By Dan Gingiss, Vice President of Strategic Group
Just when you thought you had finished the circuit of thousands of marketing technology companies asking for your business, along comes a whole new sector – Artificial Intelligence and Machine Learning. What exactly do these companies do, and how can they help you as a marketer?
These are questions I needed to learn when I was on the receiving end of these emails at several companies. In the most basic sense, Artificial Intelligence (AI) is simply leveraging a computer’s brain power instead of a human’s. And, no surprise, computers are generally faster, can retain a whole lot more information, and make fewer mistakes than their human counterparts.
A great metaphor that Persado shares with companies new to our technology involves the game of Scrabble. A human given seven letter tiles, will generally come up with an average word worth an average number of points. The machine, on the other hand, can instantly determine the total number of possibilities and which word is worth the most points.
Machine Learning refers to how these systems get “smarter” over time – the more data they can crunch and interpret, the better they perform because they can use history as a guide. This is why IBM’s Watson was able to easily beat its human competitors in Jeopardy! In the Scrabble example, machine learning allows the system to recognize that it’s seen the same combination of seven letters before and answer even more quickly – try that, humans!
This isn’t to say humans should be afraid of the computer. Quite the opposite, actually. For example, in the marketing sector, machine learning doesn’t just achieve powerful, consistent results, it performs day-to-day tasks like coming up with winning subject lines, giving marketers more time to focus on bigger-ticket items like long-term brand strategy. So as this new breed of companies approaches you with promises of game-changing results, how can you identify the truly “smart” technology? Consider this your checklist:
- Check to see if the results claims are based on actual data. Companies that are really leveraging AI and Machine Learning will have so much data and analytics that it will be hard for them to distill it into a simple presentation. That said, if they can’t explain simply what it is they do with the data, then walk away. Statistical significance is also key, so bring your Analytics team along to verify the claims.
- Make sure the company can solve your specific challenges. So many software companies today claim to solve every challenge in every channel. If you are looking for an email solution, ensure that the vendor has had measurable success in that specific channel.
- Ask if the company has been recognized by any third parties, which lends instant credibility. For example, the Retail Industry Leaders Association named Persado a winner of the (R)Tech Retail CEO Innovation Award for “groundbreaking work in promoting the industry's move towards shopping wherever, whenever through artificial intelligence and augmented reality.”
- Request actual client references, not just logos on a page. Every company shows lots of impressive logos – sometimes it feels like they are all working with the same clients! – but when I would call a colleague at one of those companies, I often found that they had no idea who this vendor was. Ask for a real person to contact.
And two to look for in any company (but especially AI)
- Make sure they’ve done their homework on your company instead of presenting you with a generic sales pitch. On the very first contact, they should know the name of your company (one vendor actually misspelled “Discover” on the cover slide of a presentation when I worked there), the type of business you’re in (another vendor confused Discover with the Discovery Channel) and something about the company’s goals or positioning that is available from a simple Google search. This may sound like a basic best practice for any salesperson — and it is — but it’s especially true in marketing and AI. If technology is going to make suggestions that will make or break an email campaign, such as the best email subject line, it’s important that the people selling the product are equally detail-oriented. Though some Discover employees were Discovery Channel enthusiasts, no one wanted a tech company mixing up our credit card deals with the latest news from the animal kingdom.
- Similarly, make sure they know something about you. Don’t you hate it when companies send the same sales email to 10 different people in the company? If they’ve taken the time to learn something about you, your background, or your interests, you will know immediately that they are thorough and personable – something that you’ll also want them to be when they work with you (and create content that speaks to your audience in an age where personalization is a must).
When programmed correctly and verified with statistically significant analytics, AI and Machine Learning software should be able to beat the humans every time. Persado, for example, takes the guesswork and subjectivity out of marketing language because the machine always finds the most effective words to use. In other words, don’t challenge it to a game of Scrabble.
About the Author
Dan Gingiss is Vice President, Strategic Group at Persado. His 20-year career has consistently focused on delighting customers, spanning multiple disciplines including marketing, customer experience, social media, and customer service. He has held leadership roles at several big brands, including McDonald’s, Humana, and Discover – the last two of which he was a Persado client. He played a key role in Discover winning its first J.D. Power Award for “Highest in Customer Satisfaction”.
He is the author of the book, Winning at Social Customer Care: How Top Brands Create Engaging Experiences on Social Media, and is the co-host of the Experience This! podcast.
Dan has a BA from the University of Pennsylvania and an MBA from the Kellogg School of Management at Northwestern University. You can find him on Twitter at @dgingiss.