Artificial intelligence hype was pervasive last year and the reality of A.I. is white-hot right now. In fact, 61 percent of enterprises say they implemented A.I. in their organization in 2017 — up 31 percent from 2016. And, 791 public companies mentioned A.I. in their earnings calls in Q3 2017.
What does this mean? It means that A.I. is nearing “big data” or “blockchain” levels of notoriety. Soon, even the most unlikely companies will claim to use the technology (imagine Long Island Iced Tea becoming a blockchain company) to ride the zeitgeist and get some good press. In fact, TGI Fridays is using A.I. to automatically customize cocktails according to your preferences, so Alexa can morph you into a mixologist in the comfort of your own home. Now, that’s a practical use of A.I. and a fantastic brand extension experience.
While the A.I. hype is real, so is the technology. The applications for artificial intelligence are endless, powerful, and deployable across an entire organization. Some of the applications today are awe-inspiring — and some are simply about efficiently scaling with optimal velocity.
But it’s not magic. Companies need a focused approach for how to use A.I. based on their desired outcomes. In fact, companies that currently use A.I. have begun to see how this technology can yield efficient results in the form of increased customer engagement and retention.
This week, I participated in a panel at the VentureBeat TRANSFORM conference for digital marketing executives, to discuss how businesses use A.I. to improve customer insight and performance marketing.
Photos of event here:
I was honored to join Ashu Garg, General Partner at Foundation Capital; Jasjeet Thind, Vice President of Artificial Intelligence at Zillow Group; and Matthew Sadofsky, Director of Growth Marketing at Tilting Point on the panel. Their ideas and commentary on the A.I. landscape were insightful.
A.I. is Driving Better and More Unique Customer Experiences
As the CEO of Sensai — an A.I.-powered social media marketing platform that delivers actionable recommendations on content, platform, and timing for engaging and growing your audience — I’ve seen first-hand how the technology can deliver insights to deeply personalize one’s experience, impact, and results while saving precious time by negating much of the grunt work or overly complex tasks.
Good A.I. tools enable your company to market to individuals, rather than personas, based on what they need, when they need it. Furthermore, robust A.I. can anticipate that need beforehand.
During #VBTransform, one of the more notable examples of this level of personalization is TGI Fridays — yes, the fast-casual dining chain.
You might not think of TGI Fridays as a candidate for leveraging A.I., but Sherif Mityas — the Chief Experience Officer — has used the technology to recommend signature cocktails, based off their personal ordering patterns.
“The goal is to increase customer engagement and, subsequently, revenues,” said Mityas in a Wall Street Journal report. “A.I. allows the brand to really, truly differentiate ourselves again.” (See the full article here.)
Brian Hovis, Vice President of Marketing at Nordstrom, also shared how A.I. not only transforms the customer experience in Nordstrom stores, but also drives the creation of completely new store concepts that take advantage of different metro geographies — which delights their customers.
Chatbots also came up time and time again as an easily implementable and highly effective application of A.I. A fireside chat between Mamie Peers of the Cosmopolitan of Las Vegas, and Clara De Soto, founder of Reply.ai, was particularly inspiring. For the Cosmopolitan, chatbots have driven a stunning increase in revenue per guest via their chatbot they’ve named Rose. Rose’s voice — confident, intriguing, alluring, sophisticated — aligns with the Cosmopolitan’s brand, and drives a 38 percent increase in customer satisfaction among those who converse with her.
As a woman in tech, I am not a fan of Uber due to its ongoing sexual discrimination and harassment offenses (I use Lyft instead), but I was impressed by the clarity of vision and clear articulation of Uber’s A.I. approach infusing artificial intelligence throughout the entire customer journey with acute focus on onboarding. The simplicity of the slide presented by Uber’s Director of Product, Jairam Ranganathan, was a great example of how to make a sometimes nebulous and evolving technology readily understood. I only wish we were able to hear more about their advancements in driverless car and other smart mobility developments.
Perhaps one of the most entertaining presentations was Vanja Josifovski, Pinterest’s CTO who describes the platform as a visual discovery engine as opposed to a social network. He said they maintain the original vision of the founders to encourage people to spend time offline. I am curious how that translates during advertising revenue and KPI discussions with investors, so I’ll have to ask them when I next see them. It’s certainly laudable in the face of so many companies that aim to exacerbate digital addictions. I suppose if you’re spending more time offline then you may be capturing and sharing more inspiring visual photos to share and pin.
A.I. Can Replace Your Employees. But It Can Also Help Retain Them
Artificial intelligence can sometimes be a pejorative term — especially when it’s in reference to employees and hiring. Most people are scared that the technology is going to render their job obsolete.
And, yes, bots are replacing many of the repetitive tasks of customer service, but not always with the same purpose.
Stewart Rogers of VentureBeat detailed how two of China’s biggest companies are using A.I. with wildly different approaches. First, there’s Xiaomi, who used bots to cut costs by replacing 66 percent of their call center staff. But Tencent used A.I. to reduce demanding and repetitive work, making its staff more effective and happy, so it could retain them. Rogers added, “So the lesson is: if you don’t want A.I. to replace you, work for nice people.”
When is It Time to Build Rather than Buy or Partner?
On our panel, I was joined by some really smart people who work for much larger companies than my early-stage startup. Yet, I’ve also served huge enterprises like Viacom and the BBC, so I’ve been on their side of the fence.
A critical point for any business is knowing when to bring capabilities in-house. Paying 10 percent of a small marketing budget to an agency may make sense, but as your budget grows, it may become more effective to hire experts in-house. Matthew Sadofsky, who runs performance marketing for multiple mobile games, said it well: “We’re spending $3 million a month acquiring and retaining customers. Before we got to that point, I realized, I could hire some really smart people for what we’re paying the agency.”
We also got a question about best practices, and I was happy to take that one on, since we have a ton of learnings about that from Sensai. Best practices are only a starting point. Every company has a unique audience, and they have to test and learn to understand how to launch effectively.
A.I. is Ready for You
Overall, the vibe at VentureBeat TRANSFORM was very encouraging. We’re seeing that businesses of all sizes are leveraging accessible A.I. tools to improve their business performance, retain employees, and grow revenue sustainably.
Amir Khosrowshahi of Intel spoke on the conference’s second day about how Intel is democratizing A.I. by reducing the computing power required. Just as Moore’s Law has helped Intel reduce computers from something you store in an office basement to something you wear on your wrist, they’re striving to bring A.I. to more people in more places with slimmer requirements. Accessible A.I. is emerging.
It’s not too early to start experimenting with A.I. tools. At Sensai we’re focused on delivering A.I. insights to SMBs that would otherwise take hundreds or thousands of humans to gather. This levels the social media playing field, which can get crowded with big budgets of enterprise level corporations.
At Sensai, it’s an honor to be part of the artificial intelligence revolution, right when it’s catching fire. We’re excited to see more A.I. applications come to market, and hopefully we’ll see even more female founders entering the space soon.
How are we implementing A.I. today?
At Sensai, we are building an accessible and affordable A.I.-driven platform and products for businesses and artists. Some examples of of A.I.-driven insights include:
- Best time to post recommendation: As I mentioned above, best practices are only a starting point. Time to post is critical on algorithmic-driven platforms, but it’s not one-size-fits-all, as many social media practitioners suggest. Not until you can sift through the myriad data of when one’s audience is active and engaging with social posts and to what degree, as well as multiple dimensions including customer verticals, their objectives and interests, their post contents and topics, paired with their audience composition — demographics, geolocation, social behaviors, seasonality, temporal, their subject of interest at the time, and more. Deep clustering and segmentation by A.I. and machine learning help surface optimal times to post, individualized for one customer.
- Topic modeling: If you were a content marketer, or more specifically a social content marketer, how could you predict what would yield your expected return? With topic modeling and clustering model using A.I., NLP, and machine learning, we’re able to identify, classify, and determine specific content topics which would ignite interaction from your audience.
- Hashtag recommendation: As many social specialists may already know, hashtags are used in determining how and to what degree social contents are being tagged, ranked, searched, discovered, and rendered to social media users. But how would you know which hashtags used in your posts would be relevant and would return an optimal level of interaction by your viewers beyond just the contents being ranked, searched, and discovered? The use of A.I., NLP, and machine learning models enables streaming of petabytes if not zetabytes of unstructured data from social media platforms, which we can then analyze, cluster, and rank, in order to surface hashtags that bear semantic relevance to the content. But you have to do it quickly: the 4 V’s of social data — volume, variety, velocity, and veracity — need to be executed in real-time while the hashtags are relevant and current.
I am looking forward to following up with many of the artificial intelligence leaders I met at TRANSFORM, so that we can all elevate our learning and collaborate in the ongoing development of new tools that improve life — maybe even humanity.