In a world saturated with marketing messages, how can brands cut through the noise and create truly resonant experiences that go beyond transactions to create genuine, long-term customer relationships?
Agility requires a willingness to constantly adapt and evolve your strategies based on real-time data and customer feedback. It also demands a culture of experimentation and a commitment to iterating quickly on what you learn.
Today, we are in New York City at Contentsquare’s CX Circle and we’re going to talk about leveraging AI and personalization to drive success in direct-to-consumer marketing. To help me discuss this topic, I’d like to welcome Kelly Soligon, VP of Consumer Digital Direct Sales at Microsoft. Kelly, welcome to the show!
About Kelly Soligon
Kelly Soligon on LinkedIn: https://www.linkedin.com/in/ksoligon/
Resources
Microsoft: https://www.microsoft.com
The Agile Brand podcast is brought to you by TEKsystems. Learn more here: https://www.teksystems.com/versionnextnow
Register now for Sitecore Symposium, November 3-5 in Orlando Florida. Use code SYM25-2Media10 to receive 10% off. Go here for more: https://symposium.sitecore.com/
Don’t Miss MAICON 2025, October 14-16 in Cleveland – the event bringing together the brights minds and leading voices in AI. Use Code AGILE150 for $150 off registration. Go here to register: https://bit.ly/agile150
Connect with Greg on LinkedIn: https://www.linkedin.com/in/gregkihlstrom
Don’t miss a thing: get the latest episodes, sign up for our newsletter and more: https://www.theagilebrand.show
Check out The Agile Brand Guide website with articles, insights, and Martechipedia, the wiki for marketing technology: https://www.agilebrandguide.com
The Agile Brand is produced by Missing Link—a Latina-owned strategy-driven, creatively fueled production co-op. From ideation to creation, they craft human connections through intelligent, engaging and informative content. https://www.missinglink.company
Transcript
Greg Kihlstrom (00:01)
When we as consumers are saturated with marketing messages, how can brands cut through the noise and create truly resonant experiences that go beyond transactions to create genuine long-term customer relationships? Agility requires a willingness to constantly adapt and evolve your strategies based on real-time data and customer feedback. It also demands a culture of experimentation and a commitment to iterating quickly on what you learn.
Today we’re in New York City at Content Square’s CX Circle and we’re gonna talk about leveraging AI and personalization to drive success in direct to consumer marketing. To help me discuss this topic, I’d like to welcome Kelly Soligon, VP of Consumer Digital Direct Sales at Microsoft. Kelly, welcome to the show. Yeah, looking forward to talking with you here. Before we dive in though, why don’t you give a little background on yourself and your role at Microsoft.
Kelly Soligon (00:42)
Great. Thanks for having me, Greg.
Of course. Well, as you said, I lead our digital direct sales business. that’s a fancy way of saying that I lead our global e-commerce business for consumers. So if someone comes to Microsoft.com and they buy an Xbox or a Surface laptop, that’s my team running that consumer engine or the direct to consumer channel for Microsoft. We operate in over 180 markets. I have over 10 billion visitors annually to our consumer facing properties. And it’s a really fun role. been at Microsoft 23 years.
about half the time in the commercial space and about half the time in the consumer space. So really love what I’m doing.
Greg Kihlstrom (01:22)
Nice, nice, love it. So yeah, let’s chat. So you just wrapped your session here at CX Circle talking about AI and personalization for Microsoft’s D2C businesses, as you said. What’s the one thing you want people to walk away with the talk that you just gave?
Kelly Soligon (01:36)
Great question. I was able to share really where we’ve been focused in the last year around AI innovation and how that affects personalization as well. We’ve been doing some really interesting things in our voice of customer business. We get over 4 million pieces of feedback from customers annually, tons of different channels. could be customer service transcripts. It can be post-survey purchase feedback, et cetera. And so we’re using AI just to get clarity within.
that feedback. As you can imagine, 4 million, the quality varies pretty wildly. And so we’re using AI to help us identify high quality feedback and then quickly take action on it to improve our customer experience. That’s one of the areas that we’re really excited about. We also launched a virtual store assistant that’s a generative AI experience. It replaced a chat bot that was pretty archaic and is really, I think, meeting the needs of customers where they are today you’re able to ⁓ cross our shopping journey from shopping all the way to support. Customers are able to interact in a human way, ask questions, the ⁓ virtual assistant gains more clarity. can do, if you’re shopping for a laptop, for example, it can do compare charts right there in the chat stream and help the customer buy the device right there or get answers to their support. And so we’re excited about that. It’s been a really ⁓ great high satisfaction experience for our customers. And then in personalization, I’ve been really proud of the
the progress we’ve made, we’ve moved from more of a of like a experimentation sidecar, if you will, to having personalization embedded across our customer journey. And again, we’re using AI to help us do this. Cause there’s, mean, I only have so many humans on my team. And so of course, you know, we’re, we, we turned to AI to say, how can you look at our high intent signals, you know, like add to cart and help us to reorder product cards, help us show additional offers to customers.
And so that’s all been, you all of these three examples or how we’re using AI just to improve our customer experience and our operations of our e-commerce channel. It’s super exciting, you know, like this pace of innovation. It’s also making me think forward to the future of what is the role of agentic commerce? And, you know, how would we take our store assistant further? How would someone shop within chat GPT? So just a lot of opportunities for us to continue to focus and really rapidly evolve our customer experience.
Greg Kihlstrom (03:53)
Yeah, definitely. some people listening out there might be struggling with four million data points. You’re talking about four billion here, few more. this just bring you kind of touched on this already, personalization at scale. And we’re talking a massive scale at this point. And so I feel like we’ve been talking about one-to-one personalization for, at this point, decades. It feels like at least.
It’s, you know, with a combination of all of those data points being collected plus AI, how has that kind of changed the perspective on, you know, being able to really meaningfully do personalization at scale?
Kelly Soligon (04:34)
I don’t think without AI you can do it at scale. I mean, we were trying, you you’re doing kind of one-off experiments and you’re making progress there, of course, and you’re creating better customer experiences. But, the type of volume that we have, I think you have to be employing AI tools and experiences to do that. I’ve really liked the ability as well to go beyond clicks and like optimizing on the click to optimizing on a high value engagement. And so as an example for us, I would say
Adding to cart is a high value engagement or cart completion. And so I’ve been able to train the AI models to be able to show, this customer is high intent. Okay, how do we show them a different set of products that matches and is a winning offer for them? Now that before would have been super manual to do, but now with the AI models that we’re employing and the personalization there, we’re able to do that at scale and we’re seeing really measurable results for our business. And so I think you also have to make it a core capability for your business and something that as a leadership team that we want to invest in. We think personalization is key to unlocking better customer experiences, making your sites stickier and more engaging for the customer. So you have to be doing it at scale to get to that. And so it’s been a really fun journey. ⁓ I like to say it’s more of our core capability now versus something that a few people worked on. Everyone is invested in personalization and we’re seeing the results from that as well.
Greg Kihlstrom (05:57)
Yeah,
I mean, it’s one of those things where the numbers were always, you know, the research was always there, they bore it out. It’s just when you actually start doing it. again, you I mostly work with enterprises too. You know, when you actually start doing it, it’s like you uncover the silos, the hurdles, the, know, the scaling issues. So.
Kelly Soligon (06:16)
And I think there’s another point there I’m talking about, like, do more of it. There’s also like really understanding what the business impact of it is and determining where you want to put your chips. Because yeah, you could personalize everything and you could be running every A model, know, 24 hours a day. But is it really resulting in measurable business impact? And so I think that’s another learning we’ve had of what, you know, I talked about the high, you know, intent actions is focusing there because if not, you’re just kind of personalizing everything and it’s hard to parse out where you’re truly having impact. And so I think that’s something to consider. Yes, it’s great to have the ability to scale, but then think about where you can have the most impact from.
Greg Kihlstrom (06:56)
Yeah, yeah. So you mentioned agentic. let’s go to that. I feel like gen.ai was like all the rage. It still is, you know. New buzzword. Yeah, agentic is the term de jour, I guess. not only are they customer facing, but they’re also employee facing as well. Talk a little bit more about how agents have impacted the buyer experience and your approach.
Kelly Soligon (07:21)
Yeah, this has been one of the things we’ve been most excited about in the past year. We launched our virtual store assistant a little over a year ago. We’ve had over 1.7 million AI conversations with our generative AI virtual store assistant. And it’s really helping customers get to their end game faster. In the shopping space, it’s helping them compare products. It’s helping them get recommendations, configure a product, buying a laptop. There’s lots of options.
You know, I work in this space and I’m still like, hey, which version do I want? And so that store assistant can really guide a customer through what are you looking to do? Like how much memory do you need? How much processing speed do you need? In human terms, I’m in tech specs. And that’s one of the things I love about it the most. I can help people complete their purchase support. It’s helping people, you know, do the basics like order lookup and returns and shipping information but it can also then determine when that’s going to be most high value to hand you to a human. Because I think there’s still a role. Like the agents can do a lot, the block can tackle things, but there’s sometimes a spot where a customer is going to really benefit and have a better experience with your brand if they talk to a human. And so it can hand off very elegantly to a human and really get the customer to the resolution that they need. So we’ve been loving it. We have plans to expand it internationally. We’re looking at new.
modes, you know, it’s chat based typing right now, but like what role would voice play in that? What role would vision potentially play in that? And so we’re excited to keep expanding there. And we’ve also are thinking about how when you come to Microsoft.com, would you just interact with this agent instead of maybe clicking through a navigation or searching in the site, you would just interact with an agent there to find, you know, maybe shopping information, maybe corporate information. And so we’re excited about the future of that within our own site experience.
Greg Kihlstrom (09:10)
So taking into account agents operating optimally, what are some of the important elements that you need to take into account to set things up for success?
Kelly Soligon (09:20)
Yes, they can go wrong. I think we’ve all seen some public examples of that. And so, I’m not a technologist, so I know that ⁓ there’s a lot more under the hood here. But I would look at three different elements when you’re creating an agent. Really being first upfront about what your use case is. Clear intent. Give the agent some guardrails, because without that context, you don’t know where it’s going to go. And so I think you have to start with, what is the use case?
For me, if it’s in a customer experience, what are the parameters of that where I want it to operate? Second, I think you got to look at data quality and governance, like garbage in, garbage out in the old data world. It applies here as well, And so you have to really be thinking about what is your data model, your governance of that to ensure accuracy, to ensure transparency, privacy, et cetera. So I think that is the second pillar. And then finally, I would say the humans are still important. Like you can’t just set it and forget it. You need to have the right level of human, you know, monitoring and checking the customer, especially when it’s in a customer experience, checking that experience and saying, Hey, is this what we wanted to represent as a brand? Is this what we wanted customers to, you know, to see next in their journey and checking that and then optimizing, tweak the model, tweak the parameters and optimize from that. But I really think you have to like,
You think about those three areas, clear intent and use case, what’s our data model and the governance of that, and then how do we monitor it and optimize it as the CX leaders.
Greg Kihlstrom (10:55)
Yeah. So customer feedback, I mean, as we’ve seen today, it comes in many forms. Can you share a moment where voice of the customer data, like surveys and messages, told you one thing while behavioral data told you something else?
Kelly Soligon (11:10)
Sure, I love voice-to-customer data. I get a lot of it too, which is great. And having the AI tools to be able to really identify the high-quality feedback is important. You know, as I thought about this question, we had an example of where we were seeing from our post-purchase surveys, five stars, this is great. So seeing that, I’m like, okay, this is a clear winner. But then I was looking at more of our behavioral and transactional data and seeing that higher than average return rate on the product. And so it’s like you’re hearing from the customer’s this, but then I’m seeing their actions be different. And I also find, I don’t know if you find this too, like the people who fill out surveys, for the most part, they’re either super happy or they’re super mad. Exactly. And so you have to take that with a grain of salt as well. And so it’s a data point, but then how do you dig further and tie that data point then to like some more, you know, behavioral,
Greg Kihlstrom (11:50)
It’s a very particular subset.
Kelly Soligon (12:03)
data that you see or your business data. If I see high returns, I’m like, oh, we got a problem there. And so I think it’s really marrying those things. like, you know, data, this is true of like stats. I remember this from a stats class. They can tell any story you want. And so it’s really, I think, coupling the data you have with what you’re seeing in your customer, true customer behavior, as well as in your business results and putting that puzzle together and just say, okay, how big is this problem or how small is this problem? Hopefully, you know.
Greg Kihlstrom (12:32)
So what is, I know every situation is going to be different, but like do you have a thought process or rule of thumb of like how do you know which sources to listen to when and stuff?
Kelly Soligon (12:43)
It’s not like an exact science. I will say we have used an AI model that identifies high quality feedback. So it’s specific, it’s actionable, and it’s linked to a real experience. If somebody gives me a piece of feedback that says slow, I’m like, what can I do with that? But somebody’s giving me some actionable feedback that’s specific. So that is our definition of high quality. Our AI model picks that up and puts it at the top of the list for our analysts to look into further.
We further than married that with what we call basically high value engagement actions. And so I would look at something like add to cart and cart completion and average order value, number of products in the cart. And so those are some high intent and high value engagements for us. And so we’re saying, Hey, the high value feedback that is also really critically linked to one of our KPIs there around add to cart or cart completion. Boom, that goes to the top of the list.
Because there I see I have a real opportunity to fix something for customers and I have a real opportunity to probably make some more money. And so that’s the mental model we’ve been using to kind of parse through all of this. And then you have to say, can I fix it? There has to be the realism factor there. Sometimes it can’t fix everything.
Greg Kihlstrom (13:57)
It be in the top, I just envision a quadrant, but like it might be in the top right, but it’s gonna take, you
Kelly Soligon (14:03)
A year. Exactly.
And so then you get creative like what else could we do? You know, I think that’s where some cool innovation and breakthrough thinking can come of like, hey, the standard way of doing this is going to take us a long time. What can we do differently here that might be better for the customer?
Greg Kihlstrom (14:18)
Yeah, yeah. So, you know, we’ve talked quite about AI being used in quite a few different ways here. What’s prediction for, you know, how AI is going to change e-commerce over the next few years? Five years feels like way too far to like… I can’t… Let’s say a year. Let’s say we’re having this conversation a year from now. What are we talking about here?
Kelly Soligon (14:38)
Yeah, I just think about my own searching behavior as a consumer and how much that has changed in the last 12 to 18 months. I used to open up a web browser and search for a product. Maybe I’d go to directly the brand if I knew what I wanted or to a retailer site, but I would search within their site, that type of thing. Now I open up Copilot or I open up ChatGPT and I’m doing my searching right there. Like what’s the best laptop for me? What hotel should I book in New York City? Looking for somebody to give me the answers quickly. And so I look at just how my own searching behavior has changed in the last 12 to 18 months and say, if I’m experiencing that, millions and zillions of other people are experiencing that as well too. And so we in e-commerce have to adapt for that. I truly believe that shopping in the future is going to be agent driven or influenced to some extent. You saw OpenAI announced yesterday, they are aogenic commerce within ChatGPT partnering with Shopify and Stripe for the payment side. That’s brilliant. I think that is where shopping is going to be. You’re using ChatGPT, your personal AI assistant to help you determine which laptop to buy. And then you’re just buying right there in the stream. And I would imagine a great customer experience where I can go back to that chat and I can see my order information. I’m getting my shipping information from when in there. You know, it’s a really great customer experience.
And so I really think we’re preparing to say, hey, gosh, all these AI agents, chat, GPT, co-pilot, Gemini, you name it, these are great new traffic sources for us because people are using this as part of their daily life to get answers, to find information, to do research on things they want to buy. Why not enable purchase within there? And so for me, I’m really thinking about what do I need to do with APIs? What do I need to do with my data to make it discoverable? Bye.
these personal AI agents that people are using and make me shopping should be seamless. I’m an e-commerce leader. I’m like, it be so easy to shop. And that to me is a new easy channel for customers to shop through. And I think it’s happening fast. Like, just like you said, like no way we can predict five years from now. Like I think this is happening so rapidly. And so just sitting back and waiting, I don’t think as an e-commerce leader, you can do that. You got to have some chips on the table and be experimenting and learning and you know, not everything’s gonna work, but what do you learn from there and take forward into the next rev? But I think our customer experiences are gonna look a lot different very quickly.
Greg Kihlstrom (17:13)
Yeah, yeah, well, we’ll have to talk in a year and compare our predictions here. So that’s great. well, Kelly, thanks so much for joining today. One last question before we wrap up here. What do do to stay agile in your role and how do you find a way to do it consistently?
Kelly Soligon (17:28)
great question. You know, I think for me, has to start with I have to have energy. think moving fast can you know, it can wear you out if you don’t take care of yourself. And so I’ll start from kind of a personal side on this. I think it is very critical for all of us to make time for ourselves for our well being, mental, physical, you know, spiritual, emotional. And so I really for me, I know I do my best work when I’m energized. And so
I exercise every day. have to take that walk. I have to take that class because that refuels me and gives me the energy then to be put in these positions where I have to be agile. And I’m getting some stress, but OK, I’ve taken care of myself. it’s kind of like you put your mask on first because you’ve got to take care of yourself first. So that applies to all of life for me, not just work. But then second, I’ve been really focused on curiosity. And how do I expose myself to different perspectives?
so that I’m constantly learning. And so I really try to actively dedicate time to meeting with people within Microsoft, meeting with people within my industry, but also just meeting with people outside of my industry. Because I feel like some of my greatest ideas and thinking have come when I’ve talked to somebody who’s in a completely different space than me and they’re like, here’s what we’re doing with our customers. I think, that can work. There’s that kernel in there that I could bring to my business and make it relevant to what we’re doing.
I really try to block out some time. try to have like coffee dates two to three times a week, meeting different people. Like I will take, know, anybody who reaches out and say, want to learn about your business. I’m like, yes. And I want to learn about yours, you know, cause I just think being curious and exposing ourselves to different perspectives is where you kind of get the best quality of work. And, know, if you keep listening to yourself and hearing the same thing over and over again, you get the same results. That’s not where breakthrough comes. And so I feel like just really been trying to lean in on curiosity, you know, to the extreme lately.
