How American Express, DoorDash, JP Morgan Chase, and more are using AI to improve business

Business Insider

CMOs from companies like DoorDash, American Express, JP Morgan Chase, and more reveal how they're using AI.

Transcript

00:00I think we see AI doing incredible things for the business. A couple of

00:04concrete examples, if you think about our business and you think about the

00:07restaurants in our business, when a new restaurant joins StorDash, there's a

00:10cold-start problem, right? You're gonna have your first 30 days, how are people

00:13gonna discover you, are they gonna find orders? So a lot of what we do is we

00:16leverage the learnings we have from all the other restaurants on the platform

00:19that look like that restaurant, package that up as a set of insights to really

00:23help them understand how do I optimize my menu, how do I set up my photos, how do I

00:26set up my store, which hours should I run? So we're able to give them a ton of

00:29information that right out of the gate helps them be successful. We can look at

00:33AI from the perspective of photo coverage, right? So we can think about for

00:36all the stores on StorDash, which might be a small-medium business restaurant

00:40that doesn't have a lot of resources to take their own photos for every dish, how

00:44might we use AI to augment that? And if we have photo coverage on every item on

00:48the platform, that's gonna drive conversion because people eat with their

00:50eyes. So that's a super concrete example about how we think about AI in the

00:53business. And then from a marketing perspective, we're really leaning to it

00:57from the copy side. You know, we send out emails, push notifications, and there's a

01:02lot of testing that we do as a business to understand what's driving more open

01:05rates, what's efficient, what's effective. And AI really helps us to generate a lot

01:09of examples of these and put those in front of people in real time and

01:12optimize. AI has been such a powerful tool for us actually for many years.

01:18Starting back in 2010, we used machine learning to help our customers from a

01:25fraud perspective. But I'm also very excited about what we're thinking about

01:29with Gen AI sort of going forward. I think it's all about the

01:34ability to deliver personalized, customized experiences, as well as make

01:39the marketers job much easier so that they can then focus on creativity and

01:45innovation. So very specifically, we're thinking about Gen AI in a couple of

01:50different ways. That content generation that I just mentioned in terms of being

01:54personalized, so the ability to use text for email messages, digital assets, and

01:59websites, again, to make the marketers job easier so they can focus on that

02:04creativity. We're also thinking about using AI from an insights perspective, so

02:09the ability to synthesize an array of insights that allow us to make smart

02:14marketing decisions. And then lastly, and I talked about this already, but using AI

02:19to create a marketing assistant. So again, to make that end-to-end campaign much

02:25more seamless, allowing the marketer to do their job in a very creative and

02:30innovative way. You have some very real AI experiences with marketing. First,

02:35using it to understand our customer research better, or apply Gen AI across

02:38an entire base of insights and customer interactions and be able to actually

02:42test marketing ideas and messages against the research. One that's been

02:46really productive, of course, using Gen AI for content creation, but more recently

02:50video creation, which is incredibly cost-efficient. And we're able to use

02:53videos now for social media and other drops that we wouldn't have before,

02:57because it's too expensive or took too long. But my favorite is we used Gen AI

03:01very early on in our ad campaign, and we actually decided we're gonna test it on

03:05ourselves. Our campaign theme was be ready, because you have to be ready for

03:07whatever's next. So we just said, hey, let's try it on ourselves. We took a team

03:11of artists, and we took Gen AI, and we took the top four headlines out of the

03:14news that morning. We had Gen AI create a kind of a picture of those four

03:19headlines. It might look all jammed together. And then the artists had six

03:22hours to refine it and make it better and make it coherent. And we literally

03:25ran it on a billboard the next day. And we did that for ten days. So we woke

03:29up every morning, picked the headlines, used Gen AI, put the artists on it, and

03:32ran it. And we said, hey, let's be ready as well. And so that was a really cool

03:35experiment. We've been using AI across the business for years, right? We've been

03:39using it to price rooms based on demand we're seeing in the marketplace. We're

03:42using it to score our content to make sure it's meeting our standards for the

03:457,600 hotels we have all across the globe. And now we're using it to actually

03:48write copy for us in some cases at scale. Or translate things for all across the

03:52world. We're in 126 countries and territories across the world. So AI is

03:55helping us create better content, more meaningfully versioned. It doesn't take the

03:59creativity out of the humans, but it's helping us amplify our efforts. And of

04:02course, I'm a little bit biased, but I think as we see AI become more pervasive,

04:05the need for human hospitality, the human connection, is even more important

04:08than ever. When I think about AI, it's really about the journey that we've been

04:11on to blend human and digital in a way that's unique to State Farm. There's

04:17several applications in terms of how we're using it and how we're exploring

04:21using it. One is just to make our people better, faster, more efficient in doing

04:26their work. We found some opportunities in just simple things like something

04:31called Simple Insights, where consumers can go online and learn more about

04:35insurance. AI has helped cultivate that in a quicker, easier way. And those humans

04:40that used to write it are now able to do higher value things. And so that's a

04:44wonderful thing. And then we're also exploring ways through creativity. It's

04:47really important to us in terms of telling our story and how we can think

04:52about hyper personalization to deliver the right message at the right time to

04:56the right audience. So that they choose State Farm, they see it as a place for

05:00them. And when you think about our industry, fraud, scams, overdraft, we have

05:06to know when those things happen. And that intelligence helps us actually

05:09reach out to our consumers and our customers and say to them, here's a

05:13moment in time, particularly if you think about like holidays, for example, or tax

05:18season, where they are more prone and more vulnerable to those kinds of scams.

05:22And so for us to be able to use that artificial intelligence to be able to

05:26get to our customers, either before something happens, if we get signals, or

05:31frankly, right after something happens to be able to help people. For us, it's

05:35very, very helpful. Now, when I think about Gen AI more recently, we've

05:39actually used it in a couple of different ways. One, it's made us much

05:43more efficient. And so when I think about the role of the team set, for example,

05:48write all of our SEO articles, they are moving from writers to editors, where

05:55they may have spent days, for example, writing an article. Now what they're

05:59doing is they're prompting. So they're prompt engineers in the sense that

06:03they're actually prompting the intelligence to say, you know, here's

06:06what I want to write about. And then they're spending 20 minutes, 30 minutes,

06:1140 minutes, whatever it is, actually then editing these articles. So it really

06:16speeds up the time. And it allows us not only to spend our time on the more sort

06:22of value added things, but it allows us to actually produce more articles. And so

06:27that's really important for us. So the last one I'll share actually is really

06:30interesting. On the JP Morgan side, we do a lot of thought leadership client

06:35events, but not a ton of clients come to those events, right? But we want to share

06:39that thought leadership and all the conversations that happen with our

06:43millions of clients. And so historically, what we've done is we've taken, you

06:47know, the scripts from the conversations that are had on panels, for example, and

06:51we painstakingly go through them. And it takes sometimes a week to be able to

06:56package the summaries of those for our clients and then send them out. Now it's

07:01taking hours, because we can take those transcripts and the AI is really helping

07:06us distill all of that down, summarize it very quickly, so that we can then send it

07:11off within 24 hours and package it to our clients. So it's really just speeding

07:16up the process internally. It's making us more efficient. And then externally, it's

07:20making us much more effective and able to connect with our consumers much more

07:24deeply.

Category

Transcript

Recommended