bumpy for consumers along the way. Consider Apple’s Siri, as an example. Being able to ask your phone anything at any time without having to type is pretty awesome, except most people would probably agree that Siri’s capabilities and accuracy are still a bit limited.
(JOSH) This is happening in three main ways for consumers. You have new user interfaces like Alexa, you have automation like Tesla’s self driving cars and you have improved customization (e.g. Netflix or Facebook). In all cases I think the trap is similar: Don’t pretend that AI can do things that it isn’t capable of, because it can lead consumers to draw incorrect conclusions or make wrong decisions. You can see this in the failure of early self-driving cars or people relying on Facebook as their primary news source.
(SCOTT) Since AI is best for solving big data problems, from our perspective this falls into a few major areas for marketers. Research and knowledge management: Can I efficiently find and use all of the the data that I own and license? Audience insights: Can I use data to gain a better understanding of my most important audiences? Media: How do I better understand past media performance to optimize my future allocations? As for impact, new technology typically brings about a level of organizational change. For example, when a process that used to take one full day now takes five minutes, you find you have more people resources available. There is also a change that takes place when data that used to take a day to get is now immediately available.
(KRISTEN) I’m a big fan of Alexa – she does everything from finding what I want on my streaming service to ordering my dog food. It’s important to note that I have given her my data to make that happen because I like the ease of use. I am trading data about me and my habits for that convenience. That is, of course, the big issue. Companies have a responsibility to be careful with this data to keep my trust. I think it’s important to keep the customer in focus: her needs, expectations, and respect for that relationship. Use this to help define experience, set standards for data, and innovate.
Q5: Along with the hype around AI, there is also an undercurrent of uncertainty and fear: How will AI impact human jobs? Will the “robots” replace us? Will we need to develop all-new skills in order to remain relevant? How would you respond to these questions?
(JOSH) From my perspective, this fear is very, very overblown. People will have to develop new skills — because they always do. Email, mobile phones and the internet all caused people to develop new skills. In the near-term, AI will augment humans and hopefully make our jobs easier. I am skeptical that most white collar jobs will be at risk in the near future.
(KRISTEN) AI is going to change the way we work, but there will still be plenty of work to do. It’s important to remember that AI has to be taught. When I ask Alexa to order my bag of dog food, she can do that because she has learned how to look into my previous orders and identify the dog food. That is a series of checkpoints that have been “trained” into the system. The same idea applies to the use of AI to manage large amounts of data. Any system has to be taught/trained how to manage the data – what to look for when answering questions. This process requires someone who understands the data and what the overall objectives are for that data. Most importantly, something still has to happen with the data: Actions must be taken. Marketers will be able to spend less time tracking down data and more time using it.
(SCOTT) I believe the fears are both warranted and unwarranted. Unwarranted in that AI today is nowhere near having the ability to simply do the job done by an individual. We aren’t going to lose Marketing Managers, Creative Directors, Account Managers etc., to AI systems that can somehow magically do those jobs. The state of AI is far, far, far away from that. But, automation brought about by technology is always disruptive. Roles will change, expectations about individual performance or client service will change. Our view is generally one that is positive. Those who embrace AI earlier will receive similar benefits to what we’ve seen with other waves of marketing technology. Those early to embrace the web, eCommerce, Content Management, Marketing Automation, Programmatic Media, etc. In each of these examples, there was a phase of early adopters who gained competitive advantage, and also laggards who waited to see if the latest innovation with real, and then fell behind.
Q6: Imagine our workplaces five years from now. What has changed because of AI?
(SCOTT) Increasingly, individuals will spend less time on the minutia and drudgery of working with data. Their AI assistants will give them command of all of the data of the organization and with that, no longer will data be locked up in silos of systems or exclusively with the experts (i.e. everyone will have access to media spend data, web analytics, performance of campaigns, latest industry research and more). Demands for performance will go up, and marketers’ AI companions will allow them to keep up with it.
(KRISTEN) We will probably have a different way of doing things. AI will handle some of the busy work we all manage. Research that now takes a few days will only take a few hours. We will still need talented people to apply that knowledge and make it work to achieve business goals. In fact, with more data available more quickly, we will need even more focus on what is going to drive results. We will also have to stay focused on the customer experience. For example, I think there are some great possibilities for retail to apply AI. If Alexa can keep track of what I’ve ordered before – why can’t my favorite store at the mall know what I may have just purchased online and use that info to make my in-store experience better? Consumer expectations are going to keep moving higher as more AI functionality becomes part of our daily experiences. We’ll all have to work hard to meet those needs.
(PAUL) In five years, a lot of what marketers do today will be done by machines. But that doesn’t mean the marketers who do those jobs will necessarily become obsolete. AI excels at doing certain very narrowly defined tasks really well and at scale, all automatically. Marketers should take an inventory of the tasks they do. With each one, asking: “Is this something a machine could do better and quicker?” If so, do some research, find a similar use case, and try using AI for that task. If those tasks constitute a large part of your job, you may need to consider how to evolve in the coming age of AI. For example, if you A/B test landing pages for a living, you’ll want to start looking for an alternate career path. But if these tasks are all taking you away from the stuff that really needs to get done, AI could be a serious competitive advantage in your career. In either case, marketers that take the time to understand and embrace AI now will have a huge competitive advantage in the years to come.
(JOSH) I honestly think that for the vast majority of people reading this, very few will see substantial change. We may be able to target ads a little better. We may have better diagnosis for cancer. But we will still be showing up and doing the work that we do today in mostly the same way.
If you enjoyed this article, you may also enjoy these other Authentic Virtual Panel topics:
Everything marketing automation: A virtual panel of expert perspectives
