Artificial intelligence and machine learning present opportunities for today’s marketers from enhancing productivity to improving audience targeting and conversion rates to bettering customer service. Jim Sterne, founding president of the Digital Analytics Association and founder of the eMetrics Summit, brings 35 years of experience in sales and marketing to our discussion about the role of artificial intelligence and machine learning in B2B marketing.
What is artificial intelligence?
Artificial Intelligence is an umbrella term that covers a lot of different areas such as natural language processing — turning words into text and then text into meaning — and computer vision — telling the difference between a cat and dog or conducting facial recognition.
On the easy end in marketing, it is finding your logo or product on social media, while on the sophisticated end it is self-driving cars. If we put all of this on top of something that moves around, we call it a robot — something that can check us into hotel or help us find a product at the hardware store.
Machine learning is applicable to much of the above, and is where marketing professionals will spend most of their time. It is where we take advantage of AI. Artificial intelligence is different from previous technology in that it has the ability to learn. Before, you wrote a computer program and it executed what it was told. The next level is predictive analytics — statistical modeling and calculating probabilities to predict what will happen. Machine learning is different because it looks at data to build its own model and when it gets new data it changes its mind — it learns.
How does it apply to marketing?
Machine learning is really good at doing very narrowly defined, specific tasks, for example, which offer we should send to which person via email. Today, you might brainstorm that a group of mothers who like fashion and yoga should get 20 percent off yoga pants. You create the rules and automate them.
In addition, you might A/B test two emails. You send half of them a 5 percent off coupon and the other half a 10 percent off coupon and then assess the results.
With machine learning, you let it do the testing and, based on the results, it changes its mind. Over time, it learns which coupon to send to which segment of your audience. This can be applied to testing email or using a bot to reply to customer service questions or social media comments.
What many marketers don’t know is they are already using machine learning. For example, bidding on Google keywords or Facebook advertising.
What do marketers need to know?
What marketers need to know is that they don’t need to be a data scientist to do this. There are a bunch of startups building systems using machine learning on the backend.
For example, machine learning can automate attribution and optimization. Invoca integrates with Google AdWords and listens to sales calls. The system tracks which ad the caller clicked on and which landing page drove the call. It will then listen to the call and give recommendations to the sales person on what to say and what the prospect might ask next. If a sale is made, it will take data from the phone call (zip code, price, etc.) and send that information back to AdWords to adjust the ad bidding in near real time.
How is artificial intelligence changing the industry and our jobs?
AI is changing marketing incrementally — it is not the big scary thing everyone is worried about. But, if you don’t start testing and using it now, you will be left behind.
AI is a tool that improves productivity and provides better customer service. This means customer expectations will grow. For example, we used to press one or two when calling a customer service line. Now we expect to be able to say "one" or "two." Marketers have to keep up with customer expectations, and AI can help.
It will change jobs in that it will do more and it will do more tasks that only humans could do before. Today, my social media management system monitors, posts and schedules. It is just automation. AI will enable social media management systems to advise me on when I should post, automate when I post, and decide on when I should post based on results I get from the past two weeks. Tools will become more intelligent.
What can we do as marketers to remain relevant?
If you want to remain in front, the best advice is to be an enthusiast for this new technology. Be the one to bring it into your company. In the early 90s, you would have been the one to say you need a website and a company email address. Now people expect chat bots — they want to go to Facebook and ask a question and get an immediate answer.
Let your company know there are some tools you can use now and some you should start exploring and you should do it before your competitors do it. Host a lunch and learn. Run around your company and find people who are working on AI already and participate.
If you see there is a group that meets every week, go. Invite yourself, and explain that you are from the marketing department and learning as much as you can. Set aside half an hour to do basic research so when someone asks about random forest vs. a support vector machine you’ll at least have heard of these terms.
How does artificial intelligence affect B2B?
AI almost applies more to the B2B space. It is impossible to guess what one individual will do, but statistically speaking, it is easy to guess what a group will do. That is a big problem for B2C.
We have more info about individuals in B2B than groups. For example, we know job titles or can look up organizational charts to find relationships in a prospect’s company. The more data you have, the easier it is for machine learning to find patterns. You can provide everything you know about your best customers and the machine will find clusters and segments. You have companies that are high lifetime value that look like this and you have companies that buy your lowest priced products only once. You should go after companies that look like the former rather than the latter.
Ultimately, B2B has even more to gain.
November Keynote with Jim Sterne, Artificial Intelligence for Marketing: Getting Started
Join the BMA Colorado and the Digital Analytics Association at a joint Keynote presentation with Jim Sterne on Nov. 8 at 4:30 pm. Find more details and register here.